Federal Regulatory Management of the Automobile in the United States, 1966–1988

Home , An Unreasonable Man, Clarence Ditlow, George Dantzig, Studebaker Canada

FEDERAL REGULATORY MANAGEMENT OF THE AUTOMOBILE IN THE UNITED STATES, 1966–1988

LEE JARED VINSEL

DISSERTATION

Presented to the Faculty of the College of Humanities and Social Sciences of Carnegie Mellon University in Partial Fulfillment of the Requirements For the Degree of

DOCTOR OF PHILOSOPHY

Carnegie Mellon University

May 2011

Dissertation Committee: Professor David A. Hounshell, Chair Professor Jay Aronson Professor John Soluri Professor Joel A. Tarr Professor Steven Usselman (Georgia Tech)

ii Dedication

For the Vinsels, the McFaddens, and the Middletons

and

for Abigail, who held the ship steady

iii Abstract

Federal Regulatory Management of the Automobile

in the United States, 1966–1988

by LEE JARED VINSEL

Dissertation Director:

Professor David A. Hounshell

Throughout the 20th century, the automobile became the great American machine, a technological object that became inseparable from every level of American life and culture from the cycles of the national economy to the passions of teen dating, from the travails of labor struggles to the travels of “soccer moms.” Yet, the automobile brought with it multiple dimensions of risk: crashes mangled bodies, tailpipes spewed toxic exhausts, and engines “guzzled” increasingly limited fuel resources.

During the 1960s and 1970s, the United States Federal government created institutions—primarily the National Highway Traffic Safety Administration within the Department of Transportation and the Office of Mobile Source

Pollution Control in the Environmental Protection Agency—to regulate the automobile industry around three concerns, namely crash safety, fuel efficiency, and control of emissions. This dissertation examines the growth of state institutions to regulate these three concerns during the 1960s and 1970s through the 1980s when

iv the state came under fire from new political forces and governmental bureaucracies experienced large cutbacks in budgets and staff.

While most previous studies of regulation have focused either on biographies of regulatory visionaries (a.k.a. policy entrepreneurs) or on legislative histories, this dissertation examines how the federal government built bureaucratic organizations and administrative capacity to regulate and force change in the automobile through performance standards. Employees of these agencies helped shape automobile design by creating routine regulatory procedures that intervened in the longstanding traditions of automobile design. Only by examining these micro- practices of governmental power, I argue, can we understand how regulatory regimes have truly influenced their intended objects. My dissertation examines how these institutions developed, learned, and evolved, with an eye to how these transformations shaped technological change in the automobile industry. By examining the mundane world of federal test procedures, scientific studies, agency meetings, and administrative hearings, I will show how low-level bureaucrats formed new networks between government and industry, established the state of the art in automobile technology, and forced innovation in automobile design.

v Acknowledgments

Over the three years of this dissertation’s making, a number of institutions have supported my research and writing. Two National Science Foundation grants funded my research and studies early on, the grant that created the Climate Decision

Making Center and the grant titled “The Socio-Political Construction of Technologies under "Technology-Forcing" Regulations: A Tale of Two Automotive Technologies,

"One" Government and "One" Industry.” I was pleased and honored to receive a

Dissertation Improvement Grant from the National Science Foundation and the John

E. Rovensky Fellowship in American Business or Economic History for the 2009–

2010 academic year. Two dissertation workshops helped me improve and winnow down my work: Sheila Jasanoff and Clark Miller challenged me to clarify and refine my thinking during the Social Science Research Council Dissertation Proposal

Development Fellowship in 2008. Pamela Laird, Mary O’Sullivan, and Steve Tolliday also helped me consider new avenues of research and unexplored corners of my topic at the Business History Conference’s Oxford Journals Colloquium in Business

History in 2009. (Thanks, too, to Jenna Alden for organizing the BHC Colloquium in

Milan, Italy.) Finally, in 2008, a Ford Motor Company Research Funding grant assisted my continuing work. Thanks also to the librarians at Carnegie Mellon, especially Kara Kreger and Sue Collins, who always effectively answered my questions.

A number of people inside and outside government helped me in locating both people and records. The librarians and record managers David Doernberg

(Department of Transportation) and Kirk Nims (Environmental Protection Agency)

vi led me to important archival discoveries. My work at the EPA may have gone nowhere at all if not for Joseph Somers, who both shared recollections from his near photographic memory and gave me the contact info of several retired EPA staff members. The most important of these contacts was Eric Stork. Eric helped me in innumerable ways, offering his time whenever I requested it, allowing me to interview him several times, sharing his large collection of press clipping and private papers, and reviewing my chapter drafts to ensure that I did not botch some matter of fact. I owe him a great deal. Karl Hellman also helped by trusting me with his memories and with his large collection of EPA memos and reports, which he gave me in the summer of 2010. Thanks also to other present and former EPA staff members who allowed me to interview them, including Ernie Rosenberg, Janet

Auerbach, Richard Lawrence, Joseph Merenda, Charles Gray, Rich Cook, and

Katherine A. Sargent. Though I chose in the end not to rely on oral histories in the chapters on federal auto safety regulations, I could not have negotiated the landscape of automotive crash safety without several interviews that gave me some insight into the workings of the National Highway Traffic Safety Administration. I am particularly indebted to Michael Finkelstein, Marilena Amoni, Joseph Kanianthra,

Sam Daniels, Clarke Harper, Harry Thompson, Jim Simons, and John Hinch. Through interviews, Bill King and Richard Klimisch gave me insight into the auto industry’s side of the story, and Steve Plotkin helped me see the complexity of auto regulation.

It has become a convention for authors’ to list a number of commentators who have improved their work, and then to write something like, “All errors are mine and mine alone.” We should equally say this about our selves. I have been

vii blessed to have many people (and institutions) in my life who have helped shape me in positive ways—including the part of me that wrote this dissertation. All remaining errors (of self) are mine and mine alone.

At Carnegie Mellon, I benefited greatly from being a member of the Climate

Decision Making Center, an interdisciplinary group dedicated to studying climate change policy- and decision-making under uncertainty. The professors and students of that group formed my thinking in ways that I did not come to realize until much later. I am indebted to Granger Morgan, Lester Lave, Marija Ilic, and Jay Apt. The students of the CDMC always enlivened me and led me to think about new subjects.

I am particularly thankful to (and miss) Inês Azevedo, Vanessa Schweizer,

Constantine Samaras, Josh Stolaroff, and Andy Grieshop. I was also lucky to be able to sit in on the seminars of SETChange (Strategy, Entrepreneurship, and

Technological Change). Being around Steven Klepper and Francisco Veloso taught me a great deal, and their student Leonardo Reyes-Gonzalez left an indelible impression. In the History Department, Steve Schlossman guided my early forays into the archive in the graduate research seminar. Paul Eiss kept my theory chops honed. Scott Sandage taught me not only about the arts of writing history and teaching but also about the art of living. It’s rare to find someone who can teach you so much, and I can’t thank him enough for it. Kevin Brown, Cian McMahon, Susan

Spellman, and Patrick Zimmerman have been true friends. Susan has frequently tightened my prose, challenged my hazy thoughts, and, generally, kicked me when I needed kicking. Thanks, sister. Thanks, finally, to the History Department’s staff members, especially Natalie Taylor and Gail Tooks.

viii I am very lucky to have found such an excellent dissertation committee in

David Hounshell, Jay Aronson, John Soluri, Joel Tarr, and Steven Usselman. Jay led me in an important guided reading on Science and Technology Studies and has helped me along in other essential ways. John also carried out an extended guided reading with me on global environmental history. John has always asked me very hard and smart questions and has pushed me to broaden my thinking on all levels.

Joel has always encouraged my broad interests and searching mind. He honed my historical skills by allowing me to assist his research and by teaching me what he was thinking as I did so. I thank him for always being supportive. Joel will have always been my intellectual grandfather . . . or, perhaps more appropriately,

Godfather.

As an undergraduate, I trained in philosophy and imagined that I would always remain in that field. Then, fatefully, I happened upon a book called

Regulating Railroad Innovation by a man named Steven Usselman. Steve was kind to reply a random fan letter from a kid in Chicago but kinder to become a mentor and friend. His work continues to provide a model of great historical thinking. Steve put me in touch with historians of technology in Chicago, who brought me into the fold.

Thanks especially to Richard John, who trusted a stranger to assist his research and taught me a great deal. And thanks to the regular attendees of the Newbery Library’s seminar in the history of technology, particularly Tom Misa. I have benefited immensely from being a member of that happy republic, the Society for the History of Technology, and of the Business History Conference. Thanks particularly to Meg

Graham, Hugh Gorman, Bill Leslie, Daniel Holbrook, Chris Rosen, Arwen Mohun, and

ix Bernie and Jane Carlson. I am blessed to have found an excellent group of peers in these societies—my own little intellectual “school.” Thanks to Hyungsub Choi,

Barbara Hahn, Eric Hintz, Eric Nystrom, Dominique Tobell, Ben Waterhouse, and especially to my historical soul mate, Andy Russell.

Three men have been instrumental in mentoring my intellectual life. After a random inquiry I made after watching the movie Clueless, my high school English teacher, Mark Wilson, gave me some Nietzsche to read in my sophomore year. He then mentored me as I moved through that irascible German to the existentialists and, finally, by my senior year, the structuralists and post-structuralists. He answered my every curiousity and counseled me in the life of the heart as well as that of the mind. I know that I would not be who I am today were it not for “Mr.

Wilson.” In my days as an undergraduate, Professor Bill Schroeder became the person who answered my long list of philosophical questions and who lectured with such furious enthusiasm that I became assured of earlier inklings that I wanted to get a PhD. Most of all, Bill taught me that study required the fire of passion as much as it did the cool order-making of reason. He remains a fast friend who continues to inspire me in important ways.

Finally, Professor David Hounshell has taught me in more ways than I can enumerate. He has always patiently humored my whim and curiousity, while fostering in me whatever discipline, rigor, and clarity I possess. His encyclopedic knowledge of history and historiography answered my every query. Most importantly, David took on a student who wrote broken and mangled prose and turned him into someone who is occasionally capable of producing a clear English

x sentence. For all these things and many more (which he would happily itemize for you), I owe David a great deal and thank him for allowing me to be his apprentice.

If my intellectual mentors have been men, all of my spiritual ones have been women. I owe much to Rev. Deborah Warren, Rev. Mary Louise McCullough, Bonnie

Thurston, Rev. Maureen Dickman, Sr. Karen Freund, and my dear late friend Sandra

Mellen.

My hometown, Joliet, was a hard place that rarely fostered aspirations.

Several adults helped me along the way, however, including Mr. Michael Reilly, Carol

Sossong, the Coughlen family, and Pat McGuire. My friends—“the circus”—always gave me hope. I thank Ramon, Kahlil, Ben (HBH), Nina, Ellen, Eliot (forever), Jean,

Bill, Andrew, and my brothers, Mike and Casey. Since those days, I have been lucky to befriend and be challenged and supported by Thom Moran, Darcy Bean, Dan

Pride, Matei Costinescu, Ben Wachter, Brian Beahan, and Sarah Ratermann Beahan.

Thanks to Sarah Ratermann Beahan and Emily Murphy for helping me edit my dissertation at various points.

I dedicate this dissertation to my family, the Vinsels, the McFaddens, the

Middletons, and to my wife, Abigail Middleton. My parents provided a home that encouraged learning, thinking, and the development of eccentricity. To them, I can only say, thank you, knowing that no words will ever suffice. Much love to my sisters, Hannah and Rachel, and especially to my brother, Jaco, who always lifts my spirits. My nuclear family is the confluence of two wonderful streams. I thank the

Vinsels and McFaddens for always being there, and I dedicate my work especially to

xi the memory of Grandma Lydia, Grandpa Loren, and Uncle Mike, and in tribute to my namesake, my Uncle Lee. I hope I have done them proud.

Abigail, we met when we both knew that we wanted to pursue graduate education but only had vague ideas of what that would entail. Thanks to the

Middletons and, especially, your parents for raising such a lovely daughter and for supporting us in every way. Thank you for the quiet, private world we have built together with Baron and Gypsy. Thank you for being strong and keeping my head level as I threw myself into my work, and thanks, most of all, for loving me regardless of the outcome it.

xii TABLE OF CONTENTS

Dedication iii

Abstract iv

Acknowledgements vi

Introduction 1

Chapter 1—Making the Crash Barrier: Medical Authority, Engineering Culture, and Bureaucratic Practice in American Automotive Safety 30

Chapter 2—Constructing Standard 201 and the Limiting of Federal Automotive Safety Regulations 62

Chapter 3—The Banality of Justice 113

Chapter 4—Organizing Federal Automotive Emissions Control: Bureaucratic Change as a Strategic Enterprise 194

Chapter 5—Establishing the State of the Art: Administering the Clean Air Act of 1970, 1972-1973 265

Chapter 6— The Long Road to Federal Fuel Economy Standards: Capabilities- Building, Bureaucratic Self-Defense, and the Fine Art of Gamesmanship 323

Epilogue 380

Bibliography 406

xiii

Introduction

Sometime during the week of January 16, 1967, in New York City, Henry Ford

II, the grandson of the founder of Ford Motor Company, gave a speech before the

National Retail Merchants Association, a trade group for retail businesses. Ford II’s theme was the dangerous precedent that federal regulation was setting. He warned the audience that there were “strong signs” that recently imposed regulations would be only the beginning.1 The federal government was beginning to take control of automobile design, he said.2 Soon these regulations could lead to “a total ban on the internal combustion engine, the development under government auspices of alternative power sources, and public policies to force people in metropolitan areas out of private cars and into public transit facilities,” he cautioned. If audience members were deluding themselves that this was only a problem in the auto industry, they should wake up, Ford II insisted. “Similar things are happening in the food and drug industries, in finance and insurance, in advertising and retailing. In virtually every industry, government looms increasingly large between business and its customers, its employees, its shareholders, and the general public.” Ford II reminded the audience not to turn the matter into an issue of “good guys vs. bad guys.” The point, he claimed, was to draw a firm line between the responsibilities of

1 The following statements of Henry Ford II are quoted from “Ford Cites ‘Strong Signs’ of New Federal Inroads,” Automotive News, January 16, 1967. 2 The Automotive News ran several stories on the production of automobiles in the USSR’s command- and-control economy during this period, often emphasizing that automobile design was state controlled in the Soviet Union. In this way, talk of the US government mandating automobile design could be tantamount to “red-baiting,” insinuating that federal programs were, in effect, “socialism.”

1 2 government and the responsibilities of industry. In Ford II’s view, this line was the very thing government was trying to undo.

A few weeks earlier, on January 2, 1967, Automotive News, a trade newspaper whose primary audience was car dealers, announced that its editorial staff had voted “auto safety” as the top news story of 1966.3 The second-place story of 1966 concerned Ralph Nader, a young lawyer who was a chief architect of federal auto safety regulations. In 1965, Nader had published a book, Unsafe at Any Speed, an exposé about the lack of safety features and the presence of “designed-in” dangers in automobiles. The book paid special attention to the Chevrolet Corvair, a sporty coupe that Nader said had “some remarkable characteristics”: “It’s one of the few cars I know that can do the Bossa Nova on dry pavement and the Watusi on wet.”4

Passage of the federal auto safety law, the Traffic Safety Act of 1966, was virtually guaranteed when Congress learned that General Motors had hired a private investigator to examine Nader’s private life, including his sexuality and whether he was anti-Semitic. The third-place story at Automotive News had to do with the decline in car sales in 1966. A series written by Joseph M. Callahan, the engineering editor of Automotive News, came in fourth. The series, titled “The Billion-Dollar

Smog Hoax,” eventually ran to over thirty pieces. In it, Callahan skewered what he saw as the fraud science of smog and its health effects, claiming both that scientists had not established the dangers of smog and that automotive emission controls were unnecessary. Thus, three of the editors’ top four choices for 1966 were related

3 John K. Teahen Jr., “’Safety’ Voted Top News of ’66,” Automotive News, January 2, 1967. 4 Quoted from archival video footage of Nader speaking in the documentary film, An Unreasonable Man (2006).

3 to federal regulation. In the same January 2 issue, the publisher of the paper, Pete

Wemhoff, wrote an open letter to auto dealers, warning, “The auto industry is in trouble—deep trouble, and not solely because of declining sales. The industry has again become the favorite ‘whipping boy’ of Washington politicians and bureaucrats, do-gooders, and some elements of the press.” He then admonished the dealers to contact their congressmen. “Today’s crisis in the auto industry is not just the makers’ problem—it’s yours, too.” Clearly, federal regulations threatened the auto industry—from the most powerful inheritor of an auto manufacturer to the lowly auto dealer.

This dissertation examines the history of federal regulation of the automobile in the United States around crash safety, emissions control, and fuel efficiency. It builds on the work of other historians, who have been studying regulatory history for decades.5 Where most previous histories of regulation have focused either on the biographies of “policy entrepreneurs,” who imagine a new regulatory regime, or on the legislation that enabled a regulation, this work takes a different tack: it examines how federal civil servants managed the regulations with which they were charged. I examine two aspects of this process, namely how federal administrators built the capabilities of their agencies—both human and physical—and then how they used those capabilities to create and enforce regulations. I focus on two

5 Thomas McCraw “Regulation in America: A Review Article,” The Business History Review, Vol. 49, No. 2 (Summer, 1975), 159-183; idem., Prophets of Regulation: Charles Francis Adams, Louis D. Brandeis, James M. Landis, Alfred E. Kahn (Cambridge, Mass.: Harvard University Press, 1984); David Vogel’s “The ‘New’ Social Regulation in Historical and Comparative Perspective” in Regulation in Perspective: Historical Essays, Thomas K. McCraw, ed. (Cambridge, Mass.: Harvard University Press, 1981), 155-185; Richard H. K. Vietor, Contrived Competition: Regulation and Deregulation in America (Cambridge, Mass.: The Harvard University Press, 1994); Stephen W. Usselman, Regulating Railroad Innovation: Business, Technology, and Politics in America, 1840–1920 (Cambridge: Cambridge University Press, 2002).

4 regulatory agencies: a division of the Department of Transportation known as the

National Highway Safety Bureau (NHSB), which was renamed the National Highway

Traffic Safety Administration (NHTSA) in 1970, and the Office of Mobile Source Air

Pollution Control, a bureau within the Environmental Protection Agency. Today, this latter bureau is known as the Office of Transportation and Air Quality.

The Performance Standard as a Liberal Technology of Governance

The hazards of technique are a universal aspect of human experience. Fire, one of the simplest of human “tools,” burned human hands, blackened lungs, and polluted the air long before the rise of agriculture, let alone modern society. People doubtlessly injured themselves while making stone tools, just as they were certainly hurt while building ancient monumental structures, such as pyramids and temples.

The dangers of technology are as indifferent to political economy as they are to history. Socialist countries have fouled their natural environments and disabled their workers just as effectively and prolifically as capitalist ones. Air pollution has never bothered to learn the distinction between constitutional monarchies and republics. Images of Chinese citizens strolling through their cities while wearing anti-pollution facemasks take their place next to photographs of Cleveland’s

Cuyahoga River ablaze.

Yet, each society in human history has chosen to control or regulate technological risk differently. The opposite extremes of governance are banning a technique or technology outright and allowing them to go on unfettered (permission can equally be a kind of governance). But within this broad swath of limitation,

5 governments, or states, have developed a wide-variety of “tools” to rein in technology. From pouring money into researching safe technologies to creating draconian laws against people who harm others, authorities have sought ways to alleviate technology’s problems while retaining its benefits.

In the United States, the car is the ultimate liberal technology. Liberalism is a political philosophy founded on the centrality of the individual, on “rights,” or freedom from coercion, the ability to make one’s own decisions as long as those decisions don’t impinge on another’s ability to make his or her own decisions, and so on. Implicit in this idea is an image of life lived independently. In the United

States, this image fostered a specific desire for a stand alone, single-family house free of the density, noise, and dinginess of urban spaces. As transportation technologies increasingly improved over the course of the late 19th and 20th centuries, people began moving out of urban centers to “suburbs,” and the suburbs became the great liberal geography. The “streetcar suburbs” were an early development, preceded by strips built along rail lines, horse cars, and other transportation systems.6 Wealthy businessmen, the only ones who could afford such luxuries, began buying property in suburbs as the streetcar allowed them to commute between their downtown workplaces and their distant homes in a reasonable and formerly unthinkable amount of time. In the Post-World War II era, however, suburban development was almost completely based around the

6 Sam Bass Warner, Street Car Suburbs: The Process of Growth in Boston, 1870–1900 (Cambridge: Harvard University Press, 1962); Clay McShane, The Horse in the City: Living Machines in the Nineteenth Century (Baltimore: The John Hopkins University Press, 2007).

6 automobile. The car became a central pillar of the “American Dream,” embodying home and land ownership and independence.

But the car is liberal beyond its ability to foster an independent life: it is also a hallmark of individuality.7 In one of his radio addresses in the late-1970s, the future president Ronald Reagan said, “Personally I’ve always believed the automobile gave us one of the truly last great freedoms. For the first time the ordinary man could go where he wanted to go, when he wanted to, choose his own departure and arrival time free at last from timetables and fixed routes.”8 If the automobile played an important role in the financial economy, it was also a central character in the libidinal economy. Once an object of wealthy conspicuous consumption, the vehicle had become democratized. Teens learned to love in the backseat of their parents’ cars. Cruising, or driving slowly up and down a town’s main drag, became a weekend pastime. Backyard mechanics built or modified cars, creating hotrods and lowriders. The automobile became a central part of the how people imagined freedom in the United States. Building on earlier ideas of the frontier and the liberating potential of the railroad, in films and popular songs, the paired technologies, the car and the open road, allowed one to escape the confines of tradition, repressive community, and even the “dysfunctional” family. In his song

“Thunder Road,” songwriter Bruce Springsteen paints the picture of a young man and woman living in a broken, stultifying, nowhere town. He sings,

7 For the automobile’s deep connection to individualism, see James J. Flink, “Three Stages of American Automobile Consciousness,” American Quarterly, Vol. 24, No. 4 (Oct., 1972), 451-473 8 Kiron K. Skinner, Annelise Anderson, and Martin Anderson, eds., Reagan’s Path to Victory: The Shaping of Ronald Reagan’s Vision: Selected Writings (New York: Free Press, 2004), 237.

Well, now, I'm no hero That's understood All the redemption I can offer, girl Is beneath this dirty hood With a chance to make it good somehow Hey what else can we do now Except roll down the window And let the wind blow back your hair Well the night's busting open These two lanes will take us anywhere We got one last chance to make it real To trade in these wings on some wheels Climb in back Heaven's waiting down on the tracks Oh, oh, come take my hand Riding out tonight to case the promised land Oh, oh, Thunder Road, oh, Thunder Road

The male narrator of the song warns the woman he is wooing that if she does not join him she will be left behind to nothingness, “tonight we’ll be free . . . Mary climb in, this is a town full of losers, and I’m pulling out of here.”

How does a democratically-elected government attempt to control the harmful aspects of a technology that is so central to a “way of life,” to how its citizens imagine themselves? How does it regulate such a technology when the government has a tradition of invading the activities of individuals and business organizations as little as possible? In the United States, one form of technological regulation that gained prominence in the 20th century was known as “performance standards.” They were well fitted to the United States’ liberal society. Performance standards are typically contrasted with “design standards” (also known as

“technology standards”), which specify the kinds of technologies firms should use to solve a technical problem. Performance standards, on the other hand, simply set a given criteria and allow the regulated firms to meet the criteria in whatever way

8 they wish. If a performance standard is set at a level high enough that producers cannot yet reach it and have to “innovate” to do so, they are known as “technology- forcing” standards. Performance standards allow the government to curb targeted risks without requiring too heavy a hand—they are a liberal technology of governance. The standards interfere minimally with the sovereignty of the consumer or the driver; they place responsibility for reducing the technological risks firmly on the shoulders of producers. But performance standards also regulate producers in a way that allow them as much flexibility as possible. Firms can meet the criteria established in the standards however they wish. Thus, this dissertation examines on how the US federal government used performance standards to regulate technological risks associated with automotive crash safety, emissions control, and fuel efficiency.

Although we may look back at the history of auto regulation and see moments set in stone, nothing could have prepared the auto industry, especially the

“Big Three” domestic automobile manufacturers, for the federal government’s decisive entry. Perhaps, then, it is best to begin the story just before the automobile became the object of considerable federal regulation, that is, about 1960.

The Automobile in 1960

In 1960, the biggest story the self-styled “The Newspaper of the Industry,”

Automotive News, was the coming of the compact car. The Volkswagen Beetle was making waves during the mid-to-late 1950s, and other foreign automakers, including Opel, Volvo, and Fiat, were lining up to get a piece of the action. People

9 wanted—demanded—smaller cars news stories and editorials said again and again.

“Market Braces for Compact Storm,” an article warned.9 Automotive News was no fan of the new, smaller cars. “This is the compact age,” a staff editorial admitted, “and it is foolish to stay in business unless you have the opportunity to offer what the public wants.” But the paper warned, “Undeniably, the compact cars have caught the imagination of the public. But the compact car is not the answer to every automotive need.”10 Another staff editorial cautioned, “Dealers surveyed by Automotive News predict that compact cars will take an increasing share of the market. But don’t sell your big car yet.”11 A cartoon depicted a car so tiny that a male driver’s and wife’s derrieres stuck out the back of a burst open trunk. American Motors Corporation

(AMC), formed out of the merger of Nash and Hudson Motor Company, had pioneered the first domestic compact car. The AMC Rambler made its appearance in

1954, giving birth to a new “market segment.” Columnists sang praises of George

Romney, President of AMC, for saving the troubled company. The respite was brief, however. Within a few years, the company would be troubled again. Fears that

Americans would forfeit their large cars in favor of more economical ones also proved a bit premature.

In many ways, the modern domestic auto industry, as we know it today, had only emerged about 1960. In 1957, a recession had set in, and it killed many of smaller automakers, including Kaiser Motors and Packard, all of whom had struggled for years. Nash and Hudson might have joined that crowd had they not

9 Robert M. Lienart, “Market Braces for Compact Storm: Fierce Competitive Clash Feared . . .” Automotive News, July 11, 1960, 1. 10 “As Dealers Eye the Changes in U.S. Auto Market,” Automotive News, April 25, 1960. 11 “Capsule Comment,” Automotive News, March 7, 1960, 12. Italics in original.

10 merged a few years earlier. It was a late-date shake out, and the hungry beast of efficiency had feasted on these smaller firms. The winds of “creative destruction” left only the “Big Three”—General Motors, Ford, and Chrysler—and a few small firms, including American Motors, Lark, and Studebaker-Packard. Moreover,

Chrysler, the Big Three’s red-headed step child, faced bankruptcy. It would not be the company’s last experience with such matters. Most small automakers crashed on the rocks. Bigness won. Columnists made rosy predictions for the decade. There would be new stability after shake out. For all the talk of compact cars, the Big

Three automakers formed the backbone of a muscular industry that made muscular machines.

In 1960, President Dwight D. Eisenhower still held the White House, and the automakers did not fear government. Indeed, the auto companies put a great deal of faith in the federal government, particularly in road construction, which found its strongest embodiment in the Federal-Aid Highway Act of 1956 . Many heralded the increased construction of highways as a major factor in pulling the nation out of the recession of ’57–’58. All members of the industry took great comfort in this automotive Keynesianism. Automakers knew the importance of roads. The

National Highway Users Conference, which Automotive News called “an important arm of the automotive industry,” put continual pressure on the President and both houses of Congress.12 “Chairman emeritus” Alfred P. Sloan, the former and famous head of General Motors, and other elites within the automakers continued to steer the group, but it had been joined by an increasing number of ancillary businesses,

12 William Ullman, “Highway Administration Angers Road Users,” Automotive News, May 23, 1960, 10.

11 including “oil and tire interests, National Automobile Dealers Assn., truckers, and manufacturers who use trucks to haul their products.”13 Municipalities, real estate agents, and land developers were also important members. The National Highway

Users Group had been a major pressure group behind the passage of the Federal-Aid

Highway Act of 1956, and it continued to press the Department of Commerce, the agency responsible for implementing the Act.

William S. Richardson, the chairman of the group and a director of B.F.

Goodrich Co., said that he was “never very enthusiastic” about the Department of

Commerce being responsible for the Highway Act, claiming “Personally, I had hoped the United States Commerce would have established a National Highway Board.”14

This vision of an agency dedicated only to transportation issues was in the air. James

M. Landis, who President-elect John F. Kennedy had asked to write a report on regulatory agencies, suggested the creation of “ministry of transportation.”15 But the idea would have to wait six more years until the Johnson administration created the

Department of Transportation.

But the government did not require a centralized authority to make one thing true: the car-dependent suburbs marched outward.16 Earlier transportation technologies—such as the street-car and even the railroad itself—had assisted homebuyers who sought single-family houses, yards as large as possible, clean air, and other benefits difficult to find in the dirty and smoky city. The automobile

13 Ibid. 14 Ibid. 15 William Ullman, “Ray of Hope Penetrates Economic Forecasts,” Automotive News, November 28, 1960, 13. 16 Adam Rome, The Bulldozer in the Countryside: Suburban Sprawl and Rise of American Environmentalism (Cambridge: Cambridge University Press, 2001).

12 allowed consumers and developers to hasten—and democratize—this process. The veterans of World War II streamed into suburban subdivisions, most famously

Levittown, New York, a completely planned, fabricated town.

Home-building and suburban development were important engines of economic “growth.” And though the automobile had its part in this expansion of wealth and increased quality of living, the car was an engine of economic growth in its own right. This fact had been realized much earlier. By the late 1920s, the

National Automobile Chamber of Commerce, which was formed in 1913 and renamed the Automobile Manufacturers Association in 1934, began publicizing facts about the car’s vital role in the overall economy. As the historian Daniel M. Albert writes, “By 1928, automobile manufacturing used 18 percent of the nation’s steel, 19 percent of its wood, 23 percent of its aluminum, 74 percent of its glass, and 85 percent of its rubber.”17 Not only those employed directly by the automobile manufacturers but also a huge number of people throughout the nation (and around the world) owed their livelihood to the US automobile industry. But these figures were not merely boosterism for the auto industry; professional economists also realized the centrality of the industry in the US economy. In 1939, Charles F. Roos penned work a titled The Dynamics of Automobile Demand.18 Only two years earlier,

Roos had published another book, NRA Economic Planning, in which he ruminated on his experiences as the director of research of the National Recovery

17 Daniel Marc Albert, “Order Out of Chaos: Automobile Safety, Technology, and Society, 1925 to 1965,” (Ph.D. Diss., University of Michigan, 1997), 9. 18 Charles Frederick Roos, The Dynamics of Automobile Demand (New York: General Motors Corporation, 1939).

Administration (which the Supreme Court declared unconstitutional in 1935).19 In

The Dynamics of Automobile Demand, Roos, who originated the notion of “disposable personal income,” examined trends and explanations for consumer demand for automobiles. His analysis fit into the emerging Keynesian belief that demand would drive economic recovery. Although Roos did not emphasize the automobile’s importance to the nation’s economy, his book was a symptom of it. The auto industry was also an important center of employment. Since Henry Ford instituted the five-dollar day in 1914, the auto industry had been an important site for the shaping of labor policy, trade unionism, and the philosophy that laborers also played a critical role as consumers.20 Under the leadership of William Reuther, the United

Auto Workers rose to previously unimagined power after World War II.21 The notion that the nation’s financial health depended on the auto industry found its apogee in the phrase, “What’s good for General Motors is good for America,” famously misattributed to Charles E. Wilson, as he moved from being the head of GM to being Eisenhower’s Secretary of Defense in 1953. Eisenhower’s appointment of

Wilson was echoed less than a decade later, when John F. Kennedy appointed Ford president Robert McNamara to the same position. The auto industry, the federal government, and the national economy were all tightly interwoven.

Yet, the industry was also changing. New production techniques entered the industry, most famously forms of “automation,” which threatened and, therefore,

19 Charles Frederick Roos, NRA Economic Planning (Bloomington, Ind.: The Principia Press, 1937). 20 Stephen Meyer, The Five Dollar Day: Labor, Management, and Social Control in the Ford Motor Company, 1908–1921 (Albany: State University of New York Press, 1981). 21 Nelson Lichtenstein, The Most Dangerous Man in Detroit: Walter Reuther and Fate of American Labor (New York: Basic Books, 1995).

14 angered workers.22 In the post-War period, automakers quickly moved to monocoque, or unitized body, construction in which the vehicle’s external shell provided its structural support, rather than an internal frame. Throughout the period, the domestic automakers and others continued to experiment with alternative engines, such as electric engines, the Wankel (or rotary) engine, the stratified charge engine, the gas turbine, and the Rankine engine. Yet, most the automakers’ “innovative activity” went into the so-called “horsepower wars,” a Cold

War-style “arms race” to see which of the automakers could make the most powerful, high-performing, and quickest car on the market.23 Meanwhile, a host of industries vied for a piece of the American automotive industry. Aluminum manufacturers and steel makers squared off over whose product made the ideal material for bumpers. Plastics corporations spun out a plethora of new applications for the car. The auto industry was giant customer. Winning its business was winning big.

No one suffered troubled sleep over the thought of serious, strict federal regulation of the automobile, or at least they did not confess to it. In 1960, the US

Congress had carried out the first major hearings on automotive emissions, but these hearings received no detailed coverage from the press. The Automotive News

22 David F. Noble, Forces of Production: A Social History of Industrial Automation (Oxford: Oxford University Press, 1984); David A. Hounshell, “Planning and Executing ‘Automation’ at Ford Motor Company, 1945–65: The Cleveland Engine Plant and Its Consequences” in Fordism Transformed: The Development of Production Methods in the Automobile Industry, Haruhito Shiomi and Kazuo Wada, eds. (Oxford: Oxford University Press, 1995), 49–86. 23 On the “horsepower race,” see David A. Hounshell, “Planning and Executing ‘Automation’ at Ford Motor Company,” 75–76. In the economic literature, the now classic article by Franklin M. Fisher, Zvi Grilliches, and Carl Kaysen, “The Costs of Automobile Model Changes since 1949,” The Journal of Political Economy, Vol. 70, No. 5 (Oct., 1962), 433–451 gave rise to an extended discussion of the issue of the horsepower race and annual model changes.

15 staff noted weakly, “Witnesses clash at air-pollution hearings in Washington. Not all the smog is in Los Angeles.”24 This statement, perhaps a play on words suggesting that the Congressional hearing room was itself a site of smog production, was true enough: Cities around the country were coming to realize that they had serious air pollution problems. Similarly, California had passed the nation’s first automotive emissions regulations, but few outside the state saw these regulations as a big deal, and the industry took the news with a shrug.

In 1960, the automakers, their suppliers, and all those tied to the industry looked forward with expectantly. They had no inkling of what awaited them.

How Cars Became “Problems”

For decades now, sociologists, thinkers in Science, Technology, and Society studies, and other scholars have examined how issues become “problems” in society. 25 A core assumption of such analyses is that problems do not simply arise on their own; problems are not “natural.” Rather, a host of different social, economic, political, and media factors come together to focus the public’s mind on a

“problem,” or if the issue is dire enough, a “crisis.”

24 “Capsule Comment,” Automotive News, March 7, 1960, 12. 25 Joseph R. Gusfield, The Culture of Public Problems: Drinking-driving and the Symbolic Order (Chicago: University of Chicago Press, 1981). The historian of technology Thomas Hughes has emphasized the role of problem-formation in his examination of technological change. See Hughes, Networks of Power: Electrification in Western Society, 1880–1930 (Baltimore: Johns Hopkins University Press, 1983). Joseph W. Schneider and Peter Conrad give a beautiful phenomenological reconstruction of how people come to realize that they have epilepsy, a disease whose very nature often involves blackouts that prevent the sufferer from forming memories of the event. Their description has wider applicability to “problems” that are not immediately perceivable. Schneider and Conrad, Having Epilepsy: The Experience and Control of Illness (Philadelphia, Temple University Press, 1985).

Following this train of thought leads one to ask, how and when did cars become problems? The answer to this question is clear enough: cars have always been problems. From their inception, cars have been hazardous, and some people have always seen them as such.

Early cars were dangerous.26 Their chassis and axles broke while the cars were moving at full speed. Early vehicles were open-bodied; seatbelts were non-existent.

When cars crashed, drivers and passengers were often thrown from the vehicle. The risk of driving, however, emerged simultaneously with a masculine culture based on proving one’s manhood through feats of speed and daring. Early drivers, almost all men, were people who embraced being “risk-takers.”27 Automobile manufacturers won attention and good press through auto races and hill-climbing competitions. This marketing of risk

26 The earliest and now classic history of automotive safety is Joel W. Eastman, Safety vs. Styling: The American Automobile Industry and the Development of Automotive Safety, 1900–1966 (Lanham: University Press of America, 1984). Other classic works include James J. Flink, The Automobile Age (Cambridge, Mass.: The MIT Press, 1988) and William J. Abernathy, The Productivity Dilemma: Roadblock to Innovation in the Automobile Industry (Baltimore: The Johns Hopkins University Press, 1978). The work of John D. Graham, though coming primarily from a public health and policy studies perspective, provides a number of important historical insights. See, John D. Graham, “Automobile Safety: An Investigation of Occupant Protection Policies,” Ph.D. diss. Carnegie Mellon University, 1983; idem., “Saving Gasoline and Lives” in Risk versus Risk: Tradeoffs in Protecting Health and the Environment, John D. Graham and Jonathan Baert Wiener, eds. (Cambridge, Mass.: Harvard University Press, 1995); idem., Auto Safety: Assessing America’s Performance (Dover, Mass.: Auburn House Publishing Company, 1989); John D. Graham, ed., Preventing Automobile Injury: New Findings from Evaluation Research (Dover, Mass.: Auburn House Publishing Company, 1988). However problematic, the best legal history of auto safety is Jerry L. Mashaw and David L. Harfst, The Struggle for Auto Safety (Cambridge, Mass.: Harvard University Press, 1990). Today, automotive history is a burgeoning field with recent works including Jeremy Packer, Mobility without Mayhem: Safety, Cars, and Citizenship (Durham: Duke University Press, 2008); David Blanke, Hell on Wheels: The Promise and Peril of America’s Car Culture, 1900–1940 (Lawrence, KS: University Press of Kansas, 2007); and Peter D. Norton, “Street Rivals: Jay Walking and the Invention of the Motor Age Street,” Technology and Culture, Vol. 48, No. 2 (April 2007), 331–359. 27 Sally H. Clarke, Trust and Power: Consumers, the Modern Corporation, and the Making of the United States Automobile Market (Cambridge: Cambridge University Press, 2007), 40-49. In the first chapter— “The Arrogance of Wealth”—of his book Auto Mania, Tom McCarthy emphasizes that early drivers were wealthy as well as being risk-takers. Owning and using a car were some of the grandest forms of conspicuous consumption during the early days of the automobile. Of course, luxury cars continue that tradition to this day. Tom McCarthy, Auto Mania: Cars, Consumers, and the Environment (New Haven: Yale University Press, 2007). In her work, Taking the Wheel: Women and the Coming of the Motor Age (New York: Free Press, 1991), Virginia Scharff describes how Suffragists and other early female drivers challenged the era’s dominant notion of womanhood, a passive, frail, and easily traumatized femininity.

17 could expand the market only so far, however. If the automakers wanted to sell large quantities of cars, consumers would need to find everyday, practical uses for their product, and they would need to feel relatively safe while doing so.

Many different actors—including “traffic engineers, traffic police, forensic psychiatrists, and driving educators”—shaped the automobile’s new safety regime. 28

Traffic signals and laws regulated the vehicle’s use. Mandatory driver education and driver licensing created a base of competency around the vehicle. Yet, most of these rules and requirements were pointed at the driver and placed responsibility firmly on the driver’s shoulders. 29 By 1960, automakers had placed few safety features in their vehicles. Earlier, when Robert McNamara headed a division of the larger Ford Motor

Company, he had attempted to include safety features in vehicles and to make safety a selling point, but his efforts failed, reinforcing the old Detroit mantra that “safety doesn’t sell.”30 At the same time, the transportation system of roads and highways was growing increasingly complex. In 1966, Lyndon Baines Johnson explained this complexity in his signing statement for the creation of the Department of Transportation. In 1946, 31 million motor vehicles crisscrossed the nation’s roadways; by 1966, the number had increased to 90 million, and government agencies estimated the number would grow to

120 million by 1975. With this growth in vehicle use came a great increase in collisions,

28 Daniel Marc Albert, “Order Out of Chaos: Automobile Safety, Technology, and Society, 1925 to 1965,” (Ph.D. Diss., University of Michigan, 1997), especially chapters 2 and 5. 29 Jameson Wetmore, “Systems of Restraint: Redistributing Responsibilities for Automobile Safety in the United States Since the 1960s,” (Ph.D. Diss., Cornell University, 2003). 30 John A. Byrne, The Whiz Kids: The Founding Fathers of American Business—and the Legacy They Left Us (New York: Doubleday, 1993), ch. 17. The quotation is typically attributed to General Motor’s long- time President Alfred Sloan.

18 and every year tens of thousands of people died on the road. Johnson called this situation an “epidemic.” 31

The car safety epidemic had become more acute in the public’s consciousness a year earlier with the publication of Ralph Nader’s Unsafe at Any Speed.32 Automakers had already been under considerable pressure from politicians and public interest groups.

In 1964, the auto companies voluntarily adopted seatbelts as standard equipment to head off possible legislation.33 This move did not appease safety advocates. As the media began to investigate accusations of the auto industry’s using private investigators to spy on Nader, momentum gathered for federal legislation and standardization. This movement eventually came to fruition in the Highway Safety Act of 1966, which created the National Highway Safety Bureau (NHSB), which in turn became the National

Highway Traffic Safety Administration (NHTSA) in 1970. The NHSB promulgated twenty standards in its first two years. Though the NHSB and the NHTSA have experienced moments of strength and weakness, the Highway Safety Act created a lasting institution to develop and support crash safety regulations.

The battle against automotive emissions has roots in the much longer and broader history of urban smoke control efforts.34 Ironically, the car was seen as a relatively clean technology when it was first introduced. In 1901, Ransom Olds advertised his gas-buggy

31 “The White House Message on Transportation,” page 1, NARA, Record Group 416, Finding Aid UD- UP-3, Box 1, Folder 2. 32 Ralph Nader, Unsafe at Any Speed: The Designed-In Dangers of the American Automobile (New York: Grossman, 1965). 33 Stan Luger, Corporate Power, American Democracy, and the Automobile Industry (Cambridge: Cambridge University Press, 2000), 66. 34 David Stradling, Smokestacks and Progressives: Environmentalists, Engineers, and Air Quality in America, 1881-1951 (Baltimore: Johns Hopkins University Press, 1999).

19 as “Odorless, Noiseless, Safe.”35 It did not spew the smoke that formed the basis of suits and movements against railroads. Nor did cars produce the odorous effluent particular to horses.36 In matters of air pollution, the state of California led the nation on almost every front.37 The Los Angeles basin had been seen as a paradise for motorists, but its natural tendency toward inversions and unchecked development transformed Southern California into the “Smog Capital” of North America.38 The science of “smog” arose from the work of A. J. Haagen-Smit and other California-based scientists, and the California Air

Resources Board (CARB) developed the earliest auto emissions standards. The federal government entered the scene primarily through the Clean Air Act of 1965. This law, however, set a very low bar for emissions controls, and the US manufacturers were able to meet the standards by making minor adjustments to their engines, leaning the fuel mix, thereby, emitting fewer of the regulated pollutants. The federal government’s powers were considerably strengthened through the Clean Air Amendments of 1970, also called the Muskie Act in recognition of Maine Senator Edmund S. Muskie’s intense support of the bill. A small office within the Department of Health, Education, and Welfare had administered the Clean Air Act of 1965, but the 1970 Act made automotive emissions control part of the newly organized Environmental Protection Agency, which

35 Quoted in Clarke, Trust and Power, 34. Emphasis added. 36 Joel A. Tarr, “Urban Pollution: Many Long Years Ago,” American Heritage Magazine 22 (October 1971), 65–69. See also, Clay McShane and Joel A. Tarr, The Horse in the City: Living Machines in the Nineteenth Century (Baltimore: The Johns Hopkins University Press, 2007) and Ann Norton Greene, Horses at Work: Harnessing Power in Industrial America (Cambridge, Mass.: Harvard University Press, 2008). 37 James E. Krier and Edmund Ursin, Pollution and Policy: A Case Essay on California and Federal Experience with Motor Vehicle Air Pollution, 1940-1975 (Berkeley: University of California Press, 1977). 38 Scott Hamilton Dewey, Don’t Breathe the Air: Air Pollution and U.S. Environmental Politics, 1945- 1970 (College Station, TX: Texas A&M University Press, 2000), 38.

20 reinvigorated already existing programs and considerably expanded new ones in the

Office of Mobile Source Pollution Control.39

Automotive fuel efficiency has the thinnest history and historiography of these three domains. Fuel shortages around World War I drove research into ways of using poorer quality petroleum. Charles “Boss” Kettering’s and Thomas Midgley’s discovery of tetraethyl lead was partly a response to this situation.40 And fuel economy was always a concern and topic of conversation for drivers. From 1936 to 1968, the Mobil Oil

Corporation sponsored the Mobil Economy Run, a coast-to-coast road test that carefully monitored the fuel economy of stock cars. Until the 1970s, however, Americans enjoyed the fruits of cheap fuel, and their lifestyle became dependent upon it.41 For most of that time, the United States’ rather large domestic petroleum sources provided most of the nation’s energy demand. Over time, however, the United States became more dependent on foreign sources. This dependence became increasingly true as the so-called

“horsepower race” took off throughout the 1950s and 1960s, wherein, domestic automakers greatly increased engine size and “performance,” also causing fuel consumption to expand markedly.42 By 1973, the United States produced roughly eleven million barrels of oil a day, but it was importing over six million barrels of crude, about

39 See Krier and Ursin, Pollution and Policy, especially chapters 10 and 11; Luger, Corporate Power, 87– 88. 40 For a treatment of Kettering’s creation of tetraethyl lead, see Stuart W. Leslie, Boss Kettering (New York: Columbia University Press, 1983). On the history and hazards ot tetraethyl lead, see Christopher C. Sellers, Hazards of the Job: From Industrial Disease to Environmental Health Science (Chapel Hill: The University of North Carolina Press, 1997) and Christian Warren, A Brush with Death: A Social History of Lead Poisoning (Baltimore: The Johns Hopkins University Press, 2000). 41 David Nye has argued that the “energy crisis” of the 1970s was, in fact, a crisis of middle class culture since shortages threatened the “American way of life.” David Nye, “ The Energy Crisis of the 1970s as a Cultural Crisis” in Living with America, 1946-1996, Ed. Rob Kroes and Cristina Giorcelli (Amsterdam: VU Press, 1997). 42 See Hounshell, “Planning and Executing ‘Automation’ at Ford Motor Company, 1945–65,” 75–76.

21 thirty-five percent of the nation’s demand.43 When the OPEC oil embargo began on

October 17, 1973, policy makers in the United States began to scramble for ways to conserve energy and produce new supplies. They used many means to attain this end. In

1975, Congress enacted the Energy Policy Conservation Act, which contained a section titled, “Improving Automotive Efficiency.” The standards formed from this act eventually became known as the Corporate Average Fuel Efficiency (CAFE) standards.44

The management of fuel economy standards has always been a complex affair. NHTSA technically creates and enforces the standards, but the EPA measures the cars’ fuel efficiency as part of the agency’s emissions control efforts, and the Department of Energy publishes the automotive fuel efficiency statistics as part of its energy information efforts.45

Expertise and Bureaucracy During the Postwar Period of Institution-Building

Both the New Deal and World War II had shown people the power of the federal government to induce change. Socially and technologically, war efforts had an enormous influence on the status quo. Yet, by the end of the war, the zeal for institution-building and for state intervention in society had waned.46 President

43 Statistics from the Energy Information Administration: http://www.eia.doe.gov/emeu/aer/txt/ptb0501.html 44 The CAFE program has always faced a number of criticisms, especially the criticism that it decreases the safety of automobiles. For a representative conservative criticism of CAFE from one of the libertarian Cato Institute’s publications, see Andrew N. Kleit, “CAFE Changes, By the Numbers,” Regulation, Vol. 25, No. 3 (Fall 2002). Additionally, many economists and policy-analysts agree that the best way to spur a decrease in driving and an increase in fuel-efficient vehicles would be to levy a substantial tax on gasoline. But many also believe that such a tax is politically impossible. 45 On the federal government’s creation of energy statistics and forecasting, Lee Vinsel, “’Every Week We Find a New Devil’: The Crusade for Credible Energy Information and Analysis in an Era of Great Mistrust, 1973-77,” History and Technology, revise and resubmit. 46 Alan Brinkley, The End of Reform: New Deal Liberalism in Recession and War (New York: Vintage Books, 1996).

Dwight D. Eisenhower had little love for the civil service, believing that its ranks had been infiltrated and colonized by New Dealers and other pro-spending forces.47

Eisenhower began policies that allowed him to push political appointees deeper into federal agencies, ensuring the loyalty of people who held key positions. Eisenhower had little faith in the supposedly “detached objectivity” of experts in the civil service.

This mindset nearly inverted with the election of John F. Kennedy, and he began a movement towards using government to protect consumers and workers.

The historian Lizbeth Cohen has listed over forty-two laws passed from 1960 to

1977 that were aimed at protecting consumers.48 Kennedy loved and surrounded himself with experts—the “best and the brightest,” as David Halberstram memorably called them.49 The White House was hip-deep in Ivy League mandarins.

Kennedy’s appointment of Robert McNamara as Secretary of Defense was one notable example. McNamara held an MBA from Harvard Business School and, as a faculty member there, specialized in accounting methods. He applied his considerable quantitative skills as a Captain in the Air Force’s Office of Statistical

Control during World War II. The Office of Statistical Control would become an important seedbed for ways of applying mathematical methodologies to an increasing number of problems. George Dantzig, the father of Operations Research in the United States, also worked at the office. After the war, McNamara went on to apply these methods at Ford Motor Company, becoming one of Henry Ford II’s so-

47 Francis E. Rourke, “Responsiveness and Neutral Competence in American Bureaucracy,” Public Administration Review, Vol. 52, No. 6 (Nov. – Dec. 1992), 539–546. 48 Lizbeth Cohen, A Consumers Republic: The Politics of Mass Consumption in Postwar America (New York: Vintage Books, 2003), 360. 49 David Halberstram, The Best and the Brightest (New York: Random House, 1972).

23 called “Whiz Kids” and eventually rising to the presidency of the company, the position he had held for only a month when Kennedy asked him to become

Secretary of Defense.50

This trust in experts only increased when Lyndon Baines Johnson took over the presidency after Kennedy’s assassination.51 Johnson idealized Franklin Delano

Roosevelt’s New Deal, and he believed in the power of government to improve society. Although, in the days and months after Kennedy’s assassination, Johnson attempted maintain a balance that honored Kennedy’s earlier work, he slowly began to move towards his more activist conception of the government’s role in society. In

January 1964, in his first State of the Union Address, Johnson declared a “War on

Poverty.” On May 22, 1964, only six months after Kennedy’s death, Johnson delivered a speech at the University of Michigan in which he put forward his vision of the Great Society. Johnson’s Great Society was comprehensive, holistic, and synthetic. For just reasons, historians have tended to focus on the Civil Rights efforts during this period. Certainly, the Civil Rights movement of the time was the most important social shift of the era, and Johnson’s signing of the Civil Rights Act of

1964 and the Voting Rights Act of 1965 is his greatest legacy. But such single- minded treatments fail to explain the how Johnson’s other programs fit within the larger whole. Johnson did mention race during his Great Society speech, but he also

50 John A. Byrne, The Whiz Kids: The Founding Fathers of American Business—and the Legacy They Left Us (New York: Doubleday Business, 1993). 51 Robert Dallek, Lyndon B. Johnson: Portrait of a President (Oxford: Oxford University Press, 2005); Bruce J. Schulman, Lyndon B. Johnson and American Liberalism: A Brief Biography with Documents (Boston: Bedford Books of St. Martin’s Press, 1995).

24 mentioned education, suburban sprawl, lack of available housing, inadequate transportation systems, environmental degradation, and poverty.

Johnson, a master legislator in the US Senate before becoming Vice President, put forward an ambitious platform of laws, many of which included new executive branch agencies. With Johnson’s blessing, the 89th Congress, which ran from 1965 to 1967 and which many people believe was the most “liberal,” or progressive, congressional session in the nation’s history, began passing a fleet of (for the United

States) left-leaning laws. These laws included Medicaid, Medicare, the Housing and

Urban Development Act [which created the Department of Housing and Urban

Development (HUD)], important education laws, the Freedom of Information Act, the act that established the Department of Transportation, and two laws that deeply influenced the automobile, the National Traffic and Motor Vehicle Safety Act and the

Motor Vehicle Pollution Control Act. Outside of the New Deal, this was the moment of institution-building in the United States.

Another marker of this transition to a more active role of the state was the changing place of think tanks in society. As historians, such as David Jardini and

Jennifer S. Light, examined, throughout this period, think tanks, especially the RAND

Corporation, moved, as Light put it, “from Warfare to Welfare.”52 Following an arc similar to the one that took McNamara from the Air Force to studying production problems at Ford, researchers at RAND saw the broader applications of the

52 David R. Jardini, “Out of the Blue Yonder: The RAND Corporation’s Diversification into Social Welfare Research, 1946–1968,” PhD diss., Carnegie Mellon University, 1996; idem., “Out of the Blue Yonder: The Transfer of Systems Thinking from the Pentagon to the Great Society, 1961–1965” in Agatha C. Hughes and Thomas Parke Hughes, Systems, Experts, and Computers: The Systems Approach in Management and Engineering, Word War II and After (Cambridge, Mass.; MIT Press, 2000); Jennifer S. Light, From Warfare to Welfare: Defense Intellectuals and Urban Problems in Cold War America (Baltimore: Johns Hopkins University Press, 2005).

25 techniques that they were developing. As the Kennedy and Johnson administrations allocated more resources towards social problems, think tanks saw an opportunity and stepped in to reap the reward. These people put a great deal of faith in the power of mathematical techniques, models, and systems analyses, and they convinced many people in government to take great faith in them. The staff members of RAND and other analytical think tanks moved in and out of the federal government, bringing their quantitative conceptions of organizing the world with them. In this way, a changing and emboldened notion of expertise accompanied the institution-building of the 1960s.

The Nixon and Ford administrations that followed Johnson did not share his ardor for expanding the role of government, but they were also institution-builders.

Through the congressionally-mandated Economic Stabilization Act of 1970, Nixon famously froze prices and wages, which made free market conservatives (whom I discuss in the epilogue) rabidly angry. Nixon also oversaw the creation of important executive-branch agencies, including the Environmental Protection Agency and the

Occupational Safety and Health Administration. In response to the “energy crisis” brought on by the OPEC Oil Embargo, Nixon also created the Federal Energy Agency, whose organization was mostly overseen by Ford after Nixon resigned. Although

Ford’s administration created fewer agencies than had Nixon’s, he oversaw important expansions of federal power, including the creation of federal fuel efficiency standards.

In 1977, Jimmy Carter entered the White House on a campaign that distanced itself from the corruption of Watergate, the horrors of Vietnam, and the Nixon

26 pardon. Carter intended to bring a different sensibility to the White House. Although both Nixon and Ford had both been institution-builders in their own ways, neither relished that role. In the early days of his presidency, Carter symbolized return to the pro-government, pro-planning ethos of the Johnson administration. He created new agencies, including the Department of Education and the Department of

Energy.53

All of the events covered in this dissertation took place during this period of institution-building, when politicians, both Democratic and Republican, believed that government could regulate the more heinous aspects of technological risk, including those inherent in the automobile, and also believed that it was in the public’s interest to do so. In the Epilogue, I briefly consider what happened to auto regulation when the era of institution-building came to an end, when a new, anti- government philosophy called neoliberalism came to reign. The developments described in this dissertation simply would not have been possible under Ronald

Reagan’s neoliberal leadership, and he sought to undo the developments that had come before him. Ultimately, his destructive efforts were limited, however. The institutions I describe in the following pages remain, though perhaps their value can be questioned.

The Plan of this Dissertation

This dissertation proceeds in a roughly chronological order. In the first chapter, I examine the formation of the National Highway Safety Bureau (NHSB), the

53 The Department of Education was split off of the Department of Health, Education, and Welfare, the remaining portions of which became the Department of Health and Human Services.

27 agency charged with managing the federal auto safety standards. Several different professional communities—including law, medicine, and engineering—were vying to lead agency, giving it their own distinctive vision of what traffic safety entailed and how it could be accomplished. Ralph Nader, who was in many ways the visionary behind the Traffic Safety Act, had a distinctly legal idea of how the agency should function—an idea that would have involved lawyers running it. But

President Lyndon Johnson appointed a public health expert to the lead as the agency’s director. The Johnson administration cast highway accidents as a “disease”; doctors became “objective” experts who would cure it. Yet, engineers eventually came to dominate the agency. I examine the organizational reasons why this happened and describe how the visions of engineers differed from those of the medical community.

In Chapter Two, I describe how the NHSB built the first national safety standards. I focus on the most controversial of this first wave of standards, entitled

Standard 201, which was meant to lessen damage to drivers and passengers when they collided with the vehicle’s interior during a crash. But the automakers fought back against the agency, claiming that the standard was unreasonable, unfair, and unworkable. The agency and the automakers then entered a heated, adversarial process, through which the matter was finally settled and the standard was constructed. In this way, I examine how the agency’s inability to make a standard that was as stringent as it wanted created a precedent of limited power that hemmed in the agency’s most ardent advocates of tough safety regulations.

In the third chapter, I describe the process through which the NHSB forced the automaker, Fiat, to recall thousands of vehicles and eventually pay a civil penalty for safety violations. I emphasize the role that mundane labor plays in enforcing and sustaining regulatory standards. After a contractor for the agency found that the steering wheels in some of Fiat’s cars were defective, the agency began a process— which eventually took more than a year—to make Fiat conform to the standard. In this chapter, I focus on the central role of testing—in this case, crash testing—in regulatory enforcement.

Chapter Four treats the development of the Office of Mobile Source Air

Pollution Control within the Environmental Protection Agency. In the 1960s, the

National Air Pollution Control Administration (NAPCA), a department within the

Department of Health, Education, and Welfare, managed the nation’s auto emissions control program. But NAPCA was famously weak, and its mission was diffuse, even confused. Centrally, because it had to base its standards on “available” technology, the agency had an significant research and development component. Because very stringent emission standards were built into Clean Air Act amendments of 1970, the

EPA’s Office of Mobile Sources moved in a different direction, focusing on technology assessment instead of R&D. In this chapter, I demonstrate how Eric

Stork, the head of the Office of Mobile Sources, reshaped federal auto pollution control efforts to fit his vision of the new agency’s mission.

The hearings that the EPA held in 1972 and 1973 to determine whether it should suspend for one year the standards set in the Clean Air Act amendments of

1970 provide the focus of Chapter Five. I argue that, contrary to the neoclassical or

29 neoliberal view that regulation hinders “innovation,” these hearings helped produce, aggregate, and disseminate important new knowledge of many different kinds. I focus particularly on how the EPA learned about the industrial organization of the auto industry, how both the regulators and the automakers learned about the

“meaning” of the law through judicial review, and how the hearings became a space for learning about the state of the art in automotive emissions control technologies.

In the final chapter, I describe the prehistory of federal fuel economy standards at the EPA. First, I explain why the agency began measuring automotive fuel economy in the early 1970s: it sought to protect itself from critics, both automakers and consumers, who claimed that emission controls were greatly reducing fuel efficiency. Once the OPEC Oil Embargo set in, however, and gas prices spiked, other agencies became very interested in the issue of automotive fuel economy. Moreover, some representatives of the auto industry wanted the economy measurement program to be moved to a more “pro-industry” executive branch agency. Thus, the EPA, particularly its Office of Mobile Sources, struggled with both other executive branch agencies and the automakers and to maintain a hold on the fuel economy program. This dissertation concludes with an epilogue.

Chapter One—Making the Crash Barrier: Medical Authority, Engineering Culture, and Bureaucratic Practice in American Automotive Safety

In 1966, the Traffic Safety Act (TSA) and its sister statute, the Highway Safety

Act (HAS), became law. The TSA created the nation’s first significant and binding mandate that all automobiles in the United States had to conform to safety criteria.

The combined acts also created a new agency, the National Highway Safety Bureau

(NHSB), later renamed the National Highway Traffic Safety Administration

(NHTSA), charged with creating automotive safety standards and enforcing them.

This dissertation focuses on how regulatory agencies have developed practices to shape the auto industry and its products, or, in other words, how regulatory practice both induces and influences technological “innovation.” In my view, these practices, along with the regulators’ beliefs, ideas, and routines, form an “organizational culture” that has important ramifications for regulation’s effectiveness.

In this chapter, I seek to answer partly one question: why did the NHSB become a society of engineers? Or, more precisely, why did engineering culture dominate the agency’s actions and self-understanding? Like nearly all federal agencies, in the beginning, NHTSA had many possibilities. It was anything but rigid.

An eager, reformist spirit possessed its employees, and the agency’s potential was open and flexible. So how did this openness close? I focus on how leadership shaped this outcome. I argue that, although William Haddon, an eminent MD and expert in public health, led the NHSB as its first administrator, his deputy, Robert Brenner, an engineer and scholar in the expanding field of traffic studies at UCLA, played a greater role in managing the agency and shaping its ultimate character. I suggest

30 31 that these men had very different notions of what kind of expertise the agency’s employees should typify and that, in the end, Brenner’s conception won out. Thus, the chapter focuses on the intellectual biographies of these men, on what comprised their thinking, and how that thought shaped the day-to-day events at NHSB, including management and decision making.1 Too often people have argued that, because Haddon was the first administrator of the agency, the NHSB closely conformed to his ideas and values. The NHSB’s archival record, which covers the broad sweep of the agency’s activities, simply does not support this interpretation

A plethora of factors, both dependent and independent, determined the agency’s character. No sufficient proximal cause lies behind the others. If such a thing did exist, we cannot discern it. Leadership queues up with other influences, but it holds the central place. This brings us to the thorny intersection between intellectual biography and history, between theory and praxis. On the one hand, when we consider the history of experts, of trust in numbers and knowers, and of scientists and their role in governance, we should attend to how these people shape their and others’ everyday behaviors and how these behaviors in turn shape the world. On the other hand, although as historians we can and should focus on concrete practices, we should not neglect the role that decision-makers play in fostering and developing those practices. Practices and leadership can be overemphasized to the exclusion of the other. I focus on this dual dynamic here.

1 Parenthetically, I believe that this history reconfirms the picture that Andrew Delano Abbott paints in his The System of Professions: An Essay on the Division of Expert Labor (Chicago: University of Chicago Press, 1988): professions do not expand in some endless void. They vie for turf, and government regulation creates new spaces for the back-and-forth of disciplinary opportunism.

This chapter begins in a, perhaps, counter-intuitive place: the intellectual biography of the world’s most famous consumer advocate Ralph Nader. If, as I argue, Haddon acted less as the manager of the NHSB than as its figurehead, we can only wonder why he was chosen. I argue that the choice of Haddon only makes sense in the context of national debates around auto safety. Consumer advocates, like Nader, acted as the prime movers of these debates: they brought attention to the issue and framed early discussions on it. Yet, they portrayed auto safety as a redistribution of power. This politicization of the issue chafed the Johnson administration, which sought to diminish technological risk while remaining friendly to industry. In this way, William Haddon’s apolitical, medical vision of auto safety appeared as a less dangerous alternative. After spelling out Haddon’s vision of accident prevention, I turn to Robert Brenner’s role in founding and managing the agency.

At the opening of the first session of the 89th Congress, Abraham Ribicoff, chairman of the Subcommittee on Executive Reorganization, was feeling around for issues to put before the committee.2 He considered a number of topics but returned consistently to an issue that he had worked on since his time as the Governor of

Connecticut (1955-1961), namely automotive and highway safety. Not long after

Jerome Sonosky, the committee’s staff director and general counsel, began to

2 Perhaps, we could characterize this not as “solutions in search of a problem” as in the famous “Garbage Can Model” but rather as “problem-solvers in search of a problem to which they would create the, no doubt ‘definitive,’ solution (thereby, winning applause).” On the Garbage Can Model, see Michael D. Cohen, James G. March, and Johan P. Olsen, “A Garbage Can Model of Organizational Choice,” Administrative Science Quarterly, Vol. 17, No. 1 (March 1972), 1–25.

33 research auto safety for Ribicoff, an attorney stopped in to make a few suggestions to him. This man was Ralph Nader. At the time, he was writing the book that would become Unsafe at Any Speed, the now-classic work of muckraking that brought the designed-in dangers of automobiles to the public’s attention. Sonosky found Nader to be extremely knowledgeable about auto safety, and since Nader was willing to work anonymously on the issue, allowing the Senators to take credit for the labor,

Sonosky invited him to help draft an auto safety bill, the bill that would eventually the Traffic Safety Act of 1966.3 In this way, Nader played a dual part in the rise of federal auto safety regulation: he acted both as the (co-)author of the law and as the issue’s most important publicist. While Unsafe at Any Speed experienced only moderate sales during its initial release in late 1965, it became an overnight blockbuster after the Washington Post unveiled that General Motors (GM) had hired private investigators to snoop on Nader, looking into his sex life and his stance on hot-button issues, like anti-semitism. GM President James M. Roche confessed this rapacious indiscretion before Ribicoff’s committee, guaranteeing nearly unanimous passage of the act.

Though Nader has been the subject of several popular biographies and at least one documentary film, we know rather little about his intellectual biography.

Some of this is deliberate on his part. He has avoided making many large-scale philosophical statements, claiming that “there is too much Athens in present day

3 Charles McCarry, Citizen Nader (New York: Saturday Review Press, 1972), 17-8.

Washington.”4 But, as I will show, Nader’s comments on philosophizing belie his engagement with the thought of his day.

Nader’s understanding of auto safety was centrally political, and his politics arose from an allergic reaction to the theory of pluralism that ruled the liberal philosophy of the day. At its core, pluralism applied the economic theories of capitalism to politics, viewing “interests” as competing in the political realm as

“firms” compete in the “market place.” Nader echoed John Kenneth Galbraith—a chief theorist of pluralism—when he asked, “What avenues for reform, for the reduction of these [social] costs, are available to the citizenry when the diversity of competing, countervailing or penalizing centers of economic power is over- ridden?”5 For Galbraith, countervailing power arose when large-scale buyers and sellers, including whole-salers and labor unions, were able to fight back against corporate monopoly power, creating an antagonistic process. But Nader noted that

Galbraith’s idea now “reads like a quaint economic fable.”6 Corporate power simply overwhelmed these “pluralistic interests,” ensuring that “the effective lifecycle of countervailing forces . . . is notoriously short.” If “unions, government, different industries, big buyers, etc.” could occasionally organize to oppose corporations, large businesses and their loyal legislators would quickly shut the opposition down.

“Devolution sets in, and the system is praised as being ‘pluralistic,’ as if that description suffices to set out minds at rest.”

4 Quoted in Robert D. Holsworth, Public Interest Liberalism and the Crisis of Affluence: Reflections on Nader, Environmentalism, and the Politics of Sustainable Society (Boston: G. K. Hall & Co., 1980), 12. 5 Nader, “Taming the Corporate Tiger” in The Ralph Nader Reader (New York: Seven Stories Press, 2000), 136. The speech was originally delivered before the National Press Club Washington, DC, December 21, 1966 6 Ibid., 137.

Thus, Nader’s politics were anti-corporate, but his view was complicated by the belief that technological risk was prone to two types of “lag.” The first fit

William Ogburn’s notion of “cultural lag”: the methods of cultural and political steering always lagged behind developments in the technological risks they were meant to control.7 Indeed, Nader quoted Francis Bacon’s famous statement, “He that will not apply new remedies must expect new evils, for time is the greatest innovator.” Nader believed that this kind of lag lived primarily in the legislative branch, which had not passed binding safety legislation around many technological risks.8 Courts had done better on this score, he believed, especially through rulings on product liability.9

The second form of lag revolved around Nader’s understanding of corporate power. Like some economists of the day, he subscribed to the view that large corporations put less effort into innovation than smaller ones.10 Bigness was a

7 Ogburn originally formulated the idea of “cultural lag” in Social Change with Respect to Culture and Original Nature (New York: Viking Press, 1950 [1922]). 8 Nader, Unsafe at Any Speed (New York: Pocket Books, 1966), vii-viii. 9 While Nader’s assessment of the courts here may have been historically accurate, it also arose from his beliefs in the importance of lawyers in social change. In his various organizations, Nader tended to surround himself with lawyers, and he had a supremely legal understanding of what auto regulation was about. This tendency in Nader also illuminates a sub-theme of this essay: when in doubt, people often believe that their own academic or professional discipline has the answer to society’s problems. Nader pointed to lawyers in his recommendations for auto safety; Haddon, to doctors and social and natural scientists; Brenner, to engineers. Each in his own way saw the issue through his disciplinary lens. 10 This issue has a long history and partly arises from Joseph Schumpeter’s reflections on the nature of large firms and innovation. Some have maintained that large firms have diseconomies of scale regarding R&D; some assert the opposite. Those in the first camp often asserted that large corporations put more exertion into “non-productive,” defensive strategies, including taking out protective patents. The difference between the truth of the two positions may come from whether large firms put R&D effort into creating so- called “disruptive technologies,” or new products that greatly change the potentials of the system, or into improving processes, which would have a less radical result. With this in mind, we could also wonder if auto safety required “radical innovation.” Nader and others would have certainly said, no, since they believed that many technologies for improving safety were already available. Regardless, Nader believed that the bigness of the Big Three and other automakers discouraged them from creating safer cars. The most conclusive statement that large firms have at least some advantages in R&D comes from Wesley M. Cohen and Steven Klepper, “A Reprise of Size and R & D,” The Economic Journal, 106 (July, 1996), 925-

36 problem for Nader. As he argued, in large corporations, “diminishing returns set in rather promptly, particularly in the area of innovation, under the grip of bureaucratic rigidities.”11 Therefore, the government had to step in to generate positive technological change.

For Nader, the government particularly had to act to protect the inviolability of the human body. Sometimes the government had acted, again primarily through the judiciary. As he wrote, “Judicial decisions throughout the fifty states have given living meaning to Walt Whitman’s dictum, ‘If anything is sacred, the human body is sacred.’”12 If Whitman wrote “I Sing the Body Electric”—the origin of that quotation—to celebrate the body in a conservative, Christian culture that diminished it, Nader sought to make the human form inviolable in a corporate culture that casually destroyed people, through unsafe products, through pollution, and through dangerous labor practices. In this way, Nader was and is an arch- liberal. His arguments about justice come from the ultimate value of the individual.13 He simply did not believe that corporations are sovereign bodies, as, for instance, libertarians do.14

Indeed, Nader’s vision of the NHSB connected and combined these three factors—corporate power, the existence of lags, and that corporations should not

951. Schumpeter on corporations and innovation: Joseph A. Schumpeter, Capitalism, Socialism, and Democracy (New York: Harper & Brothers, 1942). 11 Nader, “Taming the Corporate Tiger,” 135-6. 12 Nader, Unsafe at Any Speed, vii. 13 Holsworth argues that Nader’s focus on the individual makes his environmental arguments incoherent. Nader cannot argue for environmental conservation beyond pollution’s effect on human health, and the role that consumerism plays in environmental degradation is a blind spot for Nader. This conflict is best captured in the title and substance of one of Holsworth’s subsections: “Self-Interest and Scarcity: Nader’s Contradiction.” Holsworth, 53-56. 14 For instance, in the documentary, The Corporation, Nader recounts, with obvious disgust, how, in the late 19th century, corporations came to have the rights of persons. This, he believes, was a great mistake.

37 have personal rights. His vision of the auto safety agency was really very different than what came to be. It was far more radical. While Nader believed that safety standards were important, he also focused on the role the new agency could play in creating truth and knowledge that could be used in future lawsuits, especially product liability cases, against the automakers. In fact, only a few parts of Nader’s several congressional testimonies dealt with the issue of auto safety standards— primarily his testimony before Warren G. Magnuson’s (D—Washington) Senate

Commerce Committee on April 4, 1966, though even there Nader attended to the place of governmental knowledge in the court room.15

Two days later, Nader appeared before the Subcommittee on Public Roads. His comments focused on the issue of “secrecy,” by which he meant both corporate and governmental secrets. He especially concentrated on section 107 of the Highway

Safety Act, which states:

“No part of any report or reports of any Federal agency, or officers, employee or agent thereof, relating to any highway traffic accident or the investigation thereof conducted pursuant to this act or other applicable law, shall be admitted as evidence or used in any action for damages or criminal action, nor shall any such officer, employee, or agent be required to testify in such proceedings as to facts developed in such investigations. Any such report or reports of any such officer, employee, or agent shall be made available with the approval of the

15 Auto safety was divided into two acts—the Traffic Safety Act and the Highway Safety Act—both passed in 1966. The combined acts changed the same part of the government code. The Traffic Safety Act, which covered Titles I and II, focused on creating and researching national automotive safety standards (Title I) and researching and building a national center for researching and building a laboratory for crash and other safety tests (Title II). The Highway Safety Act covered Title III, requirements for improved highway (road) safety standards, state-level safety mandates that were tied to federal highway funds, and research money for road safety studies. Yet, in the Senate, the issue was further complicated because, following traditional Senate responsibilities, two different committees covered the acts. Until the creation of the Department of Transportation, the Commerce Committee oversaw vehicle safety issues and so was responsible for the Traffic Safety Act. The Public Work Committee’s Subcommittee on Public Roads took responsibility for the Highway Safety Act.

Secretary of Transportation only to Federal departments or agencies, State or local agencies, or persons or organizations engaged in research into highway safety.”

Nader believed that this section would inevitably brew harmful secrecy of both sorts he identified. He pointed out that most evidence in these reports “proceeded from physical evidence rather than from orally attributable evidence,” so there were no sources to protect.16 He also gestured to publicly available studies, such as those done by Paul W Gikas and Donald F. Huelke at the University of Michigan Medical

School, who also made themselves available for court proceedings. For Nader, this sense of openness, a central pillar of the so-called Mertonian norms of science, was necessary for a just society: “In the area of human safety, few practices are more deplorable than secrecy. Yet, the inducements to secrecy, absent legislative safeguards and legislative review, are numerous in an area where commercial interests have such a strong interest in the outcome of studies, investigations, and standards setting.”17

To bolster his argument, Nader pointed to several cases where, he felt, secrecy had a negative impact on consumer safety and justice. In 1965, the

Commerce Committee had held hearings on the issue of automotive tire safety.

Nader asserted that the automakers continuously “strive to envelop these tires within the rigors of corporate cost reduction,” which inevitably led to thinner, weaker, more poorly made products.18 As the Commerce Committee investigated

16 Senate Committee on Public Works, Highway Safety Act: Hearings on S. 3052, 89th Cong., 2nd sess., 1966, 248. Robert K. Merton, “The Normative Structure of Science” in The Sociology of Science: Theoretical and Emirical Investigations (Chicago: University of Chicago Press, 1973), 267-280. 17 Ibid. 18 Ibid.

39 tire production, it would have learned about the finances behind it—for instance, how much the automakers paid tire manufacturers for each piece. Nader believed that this information should be public because it would be helpful to the courts and to safety advocates to know “just how little was involved in procuring a safer tire” and about the “tire companies’ predicament in having to sell subquality tires to their giant and uncompromising buyers.”19 Nader’s liberal vision did not include corporate sovereignty; “trade secrets” should not be sacred when the public’s welfare was at stake. Nor did his vision have any room for market-centered notions of “buyer beware.” Rumors of crashes and injuries would not effectively enough discourage consumers from buying a particular car.

Nader often pointed to the Civilian Aeronautics Board (CAB) as an example of what he desired in the new auto safety agency, though he was ambivalent about some of CAB’s rules. Created in 1938 and made into an independent agency in 1940, the CAB was in charge of aviation accident investigations until 1967 when the newly formed National Transportation Safety Board took over the CAB’s investigative functions. As Nader indicated, the CAB’s final report on any accident was inadmissible in court, but the many exhibits that supported it—including “reports, photographs, diagrams, maps, charts, and personnel, training and proficiency records of the [airplane flight] crew, and history and maintenance records of the airplane”—were all admissible in precedent.20 Furthermore, CAB employees could testify about the “facts” of the case, though they could not share their “professional opinions” about the crash’s cause. The Senators who listened to Nader’s testimony

19 Ibid. 20 Ibid.

40 pointed out that there were only a few plane crashes a year to which the CAB employees had to attend, but that there were hundreds of thousands of auto accidents. Nader more or less dodged this issue. He still believed that the government should play this role of truth-maker.

To summarize, the twin auto safety acts were borne out of Ralph Nader’s highly politicized vision of auto safety. His vision of the agency differed both from what the agency became and how the eventual leaders of the agency understood it.

He put a premium on the role of lawyers and courts even in future developments of automotive safety, and he believed the government should play a central role in producing knowledge for this process. He argued that the future agency should both force the automakers to adopt basic safety standards and publicly demonstrate the automakers’ moral culpability in auto crashes.

When it came time to choose the first Administrator of the NHSB, however, the Johnson administration turned away from consumerist roots of automotive safety. Against this highly politicized background, William Haddon’s medicalized vision of auto safety must have recommended him as a safe alternative. Haddon did not discuss politics in his pre-NHSB writings. In his rendering, the automakers were not the enemies of automotive safety. The enemy was both older and more diffuse; it was superstitious ignorance, which understood safety issues in a “pre-scientific” way.

Born in Orange, New Jersey, in 1926, William Haddon, Jr. received an S.B at

MIT in 1949 and an MD from Harvard four years later. In 1957, he earned a Masters

41 in Public Health, also from Harvard. Between his MD and MPH, he worked as an intern at Philadelphia General Hospital and a fellow at the National Foundation of

Infantile Paralysis at Harvard’s School of Public Health. During that time, Haddon became interested in injury as a public health issue. When he graduated from the

School of Public Health, at the age of only thirty-one, Haddon became the director of the driver research center in the New York State Department of Health. Haddon’s youth in attaining this position was a symbol of two things: his obvious intellectual gifts and ambitions and the fact that driver safety had not become an issue that demanded an older, more established expert. He maintained the position until 1961 when he entered a residency program in epidemiology where he stayed through

1965.

Haddon became a vocal proponent of highway safety. Through safety networks in New York State, Haddon met Hugh De Haven, the so-called “Father of

Crashworthiness.”21 A professor of aeronautics at Cornell University, De Haven made many essential contributions to crash science, but one of his most important formulations was of the “second collision.” He argued that, in vehicular crashes, humans were not injured in the “first collision,” the impact between the vehicle and another object, but in the “second collision” between the person and the vehicle in which they were riding. As Nader pointed out, sharp objects, including door handles and dashboards, only aggravated this second collision. Haddon adopted this

21 In crash safety circles, De Haven’s story is legendary. In 1917, De Haven survived a mid-air plane collision that killed all others involved. He became obsessed with crash safety and argued against popular wisdom that crashes were survivable if the body’s impact could be spread across both time and space, that is, if the impact could be slowed and if more material absorbed the force of the collision.

42 concept for his future studies, and he became an advocate for changing vehicle design to mitigate injury.

Unlike Nader, however, Haddon did not focus on the idea that increased crash deaths came from the intersection of increased driving and immorally bad vehicle design. Rather, these deaths came from bad science, science that had been successfully applied in other areas. In 1964, Haddon, along with Edward A.

Suchman, a professor at the University of Pittsburgh, and David Klein, a researcher at the of the Association for the Aid of Crippled Children, published Accident

Research: Methods and Approaches. The work became, for many years, the most important book in its field, a kind of textbook. Its design was partly intentional.

Drawing together a number of historically important studies dating to 1948,

Accident Research included a running commentary in which the authors assessed both what advances the studies had made and where they had gone astray. The book was meant not just to bring together the state of accident prevention knowledge but also to suggest a unified approach that others should adopt; it desired to be authoritative, if not hegemonic. In the unsigned and probably team- written first part of Accident Research, Haddon and his co-authors asserted that accidents were the primary cause of death for young people, but that this was the result of gains in decreased mortality in other areas, such as childhood disease.22

They wrote, “Accident mortality, at least in Western countries, has ‘risen’ by remaining essentially unchanged while dramatic reductions have occurred in the

22 Both as a scientist and as a federal employee, Haddon often co-wrote essays and reports. I will tentatively attribute an idea in one of these coauthored works to Haddon if the idea also occurs in one of his solo written pieces, especially if his single-author work appeared before the coauthored one.

43 incidence of deaths of other types.”23 They went on to claim that “changes in culture and technology change the incidence of specific kinds of accidents (railroad fatalities decrease and motor vehicle fatalities increase), but the substantially unchanging over-all totals tend to lend credibility to the widespread popular belief that accidents, like the poor, we shall always have with us.”24 Haddon and his coauthors cast themselves as Enlightenment figures, often referring to received notions of accidents, accident prevention, and fate as “folklore” and as “traditional, prescientific wisdom.”25

That the Johnson administration and agencies produced as part of the Great

Society would depoliticize an issue by making it a technocratic one is not in itself surprising. More precisely, the Johnson administration and Haddon were able to legitimate an intrusion into the private spaces of corporate decision-making and consumer behavior by discussing auto crashes as a disease. This move was a Great

Society tact par excellence. The tendency was probably best embodied in the person of Daniel Patrick Moynihan, whose work on both race and highway safety ties together two seemingly disparate wings of the Great Society movement. Moynihan had worked closely with Haddon beginning in the late 1950s, and Moynihan wrote

23 William Haddon, Edward E. Suchman, and David Klein, Accident Research: Methods and Approaches (New York: Harper and Row, Publishers, 1964), 2. 24 Ibid. It is impossible to tell from this passage whether Haddon and his co-authors believed that poverty, like auto crashes, did not need to always be with us but rather could be aided through expert intervention. Yet, perhaps not coincidentally, only a few years later, Haddon would become the Administrator of a federal agency at the same moment that the President of the United States pushed his “war on poverty.” 25 At a deep level, this belief that traditional views of accidents were caught up in folklore and superstition was tied to science’s suspicion of what would later be called “indigenous knowledge.” This suspicion went hand-in-hand with top-down Cold War “Third World” development initiatives, and it came under heavy fire from academics during the ascendancy of postmodernism.

44 his essay, “Epidemic on the Highways,” under Haddon’s tutelage.26 Furthermore,

Johnson echoed this language when he signed the acts into law: “For years, we have spent millions of dollars to understand and fight polio and other childhood diseases.

Yet until now we have tolerated a raging epidemic of highway death—which has killed more of our youth than all other diseases combined. Through the Highway

Safety Act, we are going to find out more about highway disease—and how to cure it.”27 Later, in 1965, Moynihan published the perennially controversial The Negro

Family: The Case for National Action, which included the infamous suggestion that structural racism created a “tangle of pathology” in black families. That Moynihan turned to this rhetorical strategy (at least) twice suggests that the rhetoric of disease and faith in expertise were an easy match. Inversely, Haddon’s rendering of safety as an issue of superstition and ignorance also fit with Moynihan’s and the

Democratic party’s portrayal of racism as arising from slack-jawed yokelism.

In opposition to pre-scientific ideas of crash safety, which Haddon denigrated as merely “descriptive,” he proposed an “etiological” approach. In spelling out this divide, he turned to analogies from bacterial and viral diseases. For instance, prescientific medicine relied on description when addressing the symptoms of fever and wasting. By focusing on the symptoms instead of root causes, doctors improperly believed “wasting” to be an independent illness instead of a symptom of a number of diseases, including protein deficiency, amebiasis, and tuberculosis. For

26 Daniel Patrick Moynihan, “Epidemic on the Highways,” The Reporter, Vol. 20, No. 9 (April 1959), 16– 23. Later, as the Director of the Joint Center for Urban Studies of the Massachusetts Institute of Technology and Harvard University, Moynihan led and wrote a less well known study on highway safety. U.S. Department of Health, Education, and Welfare, Report of the Secretary’s Advisory Committee on Traffic Safety (Washington, DC: US Government Printing Office, 1968). 27 “The White House Message on Transportation,” National Archives and Records Administration, Record Group 416, Finding Aid UD-UP-3, Box 1, Folder 2, 1.

Haddon, the transition from description to etiology was the paradigmatic mark of modern medicine. In a Kuhnian way, Haddon believed that medical science was hindered by people who held tightly to the paradigm of the descriptive model with diseases that included “‘chronic bronchitis,’ ‘cancer,’ and ‘diabetes.’”28 He insisted that, if crash science and accident studies more generally were going to move forward, they had to make this transition. As Haddon wrote, “The notion of an accident is descriptive, not etiologic.” He went on:

“It has a long history and close relationship to notions of personal vulnerability and invulnerability. This is illustrated, for example by the themes of the stories of Job and Achilles. Redolent of the extrarational and supernatural and prescientific, they remind one of Malinowski’s natives in their approach to the hazards outside the reef, which they did not understand, in comparison with those inside which they approached in more rational terms.”

The etiology of human injury was clear, however; its “basis is the various forms of energy exchange which must occur in excess of body injury thresholds for the injuries which make the field of such current social concern to occur.”29 These energies could be further broken down into their constituent forms, including

“thermal energy, ionizing energy, mechanical energy, electrical energy, and chemical energy.”30 The enemy of auto safety was not any human agent, whether individual or corporate; it was energy.

To conceive fully of human injury as a disease, Haddon created his famous

Haddon Matrix, a graphical device used to model injuries and create distinctions

28 Haddon, “The Changing Approach to the Epidemiology, Prevention, and Amelioration of Trauma: The Transition to Approaches Etiologically Rather than Descriptively Based,” American Journal of Public Health, Vol. 58. No. 8, (August,1968), 1433. 29 Ibid. 30 Ibid.

46 between them. Again, Haddon turned to a viral infection—this time polio—as an analogy to injury. The Haddon Matrix broke injury and illness into three moments, or “phases of social concern.” The first phase dealt with the agent reaching the human host and ways to prevent it. In polio, of course, this phase arose from humans’ coming across the polio virus, and, as Haddon noted, it was avoided by different means including keeping children away from swimming pools and movie theaters. In auto safety, Haddon called this the “pre-crash phase.” In the second phase, or “crash phase,” the question was what to do once the agent and host had come together. With polio, this included forms of drug treatments. In auto safety,

Haddon focused on how the automobile, especially its interior, could be transformed to lessen injuries during crashes. Over the next few years, Haddon predicted, automakers would bring about “the far better ‘packaging’ of human cargo, accomplishing with people what we long ago achieved with property.”31 Finally, the third stage, or “post-crash phase,” in Haddon’s words, “involves maximizing salvage, once damage has been done to the susceptible structures.”

Strikingly, Haddon’s language not only ignores the issues of morality and politics, it also dehumanizes the person. Haddon reduced the human body to a

“susceptible structure.” If Haddon sang the body electric, he did so only when the body became a conductor of energy and, thus, was at risk of injury.

Haddon’s matrix was his most lasting contribution both to the safety literature and to the National Highway Traffic Safety Administration itself. The

Haddon matrix became embodied in the very classification of the agency’s

31 Thus, in Haddon’s mind, Hugh De Haven’s dropping eggs onto foam was not at all a far-fetched analogy to human safety as cargo.

47 standards. The agency’s standards are broken down into 100-, 200-, and 300- level standards. Thus, Haddon’s pre-crash phase comprises the 100-level standards; the crash phase, the 200-level standards; and the post-crash phase, the 300-level standards. For instance, Standard 108 deals with “Lamps, Reflective Devices, and

Associated Equipment,” which serve both to illuminate the road for the driver and to make his or her vehicle more conspicuous to other drivers, both functions that decrease the likelihood of accidents.32 Standard 301 addresses “Fuel System

Integrity.” Historically, the 200 standards have been the most controversial and have elicited the fiercest battles from automakers. Dealing with the moment of the crash, the 200 standards go to the heart over disagreements over the moral culpability of product design and the responsibilities of drivers, disagreements that remain unresolved to this day.

Most importantly, the Haddon Matrix divided research labor. A section from the Moynihan-led Report of the Secretary’s Advisory Committee on Traffic Safety, published in 1968, closely mirrors Haddon in this regard:

“Such an environmental health and consumer protection activity should be built around etiologic rather than descriptive concepts. This is to say that the mission of any such organization must be seen as the control of environmental hazards rather than, for example, the prevention of ‘accidents,’ and the characteristic leadership role in such an organization must be held not by ‘experts’ in traffic safety, food poisoning, household falls, or whatever, but rather by persons trained in various intellectual disciplines—medicine, epidemiology, biology, mathematics, psychology, physics, political science,

32 400 and 500-level Safety Standards describe requirements that do not fit in the Haddon Matrix. For instance, Standard 403 contains safety requirements for automotive “platform lift systems” used to move people who use, for instance, walkers or wheelchairs. Standard 531 deals with average fuel efficiency standards, also known as CAFE (Corporate Average Fuel Efficiency) standards.

sociology—who bring their skills to bear on this particular problem area.”33

What Haddon called “Technics in Use” followed out of this division of research labor and was only effective if it was preceded by a shift to the etiological way of thinking. After scientists and doctors shifted thinking away from its prescientific roots and examined the causes of human injury, technicians could come in and apply their design lessons. Pure, abstract research would precede application.

In this way, Haddon was very much a child of his times. As historians have discussed, after World War II, scientists and policy makers adopted a vision of innovation that has come to be known as the “linear model.”34 This framework holds that science aimed at gaining propositional knowledge about how the world works precedes prescriptive knowledge, or knowing how to remake the world to fit our desires.35 Where most advocates of the linear model focused on the “hard sciences,” such as physics, chemistry, and biology, Haddon’s version also included room for the social scientists, but he valued even their work for its potential as ways of coming to know the world fundamentally, rather than for the practical knowledge that the social sciences had already won. As we will see, this is almost diametrically opposed to Brenner’s understanding of the safety issue.

33 Report of the Secretary’s Advisory Committee on Traffic Safety, 23. 34 On the linear model, see Donald Stokes, Pasteur’s Quadrant: Basic Science and Technological Innovation (Washington, D.C., Brookings Institution Press, 1997) and Benoît Godin, “The Linear Model of Innovation: The Historical Construction of an Analytical Framework,” Science, Technology, and Human Values, Vol. 31, No. 6 (November 2006), 639–667. 35 Joel Mokyr spells out the helpful distinction between propositional and prescriptive forms of knowledge in his The Gifts of Athena: Historical Origins of the Knowledge Economy (Princeton: Princeton University Press, 2002).

Haddon was willing to go quite far in his abstraction of accident prevention.

He wrote,

“For example, within the highway area, this broader approach, based on the etiologic agents which must be involved, picks up in the first phase [the crash phase] the occasional murders and suicides which many of us working in this field have occasionally come across. The approach also picks up a number of equivalents that involve individual and organized violence and aggression whether on the civilian scene or in war. Thus, one can identify for study, across all human violence, the roles of alcohol, views of risk, notions of individual vulnerability, and subjective probability.”36

When dealing with automotive crash safety, Haddon believed that the focus should be put on changing the automobile, not on shaping driver behavior. At root,

Haddon believed that humans were essentially fallible and beyond perfection or perfecting. Throughout his career, Haddon made many arguments for his focus on technology. Yet, one of the simplest is this: Haddon made an analogy between auto safety and the dangers of early electrical systems. He pointed out that, instead of training people to use electrical systems safely to prevent fires and other hazards, electrical engineers created fuses.37 Years later, we see fuses as “natural” components of electrical systems, but they were erstwhile engineering interventions that circumvented human fallibility (and non-human system failure as well). But Haddon had little understanding of how these technologies would develop or ideas about the nature of technical change.

36 Haddon, “The Changing Approach,” 1437. 37 Haddon, “The Changing Approach,” 1436.

Often, the heads of federal agencies play public roles, including interacting with the

White House and Congress and making speeches before the public, that pull them from their supposed jobs. In these instances, they have to rely on their subordinates to carry on the work of their agencies. Without doubt, in such circumstances,

Haddon turned repeatedly to one man, his Deputy Administrator, Robert Brenner, who far more than Haddon, shaped the NHSB in ways the endure.

Brenner occupies our focus for two primary reasons. First, when the Johnson administration created a group within the Department of Commerce to plan the future Department of Transportation, it was Brenner, not Haddon, who was a member. Brenner took part in the day-to-day activities of the several-month project, and he had a major influence on the organizational structure of the NHSB.38

Although we can imagine that Brenner was in regular contact with Haddon about the plans, Haddon was not a part of these quotidian affairs. Brenner was able to build his own conceptual understanding of auto safety into the agency’s “genes.”39

Second, of the thirty-three NHSB’s Executive Meetings that have records in the archives, Haddon attended exactly six of them. Moreover, the last meeting we know

Haddon attended, occurred on January 16, 1968. For the rest of the year and until

Haddon's departure in February 1969, Brenner ran the Executive Meetings. At most meetings, Haddon's personal assistant, Joan Claybrook, attended, and we can imagine that she fostered communication between Haddon and those who attended

38 The records of this Federal Highway Agency Task Force are contained in National Archives and Records Administration, Record Group 416, Finding Aid UD-UP-3, Box 2, Folder 11, “FHA Task Force Proposed Organization.” 39 The economist Kenneth Arrow occasionally used such “genetic” language in his The Limits of Organization (New York: Norton, 1974).

51 the meeting. (Claybrook would herself become NHTSA's administrator under the

Carter Administration.) From his position as chairman of this weekly meeting,

Brenner had a powerful way to shape the agency. For instance, the November 21,

1968 meeting notes that Brenner had been giving presentations on “Dr. Brenner’s underlying philosophy of NHSB. Found to be most helpful to NHSB personnel in their approach to work.” Moreover, Brenner had several other means by which to shape the agency. After Haddon left the agency, Brenner became the Acting

Administrator through 1970. Then, from 1971 to 1986, Brenner became the agency’s Chief Scientist, an important, if largely honorary, position in the organization. Brenner’s affiliation with the agency ran deeper and longer than

Haddon’s.

Born in 1922, Brenner attended the University of Pennsylvania as an undergraduate, receiving a Bachelors in Mechanical Engineering in 1943. He worked first as a designer at Lockheed Aircraft for the remainder of World War II, and after the war he moved to American Canning Company, where he worked as a plant engineer. In 1948, Brenner entered a Masters program in Engineering at the

University of California Los Angeles (UCLA). After receiving his degree in 1949, he became a faculty member at UCLA’s Institute for Transportation and Traffic

Engineering (ITTE), a position he held until he began working for the federal government in 1966.40

40 The California Legislature formed the ITTE in 1947. Partly, the Institute arose because Los Angeles and other major Californian cities developed in ways that were heavily dependent on the automobile. Complaints about traffic in the City of Angels went back at least to the late 1930s. Additionally, legislators believed that infrastructural maintenance had been neglected during the Depression and World War II. Thus, the ITTE was to study comprehensive California’s transportation systems. The institute was an

In 1950, Brenner published an essay, “A Segmented Electrical Element for

Detecting Vehicular Traffic,” which he co-wrote with other members of the ITTE, including John H. Mathewson, who played a central role in the ITTE and in highway safety research more generally. This was his first major publication as well as his first work published through Highway Research, a publication of the National

Research Council’s Highway Research Board, a part of the NRC’s Division of

Engineering and Industrial Research. Many of Brenner’s early works were necessarily focused on developing methodologies and instruments for the study of traffic and crash safety.

In 1960, Brenner became the Secretary of the Highway Safety Committee of

NRC’s Highway Research Board. Charles W. Prisk, the Deputy Director of the U.S.

Bureau of Public Road’s Office of Highway Safety, chaired the committee. Though

Brenner had been a member of and made presentations before a number of national organizations, this was the first time he held a leadership position in one. His rise from that point was quick. In 1962, the same year that he received his Ph.D.,

Brenner became a member of the National Institutes of Health’s Accident Prevention

Panel, which was formed through a special grant from NIH’s Division of Grants and

Research.41

At least four aspects of Brenner’s thought and action distanced him from

Haddon. First, Brenner placed much higher value on the role engineers and practitioners of highway and road management in increasing automobile safety. In important center for automotive safety research, and Brenner worked on both auto safety and traffic control issues. 41 Brenner, “Analysis of Traffic Flow Using ‘Time’ as the Dependent Value” (PhD. diss., University of California Los Angeles, 1962).

1965, Brenner presented and published a brief essay entitled “The Future of

Highway Research.”42 While Brenner noted that “more widely diversified disciplines and skills are now being directed toward accident research than ever before,” he also emphasized that “operational personnel,” such as engineers of California

Division of Highways, must play a central role in highway safety research. These engineers were precisely the kind of experts that Haddon sought to move away from. Brenner recounted a study published that year, showing that many lives could be saved by “a 5-cent washer inserted behind the head of bolts and nuts used to hold guardrails in place.” Brenner said that he was unaware of how researchers discovered this simple but important insight. It could have come from a detailed study of guardrails, but he believed that it more likely came from “luck.” He asserted that luck had no “derogatory” meaning here and that many important discoveries—including radioactivity and penicillin—were discovered largely through luck. The important thing, in his view, was that chance only became serendipity if the person looking at the data had keen eyes for what could be fortuitous breakthroughs. As Louis Pasteur famously noted, “Chance favors the prepared mind.” The person had to be able to see the signal in the noise. As he wrote, “But these discoveries were made by people who had the skill and experience to understand a problem and recognize a solution, even when it appeared by chance.”43 For Brenner, unlike Haddon, this essential experience came from familiarity with the problems and dangers of the road.

42 Brenner, The Future of Highway Research,” Highway Research Record, No. 79, 1965, 58-9. 43 Ibid., 59.

Thus, if Haddon was an advocate of the linear model, Brenner was a champion of what Donald Stokes called “Pasteur’s Quadrant.”44

Second, in contrast to Haddon’s supposed single-minded focus on changing automotive design, Brenner was well aware of the road’s role in driving safety, and several of his colleagues at ITTE studied the human dimensions of crash risk. For

Brenner, the key metaphor and model was always the system. In this way, Brenner was very much a product of his times. The paradigm of systems reigned supreme in many intellectual circles, including at the RAND corporation and in Jay Forrester’s studies of system dynamics.45 Brenner was willing to go quite far in his systems thinking. He viewed the network of roads as a thermodynamic system made up of several interconnected subsystems.46 Whereas classical thermodynamics focused on energy, Brenner’s fundamental variable was “travel time.” He took pains to emphasize that he saw thermodynamics as a model and metaphor. “Clearly energy transformations take place during the accomplishment of the transportation factor,” he argued, but where others had taken lessons from gas laws and classical mechanics when applying such models to traffic—arguing, for instance, that, like gases, “the people in a city will move about more and more as the total energy supplied to a city is increased”—Brenner called traffic “thermodynamic-like.”47 Yet,

44 Stokes, Pasteur’s Quadrant. 45 On RAND, see David Hounshell, “The Cold War, RAND, and the Generation of Knowledge, 1946- 1962,” Historical Studies in the Physical and Biological Sciences, Vol. 27, No. 2, 1997, 237-267; David R. Jardini, “Out of the Blue Yonder, The RAND Corporation’s Diversification into Social Welfare Research, 1946-1968 (PhD diss., Carnegie Mellon University, 1996); and Jennifer S. Light, From Warfare to Welfare: Defense Intellectuals and Urban Problems in Cold War America (Baltimore: Johns Hopkins University Press, 2003). 46 Brenner diss., 1. 47 Brenner diss., 2-3. Brenner’s quotation on the relationship between people and gas laws comes from R. T. Howe, “Can the Urban Transportation Problem Be Solved?”, Traffic Quarterly 14: 85-95, Jan. 1960.

55 if Brenner was not willing to push the analogy between thermodynamics and traffic too far, he did see the world in systemic terms. After pointing to H. G. Wells’s argument that the structure of a nation’s cities and the distribution of its population were essentially dependent on its “transportation facilities,” Brenner approvingly followed LLewellen Boelter, UCLA’s dean of engineering, in claiming that “a city is a

‘system’ having as a general purpose the optimization of the transportation function.”48

When designing the NHSB’s organizational structure, Haddon and his team members turned once again to this organizational motif and built “system” into their conception of the agency. As their proposed mission statement reads:

“The end results of concern—the safe and efficient movement of people and goods to achieve defined social goals—depend upon several mutually complementary factors comprising a coherent total system. The [NHSB] deals with the central factors which include (1) the vehicles, (2) the traveled way, (3) human factors in handling vehicles on the traveled way, and (4) the storage of vehicles and goods in transit.”49

Third, where Nader placed the failures of auto safety at the feet of politics and Haddon placed them at the feet of superstitious ignorance, Brenner attributed unsafe cars to the misfortunes of economics, to what is sometimes called “market failure.”50 In this case, the failure arose because automakers had no incentive to

48 Brenner diss., 12-13. H. G. Wells, The Works of H. G. Wells, Vol. 4, “Anticipations and Other Papers” (New York: Scribner, 1904). Llewellen M. K. Boelter, “Comments on the Theory of Planning,” Paper presented at the National Conference of American Society of Civil Engineers, Los Angeles, Feb. 1959, 9p. 49 “Proposed Statement of Mission. Federal Highway Administration. Department of Transportation,” National Archives and Records Administration, Record Group 416, Finding Aid UD-UP-3, Box 2, Folder 11, “FHA Task Force Proposed Organization,” 3. 50 The connection between liberalism and engineering in the United States has not been systematically explored, but important earlier studies, including Edwin Layton’s The Revolt of the Engineers, David F. Noble’s America by Design: Science, Technology, and the Rise of Corporate Capitalism, and Bruce

56 create safer cars given the nature of consumer demand. Brenner was part of the trend in engineering (that continues to this day), which saw an increasing role of economics in engineering practice. Once called “engineering economy,” later known primarily as “cost-benefit analysis,” this development applied the “lessons” of economics, including later the mathematization of economics, to engineering projects, for instance, by comparing the costs and efficiencies of different solutions.51 Economics—especially the idea, if not the language, of “market failure”—was Brenner’s primary tool for thinking about social problems and their solutions. He used the example of tire low pressure warning systems to demonstrate the problem.52 He argued that the “social need and attendant market alone were not enough” to get warning systems into vehicles, but a new standard that the agency was considering would create a “major market” and induce automakers and their suppliers to create such technologies.

Fourth, when Brenner considered the safety features of the automobile, which constituted but one part of the overall system, he thought about it in terms of engineering design. He believed this aspect of safety was best accomplished by creating new and expanding connections—or networks— between people who were focused on design, including automakers, parts suppliers, academics, and government bureaucrats from around the world.

Sinclair’s Philadelphia’s Philosopher Mechanics, suggest that engineering in the US has always been a thoroughly liberal institution and that engineering education inculcated its students with these values. 51 The full history of this trend remains to be written, but Theodore Porter has written about the development of Cost-Benefit Analysis in Trust in Numbers: The Pursuit of Objectivity in Science and Public Life (Princeton: Princeton University Press, 1995). 52 Brenner, “Automobile Design and Safety—Technology Tradeoffs and Transfer Processes” in Space Technology Transfer to Community and Industry, eds. Ralph H. Tripp and John K. Stotz, Jr. (Tarzana, CA: AAS Publications Office, 1972), 173.

In March 1972, Brenner attended a meeting of the American Astronomical

Society, which was holding its Tenth Annual Goddard Memorial Symposium on the theme “Transfer of Space Technology to Community and Industrial Activities.” The eminent Wernher von Braun gave the opening speech, entitled “Earth Benefits from

Space and Space Technology.” Von Braun’s opening remarks give little doubt about the context of that year’s symposium: the space program was under siege. Its budget was being cut, the public had become bored with the Apollo moon program, and critics claimed that money spent by NASA was money wasted on frivolous projects that would recoup little of practical value. Von Braun believed that the space program’s defenders had to demonstrate the program’s practical consequences in both tangible results, such as technological “spillovers,” and intangible ones, including “national prestige, spirit, and life concepts,” which von Braun said “may be the most important legacies of our activities in space.”53

Brenner actively participated in this ideologically-charged project. His involvement in the symposium suggests a belief in the progress-oriented developments in postwar technology, but, more importantly, his words there also display how deeply an engineering-centered paradigm informed his understanding of automotive safety. Auto safety was plainly and simply an engineering problem.

The space program, with its attention to difficult design problems, was a perfect partner to auto safety: “At the moment of impact—be it vehicle to vehicle, vehicle to roadside object, or vehicle to some child or elderly person in a pedestrian

53 Wernher von Braun, “Earth Benefits from Space and Space Technology” in Space Technology Transfer to Community and Industry, eds. Ralph H. Tripp and John K. Stotz, Jr. (Tarzana, CA: AAS Publications Office, 1972), 3. Speaking of “technological spillovers” is strictly anachronistic. None of the conference’s participants used this current term. Nearly all of them used the more vague word “benefits.”

58 crosswalk—we have one of the most hostile environments of modern civilization.

And one created entirely by man. You of aerospace community have successfully dealt with hostile environments, and in the process have generated a vast body of new scientific knowledge.”54

To illustrate his point, Brenner discussed a technology known as “Tor-Shok,” which, though created by NASA, had found uses in highway safety applications.55

The device’s name was a clever play on its configuration; it was a torus-shaped piece of alloyed metal that utilized the “principle of cyclical straining of metal” to form an energy-absorbing structure. For Brenner, the most important aspect about the story of the device’s transfer from NASA to auto safety was that, in NASA’s view, it was a failed technology. Tor-shok was not used in the Apollo Lunar Excursion Module, for which it was designed. Yet, because of “personal contact” between people in the space field and people in the automobile safety field, its wider potential applications became evident. Moreover, because regulators at the NHSB knew of Tor-shok and, thus, knew that the technology was “available,” they were more confident about making new regulations.

For Brenner, the analogy about Tor-shok was important because it showed an alternative model for effective government regulation. He pointed out that people traditionally took two stances regarding regulation—either for or against technology-forcing—and, either way, all of the research was on the industry’s

54 Brenner, “Automobile Design and Safety—Technology Tradeoffs and Transfer Processes” in Space Technology Transfer to Community and Industry, 169. 55 Once he arrived at NHSB, Brenner had the agency open a contract to investigate Tor-shok’s potential in highway protective barriers. See Bureau of Public Roads and Aerospace Research Associates, “A Reusable Energy Absorbing Highway Protective System for Median Areas,” Bureau of Public Roads Report #96, published 1968.

59 shoulders.56 But Brenner saw a third path, “namely, one in which Government would be financing a technology resource that has the mandate to advance the state of the art of vehicle design free of immediate marketplace constraints or pressures to regulate production vehicles.”57 Brenner believed that NASA and the rest of the federal government had already played this role, but that it could do so in a more

“systematic” way. Again, as he had done earlier in the planning of the DOT, Brenner highlighted the importance of systems thinking. He argued that Tor-shok, along with other technologies he mentioned, “was not a chance occurrence insofar as it arose out of a deliberate program to identify non-space applications.”58

Brenner believed that the most important program within the NHSB in these terms was the Experimental Safety Vehicle (ESV) program. Through this program, the NHSB had hired four independent contractors—Fairchild Industries, American

Machine and Foundry, General Motors, and Ford—to build ESVs, cars that had a plethora of cutting-edge safety technologies meant to explore the outer limits of vehicle safety and create a “quantum jump in vehicle performance.”59 Beginning in

56 In regulatory thought, a regulation is called “technology-forcing” if it requires producers to create a technology or enhance some technological capability that currently does not exist in order to satisfy regulatory requirements. The technology is not “off-the-shelf,” that is, ready for production, and it will require research before being available. Since the beginning of the NHSB, there has been controversy over the question of whether the agency has legislative authority to create technology-forcing regulations. The Traffic Safety Act is ambiguous on this point, and a person’s stance, pro or con, on the issue often unveils his or her political stripes vis-à-vis auto safety and government more generally. Nader has always maintained that the goal of the Traffic Safety Act was to push the industry to develop and adopt new, better technologies; the auto industry has invariably insisted that the agency’s regulations must be based on already-existing standards and off-the-shelf technologies. For its part, the NHSB and NHTSA adopted and continues to maintain the policy that the agency must demonstrate that technology is available at a reasonable cost before the agency can promulgate new standards. 57 Brenner, “Automobile Design and Safety—Technology Tradeoffs and Transfer Processes,” 184-5. 58 The literature on technological spillovers and the issue of whether they result more often from “pure” or “applied” research is huge and dizzying, but the best, most succinct synthesis of the issue is Donald Stokes, Pasteur’s Quadrant. 59 Brenner, “Automobile Design and Safety,” 176, 178.

1971, the NHSB also took part in the first international ESV conference, held that year in Paris. At both levels—first national, then international—Brenner believed that the ESV program achieved several things for the agency: it created networks between different relevant social groups. It created knowledge for the agency about state-of-the-art safety technologies, which was important to bolster its case for new or more stringent standards. Finally, related to an earlier point, it went part way to solving the market failure problem by generating new technologies for the auto companies.

Over time, the NHSB and its successor, NHTSA, became an agency where the vast majority of employees (not counting secretaries, janitors, and other essential support staff) were and are members of the Society of Automotive Engineering

(SAE). When the NHSB created its first auto safety standards, most of them were based on pre-existing SAE standards. Engineering culture has thoroughly dominated the agency. While the agency arose from the politicized discourse of consumer advocacy, over the years, it has favored a reserved spirit of technical authority.

Where Haddon favored the “pure” research of doctors and scientists, the agency he briefly headed has always emphasized the more practical approach of engineers.

Doubtless, many factors encouraged the agency to develop in such a manner.

Legislators put emphasis on auto safety standards, and perhaps engineers were best equipped for this mission. Perhaps, they simply fit the prevailing understanding of the agency’s division of labor. But, in the end, understanding Brenner’s role in the agency helps to explain a seeming paradox—that the agency did not develop even in

61 its early stages according to the medicalized visions of its first “leader,” William

Haddon. As Robert Brenner attended to the tedium of planning and the work of mundane details, he made the NHSB a society of engineers.

Chapter 2—Constructing Standard 201 and the Limiting of Federal Automotive Safety Regulations

“We regret that this proceeding may take on the outward appearance of a contest between adversaries, but we have not approached it in that spirit”

—Lloyd Norton Cutler of the law firm Wilmer, Cutler, & Pickering, which represented the Automobile Manufacturer’s Association at the National Highway Safety Bureau hearings on Standard 201, May 22, 1967.1

“This proceeding was supposed to be a fact-finding investigation and not an opportunity to put on a show and try to make idiots of the federal government. And that is what it has done. And that is what they are trying to do”

—Robert M. Mahoney, NHSB’s counsel at its hearings on Standard 201, May 22, 1967.2

Introduction

From 1966 to 1967, the National Highway Safety Bureau (NHSB) designed

and promulgated the first twenty, nationally-binding safety standards. NHSB

Administrator William Haddon, Deputy Administrator Robert Brenner, and the civil

servants who comprised the new agency brought with them an eager, reformist

spirit. Yet, their work did not go unopposed. Like all federal agencies, the NHSB had

to put its standards through the steps mandated by the Administrative Procedures

Act. This entailed first publishing the proposed standards in the Federal Register,

then allowing for a public comment period, and, lastly, publishing the final

standard.3 As Brenner and others organized and built the agency, NHSB staff

1 “Transcript in the Matter of Reconsideration of Initial Motor Vehicle Safety Standard 201,” May 22, 1967, 9, Microfilm copy of selected subject correspondence relating to motor vehicle safety standards, 1966-70, General records, Records of the National Highway Traffic Safety Administration 1966-91, Record Group 416, National Archives at College Park, MD. 2 Ibid., 137. 3 I have heard rumors of a, perhaps apocryphal, flow chart used in the EPA to explain the agency’s standards-making process. After illustrating the steps mandated by the Administrative Procedures Act, the chart moves to final step wherein at least one regulated firm sues the agency, claiming that

62 63

prepared for the coming industry opposition that would undoubtedly greet their

efforts. They made ready for a fight.

This chapter examines the course of this standards-making process. At least

since the publication of Abbot Payson Usher’s A History of Mechanical Invention in

1929, historians of technology have taken interest in how technological artifacts

take their shape.4 Yet, I examine the shaping not of artifacts themselves but of the

the rule is impracticable, unworkable, and unfair and that it will most certainly destroy the industry. Thus, the courtroom becomes a de facto step and space for the regulations. 4 Abbot Payson Usher, A History of Mechanical Invention (New York: McGraw-Hill, 1929). Historians, economists, and other scholar have been trying characterize the geneses and patterns of technological change for decades. In his book, The Shape of Time: Remarks on the History of Things (New Haven: Yale University Press, 1962), the art historian George Kubler imagined the history of human artifacts as a series of “prime objects,” or true invention, and “replications,” mere emulation. The history of technology, therefore, was punctuated by brief periods of significant “invention” followed by longer periods of what other historians would call “normal” technology, spans of time in which “innovation” was incremental and slow-going. The historian of technology Edward Constant borrowed the language of “normal technology” versus “paradigm shifts”—moments of transformative and “revolutionary” change—from the historian and philosopher of science Thomas Kuhn. Kuhn had used the terms “normal science” and “paradigm shift” in his analysis of the Scientific Revolution. In Constant’s historical examination of aeronautical engineering, propeller-powered planes had assumed the state of “normal technology.” Engineers were slowly making small improvements in this technology. Engineers also realized that the propeller had a logical upper-limit, however. Because this limit had not yet been reached, Constant—following Kuhn’s description that revolutions in science came from answering “anomalies” between theoretical expectations and empirical observations—called the logical upper-limit a “presumptive anomaly.” The anomaly was “presumptive” because the propeller had not yet reached it. This presumptive anomaly eventually led to a paradigm shift and the birth of the “turbojet revolution.” Edward W. Constant, II, “A Model for Technological Change Applied to the Turbojet Revolution,” Technology and Culture, Vol. 14, No. 4 (Oct., 1973), 553–572; idem., The Origins of the Turbojet Revolution (Baltimore: The Johns Hopkins University Press, 1980); Thomas S. Kuhn, The Structure of Scientific Revolutions, 3rd ed. (Chicago: The University of Chicago Press, 1996 [1962]). The historian of technology George Basalla built, in part, upon Kubler’s insights in developing an evolutionary theory of technology. Basalla emphasized how variation and selection led to periods of continuity and discontinuity in technological change. Basalla, The Evolution of Technology (Cambridge: Cambridge University Press, 1988). In the 1970s and 1980s, James M. Utterback and William J. Abernathy began describing patterns of “innovation.” In their account, products, including automobiles, originally go through a period of “flexibility” and flux before a “dominant design,” or stable product structure, finally emerges. Utterback and Abernathy, “A Dynamic Model of Process and Product Innovation,” Omega, Vol. 3, No. 6 (December 1975), 639–656. The historian of technology Thomas Hughes theorized about change in “technological systems.” In Hughes’ model, technological systems move from moments of “invention” (such as Thomas Edison’s early experiments with the electric light) through a period of “system growth” until they finally attain what Hughes termed “momentum,” wherein a system acquires “mass, velocity, and direction.” Thomas P. Hughes, Networks of Power: Electrification in Western Society, 1880–1930 (Batimore: The Johns Hopkins University Press, 1983), 15; idem., “The Evolution of Large Technological Systems” in The Social Construction of Technological Systems, Wiebe E. Bijker, Thomas P. Hughes, and Trevor

tests and criteria that comprise performance standards by which, artifacts were

designed, manufactured, deployed, and, ultimately, judged to contribute to

automobile safety. Test-building involved artifacts and the construction of complex

systems—interconnections between instruments, infrastructure (such as labs and

crash barriers), humans running the tests, and the artifacts to be tested.

This chapter focuses on the history of federal automotive safety Standard

201—titled “Interior Occupant Protection.” (I will refer to this standard simply as

“Standard 201” throughout the chapter.) None of the first twenty automotive safety

standards offers a “representative” case, as if studying how regulators made one

standard would show how all twenty came to be. Each standard has its own

Pinch (Cambridge, Mass.: The MIT Press, 1987). Hughes applied his model to electrification, but it equally pertains to other systems. In terms of the automobile, Hughes’ model applies both to the automobile itself, which is certainly a complex system and one whose complexity increased over time, and to the infrastructure (roads, parking garages, etc.) and supportive institutions (such as automotive garages and dealerships) upon which users of the automobile depend. Some attacks on the automobile’s role in the United States resemble Hughes’s notion of momentum. Multiple actors in the United States—including the government, developers, and homeowners—have built the US physical environment, including its “suburban sprawl,” in such a way that its function depends upon the automobile. See, for instance, Jane Holtz Kay, Asphalt Nation: How Automobiles Took Over America, and How We Can Take It Back (Berkeley: University of California Press, 1998). For a more balanced view of the process of suburbanization, see Adam Rome, The Bulldozer in the Countryside: Suburban Sprawl and the Rise of American Environmentalism (Cambridge: Cambridge University Press, 2001). In the 1980s and 1990s, thinkers associated with the Social Construction of Technology (SCOT) “movement”—for lack of a better word—attempted to correct what they saw as a central error in writing on technology: people tended to treat technology as if it developed naturally, along its own path, and under its own powers. Historians and theorists of SCOT were replying to views such as the one inherent in the title of Langdon Winner’s Autonomous Technology. In the view of SCOT, interested parties “constructed”—or determined the ultimate shape of—technology through a process of negotiation. Yet, some of the concepts of SCOT resembled earlier ideas. For instance, the idea of “closure,” wherein a technology answers a previous problem and becomes “stabilized,” bears strong resemblance to Utterback’s and Abernathy’s notion of “dominant design” and Usher’s processes of emergence Trevor J. Pinch and Wiebe E. Bijker, “The Social Construction of Facts and Artifacts: Or How the Sociology of Science and the Sociology of Technology Might Benefit Each Other” in The Social Construction of Technological Systems, Wiebe E. Bijker, Thomas P. Hughes, and Trevor Pinch, eds. (Cambridge, Mass.: The MIT Press, 1987). 44–46; Wiebe E. Bijker, Of Bicycles, Bakelites, and Bulbs: Toward a Theory of Sociotechnical Change (Cambridge, Mass.: MIT Press, 1995); Langdon Winner, Autonomous Technology: Technics-Out-of-Control as a Theme in Political Thought (Cambridge, Mass.: MIT Press, 1977). While I believe that SCOT has its limits, it—along with all of the other theories described above—has greatly influenced my own examination of how multiple parties “constructed” regulatory tests.

background, its own story of becoming. Standard 201, however, was the most

contentious standard of the first score and so, is in some ways, the most interesting.

Standard 201 emerged from a complex process with many stages. Yet, throughout

each stage, the discussions and battles were waged not only between the two

“sides” of the regulatory equation—the regulators and the regulated—but also

between a host of other actors, especially experts from academia.

The sociologist of science and technology, Bruno Latour, and other theorists

of “Actor Network Theory” have argued that science, technology, and policies

thereof are matters of making and unmaking human and technical networks.5 The

interest-based politics of regulatory rule-making provides another example of

networks working to enlist new adherents. Regulatory rule-making is a process full

of interests, and often regulatory agencies themselves act as interest groups. In the

case of automotive crash safety regulations, the men and women who took the helm

of the agency were reformers who pushed an agenda of making cars more

“crashworthy”; they were not mere adjudicators of other external concerns. Indeed,

5 Bruno Latour, Reassembling the Social: An Introduction to Actor-Network Theory (Oxford: Oxford University Press, 2005). Latour’s image of science fits nicely with the “pluralist” account of politics, which depicts different “interest groups” vying for supremacy. Of course, gaining the allegiances of other groups is crucial for winning that supremacy. Many trace inklings of political pluralism back to the early history of the United States, particularly to the Tenth Federalist, but theoretical pluralism was definitively a twentieth-century development. Arthur F. Bentley, a student and sometimes collaborator with John Dewey, first wrote of the idea in his The Process of Government: A Study of Social Pressures (Chicago: University of Chicago Press, 1908). Yet, Bentley’s ideas languished in relative obscurity until the mid-twentieth century when it was “rediscovered” and reinvigorated by such important works as David B. Truman’s The Governmental Process: Political Interests and Public Opinion (New York: Alfred A. Knopf, Inc., 1951). Truman, who received his PhD from the University of Chicago and who was a professor at Columbia University, where John Dewey himself had moved after leaving Chicago, wrote The Governmental Process after teaching Bentley’s work for several years. Pluralism became the prominent theory during the 1950s and 1960s. Harmon Zeigler’s Interest Groups in American Society (Englewood Cliffs, NJ: Prentice Hall Inc., 1964) is a representative work, and it is helpful both for its historiography and for its special attention to “business associations,” which form the basis of chapter four. For a historical narrative about the birth of theoretical political pluralism, see Louis Menand, The Metaphysical Club: A Story of Ideas in America (New York: Farrar, Straus, and Giroux, 2001).

66 other forces within the federal government, including courts, sought to maintain balance between the agency’s concerns and those of other parties, overwhelmingly the automakers’. Of course, the goal of each party in such adjudicative processes is to win the decider’s approval, whether that person is a judge or a bureaucratic administrator. Winning the battle over the adjudicator—the judge or the hearing officer—often goes a long way towards winning the overall regulatory war.

In the conclusion of the chapter, I will explore a number of other works that have addressed the history of auto safety regulation in the United States. For many years, scholars—focusing primarily on air bags—have argued about why the NHSB and its successor agency, NHTSA, were ineffective at generating “technology forcing” regulations that would have compelled automakers to create new,

“innovative” safety technologies. Yet, as I argue, these scholars have systematically overlooked the process by which the NHSB set the first standards. Fallout from

Standard 201 severely limited the agency’s ability to create more stringent standards that were not based on already “existing standards.” Thus, where many scholars and safety advocates have discussed the “technology-forcing” nature of the auto safety standards, the real practical upshot of these first standards was just to get into mass-produced automobiles the “available” technologies, which were by no means new but which automakers had failed to adopt because they would raise costs and because manufacturers believed that “safety doesn’t sell.” If one believes that NHSB/NHTSA put too little pressure on the auto industry, one should look to

Standard 201 to find the roots of the agency’s neutered nature.

Making the Standards

Standards-making is always a complex process that involves many actors.

When governments are involved, legislative requirements often add to this complexity. In the case of auto safety standards, the Traffic Safety Act mandated

NHSB’s standards to be based on “already existing standards. How this requirement made it into the law is difficult to reconstruct from the legislative history. Ralph Nader played a large part in creating the first draft of the law, and he certainly would not have included such a requirement. A more likely scenario is that the stipulation was an eleventh hour compromise with the auto industry. The spirit of the requirement was to keep regulators from creating ex nihilo excessively strict standards. By requiring the new rules to be based on “already existing” ones, legislators hoped the new standards would be reasonable and achievable. This stricture necessitates a look further back in history to the origins of some of the

“already existing” standards that the NHSB used, particularly some earlier federal standards that were meant to improve auto safety for government employees.

The process by which these “already existing” standards came into being is largely invisible outside of reports that retroactively describe it. Archival records don’t give us a vision of the group who composed them. As the NHSB read administrative law, “An agency is under no legal duty to reveal the internal processes that shaped the project, and interested persons are not entitled thereon.”6 The NHSB probably adopted this reading as a defensive tactic: agency staff members simply did not have enough time to support every iota of rule-

6 Federal Register, Vol. 32, No. 23 (February 3, 1967), 2408.

68 making with scientific evidence.7 They had to act faster and more cunningly than complete openness would allow. So there is almost no paper trail for some moments of the standards-building process. Yet, we can reconstruct some it retrospectively.

The Pre-History of Federal Automotive Safety Standards: The State as Consumer and User

In the United States, the federal government came to think about auto safety not only because it was worried about the health of its citizens, but also because it was worried about the well-being of its employees, the nation’s civil servants.

During the late 1950s and early 1960s, federal administrators pushed for increased safety standards in government-owned vehicles. Often the injury and death of government employees hastened and intensified the views of agency leaders. The primary way these internal interests played out was through pressuring the

General Services Administration (GSA) to purchase vehicles with safety features built in. The federal government was an enormous consumer of vehicles, and the

GSA was the agency in primary control of purchasing. Even agencies that managed their own vehicles, like the Postal Service, went through GSA for purchasing and selling. The GSA also maintained a large number of motor pools used by government employees throughout the nation.

7 At the same time, the hidden nature of the NHSB’s standards-making relates deeply to reflections on the relationship between science and democracy by scholars of Science and Technology Studies, such as Yaron Ezrahi, Sheila Jasanoff, and Mark B. Brown. Ezrahi, The Descent of Icarus: Science and the Transformation of Contemporary Democracy (Cambridge, Mass.: Harvard University Press, 1990); Jasanoff, Designs on Nature: Science and Democracy in Europe and the United States (Princeton: Princeton University Press, 2005); Brown, Science in Democracy: Expertise, Institutions, and Representation (Cambridge, Mass.: The MIT Press, 2009).

In October 1960, the Secretary of the Interior complained to Franklin Floete,

Administrator of GSA, that employees of the Department of the Interior believed that seat belts were a necessity in government vehicles.8 The Secretary claimed that a number of injuries and deaths of Interior’s civil servants in auto accidents could have been avoided if someone (automakers, post-market suppliers, or government workers) had installed seat belts. Floete responded, “Seat belts have been carried as a permissible option in Federal Standard 122 for automobiles, station wagons, and light trucks, since June 8, 1958.”9 But, he noted, the

Comptroller General had ruled (Decision B-131418) on April 19, 1957 that seat belts could not be added to vehicles unless their price with the seat belts remained within the “maximum allowable price for the purchase of automobiles established each year in the General Government Matters Appropriation Act.” Since this price was typically set very close to market price of an automobile, there was often little margin for the purchase of safety features, thus putting seat belts out of reach for federally owned vehicles.

In another example, Rufus E Miles, Jr., the Administrative Assistant Secretary of Health, Education, and Welfare, wrote to Floete in 1960, saying that HEW, along with the American Medical Association and the National Safety Council, were pushing for seat belt use by all drivers and passengers.10 Floete responded that the

GSA had mandated that seat belt anchors must be included in all government

8 Letter from Franklin Floete to Sec of Interior, Oct 18, 1960, Record Group 269: 36/1/4, Central File, 1959-1961, Box “Vehicles,” Folder 1. 9 Ibid. 10 Letter from Franklin Floete to Rufus E. Miles, June 22, 1960, Record Group 269: 36/1/4, Central File, 1959-1961, Box “Vehicles,” Folder 1.

70 vehicles. But he reiterated his previously noted statements about prices.

Moreover, he claimed that the GSA had not equipped its vehicles with seat belts because many believed that no one used them, though the agency was using interagency motor pools to conduct “controlled tests” of this hypothesis.

This impasse—the possibility of safety standards blocked by price concerns and the inability to create real, effectual mandates—remained in effect for a number of years. Representatives, administrators, and outside interest groups would write to the GSA and encourage it to require safety features, “anti-smog”

(emissions) controls, and items such as highway warning kits (flares, etc.) in all federal vehicles. The GSA Administrator or one of his employees would write back, saying that the agency lacked authority or that the feature was price prohibitive.

Change came with the passing and signing of the “Roberts Act” of 1964, named for Congressman Kenneth Roberts (D–AL), who had pushed for the law but who had actually been voted out of office by the time the bill had passed.11 This law gave the GSA standard-setting powers for government-purchased vehicles. The

GSA published its first seventeen standards on June 30, 1965 and then added nine new standards and revised eleven of the original ones on March 8, 1966.12 Since the federal government bought such a large number of vehicles, auto manufacturers moved some of these standards into the entire national fleet, so they did not have to retool just for vehicles made for the government.13 Though these market-wide

11 Public Law 88–515 12 Federal Register, Vol. 30 (1965), 8319; Federal Register, Vol. 31 (1966), 4088. 13 On the idea that users often play a role in shaping technology, see Eric Von Hippel, “Economics of Product Development by Users: The Impact of ‘Sticky’ Local Information,” Management Science, Vol. 44, No. 5 (May 1998), 629–644; idem., Democratizing Innovation (Cambridge, Mass.: The MIT Press,

71 effects of GSA’s standard setting may have originally been unintentional, later the

GSA used the mechanism in a self-conscious manner to encourage safety improvements. Moreover, the GSA’s powers did not simply disappear when the

NHSB came into being. The two agencies—GSA and the Department of

Transportation—signed an interagency agreement, stating that GSA could continue to set more stringent standards than the NHSB required.14 Indeed, some seemed to suggest that the GSA could “pull” automakers to innovate in ways that NHSB could not (particularly through government demand).

The First Twenty Auto Safety Standards

On October 8, 1966, the NHSB published an Advance Notice of Proposed Rule

Making in the Federal Register, declaring that the agency would issue initial standards by the statutory deadline of January 31, 1967. The law required that interested parties be allowed “to participate in the making of [the] standards.”15 In the end, interested parties sent 3105 pages of technical commentary on the proposed standards by November 1. A month later, the agency had prepared formal proposals for the initial standards, which in theory reflected the input of the interested parties. The NHSB published the Notice of Proposed Rule Making on

2005); and Joanne Yates, Structuring the Information Age: Life Insurance and Technology in the Twentieth Century (Baltimore: Johns Hopkins University Press, 2005). As I will discuss in the next chapter, there is clear, if somewhat anecdotal, evidence that the GSA had a real effect on the national fleet. Indeed, automakers protested some of the NHSB’s new standards because the companies had already retooled their production lines to meet the GSA standards. 14 “Inter-Agency Agreement,” March 3, 1967, NARA, Record Group 416, Finding Aid UD-UP-3, Box 1, Folder 5, "Organization History", 49-52. 15 United States Department of Commerce, National Traffic Safety Agency, “Report on the Development of the Initial Federal Motor Vehicle Safety Standards Issued January 31, 1967,” March 17, 1967, NHTSA Microfilm, Reel One, ii.

December 3. The issuance allowed interested party submissions on these proposed standards until January 3, and automakers and other parties submitted 4,525 pages on these proposals, an increase in heft of almost fifty percent from round one commentary. The agency then had “slightly more than three weeks” to sort through these documents and publish the initial standards by January 31. Things were moving very quickly.

Beyond the requirement that the new standards had to be based on previously existing ones, another stricture of the law had a profound effect on the eventual shape of the standards. Standards had to be based on performance, not design. That is, they had to set criteria for at what level the technology was to perform, not which technologies should be used. The NHSB originally proposed twenty-three standards, but three were found to be impracticable, so only twenty were finally issued. As the NHSB wrote when the agency promulgated the standard:

“These twenty standards . . . were based on a total of forty-eight existing standards. These include nineteen from the General Services Administration (GSA) [which were issued as part of the Robert’s Law], one from the Interstate Commerce Commission (ICC), twenty-two voluntary practices recommended by the Society of Automotive Engineers (SAE), one Swedish National Road Board Standard, one of the Uniform Vehicle Code, one of the Post Office Department, one of the U.S.A. Standards Institute, one from State laws, and one from the National Bureau of Standards (NBS).”16

16 Ibid., iii.

Thus, in the agency’s staff’s reading, the law forced them to cobble together

standards, basing some individual standards on aspects drawn from multiple

“previously existing” standards.17

The law hemmed in the agency in other ways as well. The GSA standards,

which formed the core of the proposed standards, were meant for government

vehicles and were not entirely practicable for the national fleet. But standards that

currently applied to civilian vehicles were voluntary and lacked teeth. The NHSB’s

goal, therefore, was to adapt the stringency of the government standards while

culling the broadness and flexibility of the civilian ones. Yet, the GSA standards

ultimately did the lion’s share of the work. As the NHSB report noted, “In this

context, it should be pointed out that the Agency was able to accomplish as much as

it did substantially because of the fact that the industry worldwide had been served

notice by the earlier GSA standards, largely adopted by the Agency, as to the likely

direction of future Government vehicle standards.”18

Additionally, the science of crash safety continued to change very quickly.

Two of the standards were revised between the initial and final rules because

17 In the 1980s and 1990s, academics made a big deal and much use of the anthropologist Claude Levi-Strauss’s notion of the “Bricoleur,” a kind of handyman or tinkerer who uses preexisting, available materials that lay at hand. Scholars have paid less attention to the fact that Levi-Strauss opposed the Bricoleur to an ideal type that he called the “Engineer,” a person who, in Levi-Strauss’s view, creates whatever new tools are necessary to complete a task. Yet, while it is true that engineers do build new tools and instrumentation, at least in terms of standards-making, Engineers too are Bricoleurs. The engineers, technicians, and scientists at the National Highway Safety Bureau cobbled together standards and the tests that comprised them from a host of sources. Among other sources, they looked to existing engineering methods formalized in the Society of Automotive Engineering’s Recommended Practices. Claude Levi-Strauss, The Savage Mind (Chicago: University of Chicago Press, 1966). Historians of technology have long drawn attention to how “innovators” draw on previous sources. See Usher, A History of Mechanical Invention; Louis C. Hunter, Steamboats on the Western Rivers: An Economic and Technological History (Cambridge: Harvard University Press, 1949); Kubler, The Shape of Time; Basalla, The Evolution of Technology; Angela Lakwete, Inventing the Cotton Gin: Machine and Myth in Antebellum America (Baltimore: Johns Hopkins University Press, 2003). 18 “Report on the Development of the Initial Federal Motor Vehicle Safety Standards,” iii.

74 significant studies changed the rational basis of the earlier proposed standards.

The clearest example of this was the segmented, “energy absorbing” steering column, discussed in the next chapter.19 Moreover, by the time of the statutory rule- making, the agency still did not have sufficient evidence to make a number of standards that it wanted to put into effect, including rules for tires, head restraints

(almost always head rests, but automakers technically could have used other technologies to meet the test), and windshield wipers for small vehicles. So the

NHSB was forced to delay these standards.

In its “Report on the Development of the Initial Federal Motor Vehicle Safety

Standards Issued January 31, 1967,” the NHSB detailed how submitted comments were incorporated into the final rules by illustrating the development of Standard

103, “Windshield Defrosting and Defogging.” NHSB staff members derived the federal standard from an SAE-recommended practice, which involved “a highly specialized climatic test chamber.”20 The NHSB quickly realized, however, that foreign manufacturers did not possess such test chambers, which were employed only by SAE-allied manufacturers in the United States. Moreover, regulatory agencies have almost always had trouble making rules that are based on new testing instruments. Not only were the instruments difficult to obtain, but also the new market for the devices often created large backorders, sometimes even for years. With this information in hand, the NHSB took two measures: first, it promulgated the simple rule that all cars must have some form of defrosting and defogging. (This was not a design standard because it did not specify how the

19 As detailed on Ibid., v. 20 Ibid., vi.

75 manufacturers would meet this requirement.) Second, it simplified the test procedures so that manufacturers could use more readily-available and widely- adopted cold storage facilities. The agency, thereby, avoided the pitfalls of the specialized test chamber while maintaining the spirit of the rule. The NHSB ended its report by saying that it planned on giving the industry more lead time in future rule-making. The Traffic Safety Act’s statutes had simply limited the agency’s ability to do so in this first instance.

The NHSB’s staff members, many of whom were new to government service, continued to overestimate the industry’s willingness to adopt safety standards voluntarily, or at least they did so rhetorically, perhaps in an effort to spur the industry to make this shift. The report claimed, “There is every evidence that manufacturers will soon be competing actively in the safety field on a voluntary basis, and will on their own introduce numerous safety features as a stimulus to sales.”21 This prediction eventually came to fruition, but not for another twenty years. Members of the NHSB apparently had no concept of the fight that awaited them or of how long they would have to promote safety before consumers themselves began to demand it.

Building Standard 201

William Haddon, Robert Brenner, and the crash safety community cared deeply about Standard 201—Interior Occupant Protection—because it was meant to combat the problem of the “second collision,” described in Chapter 1. As the New

21 Ibid., iv.

York Times wrote, ”The rule is considered by many the most important of the twenty. Its purpose is to minimize injuries resulting from the ‘second collision’ that occurs when the occupant of a car is thrown forward by the impact of a crash.”22 Of all the standards, only Standard 201 fully embodied the idea of the “second collision,” and to the degree that the auto industry rejected this notion, the standard was a philosophical battleground. Moreover, Standard 201, both literally and symbolically, penetrated the most private relationship between the automakers and consumers, the vehicle’s interior, which was the basis of many visions of luxury, comfort, and contentment.23

Standard 201 was based on six GSA standards, all related to the vehicle’s interior.24 The GSA standards that dealt with impact areas stemmed from two SAE

Recommended Practices—J826, which used various devices to figure out seating accommodations, and J921, which used an “impact pendulum” test to determine the

“safety characteristics of padded instrument panels and other passenger compartment components.”25 Engineers used J826 to determine where head impact zones were in the vehicle, and they then used the pendulum device described in J291 to smash an accelerometer-equipped head form into those impact

22 “Car Safety Rule Termed Illegal,” New York Times, Mar. 7, 1967. 23 In the early 1960s, legislatures and courts were still working out the degree to which a car could be considered a private space that would, for instance, enjoy 4th Amendment protections. In 1960, a federal court had decided for the first time that cars involved in drug busts could be seized. Cars were personal technologies that used public infrastructure and were highly visible, and their claims to “privacy” were unsettled. 24 Ibid., 58, 110. Standard 201 was based on GSA standards 515/2a—Padded Instrument Panel and Visors for Automotive Vehicles, 515/3a—Recessed Instrument Panel Instruments and Control Devices for Automotive Vehicles, 515/18—Window and Door Controls, 515/19—Ashtrays and Lighters, 515/20—Armrests, and 515/21—Padding for Automotive Seat Backs. The GSA published its finalized standards in the Federal Register on June 30, 1965, pages 8319–8327. 25 The tests remain largely unchanged today. See, for instance, the SAE website for the standards.

77 zones. The J291 pendulum test was often referred to as the “lollipop test” because observers thought that the head form looked like a lollipop when it was placed on the end of the shaft that was attached to the pendulum.

The Chrysler Corporation played an influential part in constructing both of these SAE testing processes, which is surprising given how small of a role Chrysler played in developing emission controls in the 1970s, when the firm almost completely relied on GM’s R&D. The SAE had tried to standardize seating arrangement practices for a number of years, including an attempt by the SAE Body

Activity Committee in 1959.26 Until the SAE established Recommended Practice

J826 in 1963, however, Chrysler, Fisher Body (GM’s auto body division), and Ford

Motor Company all had different methods for determining internal configurations of their vehicles’ body. 1n 1961, the SAE Body Activity Committee made another push for standardization and had created a Manikin Subcommittee. The core

“innovation” in “interior dimensioning” came out of GM, which took the best features of each company’s practices and combined them into a single manikin and methodology, which became J826.27

Recommended Practice J921 had a similarly long gestation. Chrysler’s

Rubber and Plastics Laboratory developed the precursor of the pendulum used in

26 R. C. Bazzell, “The SAE Manikin Interior Space Evaluator,” a paper given at the National Automobile Meeting, Detroit, MI, March 19-21, 1963, SAE Paper 630176, 1. Bazzell was an engineer in Chrysler’s Engineering Division. The difference between Chrysler’s role in auto safety versus emissions control was partly related to their relative relationship to the core of the firm: Chrysler designed its own bodies and, thus, had vested interests in playing some part in shaping the industry standard, but the firm had much fewer capabilities around chemistry and chemical processes and often relied on suppliers, including GM in the case of emissions control, for developments in these fields. 27 Vincent D. Kaptur and Michael C. Myal, “The General Motors Comfort Dimensioning System,” paper presented at SAE International Congress of Automotive Engineering, Detroit, January 1961, SAE Paper 610174.

J921 in 1956-1957.28 When the SAE Body Engineering Committee formed the

Passenger Compartment Energy Absorption Subcommittee in response to the passage of the Roberts Act, Chrysler’s Engineering Office redesigned the pendulum.

GSA picked up J921 as part of GSA standard 515/2, which suggests that industry played an influential part in shaping the federal standards.

The NHSB admitted from the beginning of the project that it had tightened certain aspects of the GSA standards to create a more stringent interior occupant standard. Standard 201 added a “knee and leg” impact requirement to the already existing “head” and “hip” impact criteria. It lowered the acceptable g-forces that the crash dummy could sustain, and it “changed the wording of the requirements for protrusions [items such as door handles, radio knobs, and window cranks that projected into the passenger compartment].”

NHSB employees stipulated that the test should be run for both a 95th- percentile adult male (i.e., a manikin of which only five percent of adult males were larger) and a 5th-percentile adult female (i.e., a manikin of which only five percent of adult females would be smaller). Since men tend to be larger than women, if a car could safely accommodate both a large male manikin and a small female one in a crash, NHSB employees believed it would be safe for the wide range of human figures, making exceptions for people who had forms of either dwarfism or

28 A. M. Hansen,”SAE Test Procedure for Instrument Panels: An Impact Pendulum for Testing Safety Characteristics of Padded Instrument Panels and Other Passenger Compartment Components,” October 10, 1965, SAE Paper 650963, 1. Hansen worked as the Supervisor of Material Development in Chrysler’s Engineering Office. Of course, materials testing has a much longer history than the one described here. See for instance, Stephen Timoshenko, A History of the Strength of Materials (New York: McGraw-Hill, 1953) and Amy E. Slaton, Reinforced Concrete and the Modernization of American Building, 1900-1930 (Baltimore: Johns Hopkins University Press, 2001).

gigantism.29 The measurements of these manikins were based on anthropometry

research published by the Public Health Service.30

Fighting Standard 201

In part, the controversy between regulators and automakers over Standard 201

arose from different visions of engineering practice. The automakers claimed that

NHSB employees had an insufficient understanding of existent industry practices,

while the regulators argued that the auto companies were too hidebound.

More importantly, however, the contest over Standard 201 was a fight about

the philosophy of crash worthiness and the auto safety advocates’ adoption of the

theory of the “second collision.” As discussed in the last chapter, Hugh de Haven,

the “Father of Crashworthiness,” used the idea of the “second collision” to argue

that humans were injured when their bodies impacted the interior of their vehicles,

not from the first impact between the car and some other object. Through Standard

201, the NHSB sought to rectify the second collision by making the vehicle’s interior

29 The NHSB never spelled out the intentions for the larger and smaller manikins in its publications. But, if the intentions were not plain enough, at a hearing, one of the NHSB’s attorneys, Robert O’Mahoney, who is discussed below, asked one of the auto industry’s engineers, “Let me ask you, sir: Would it be your understanding of the definition which uses as one outer limit the 95th percentile male and as another limit the fifth percentile female, that it was intended to protect the segment of the population that has sizes between the 95th percentile male and 5th percentile female?” To this question, Edward C. Adkins, a representative for General Motors replied, “I object to this line of questioning. This came up several times. The intent of the standard is not in issue here. I think that such questions are irrelevant and particularly so since this witness is asked to give an opinion as to what he might conceive to be the intent of the standard.” (“Transcript of Proceedings in the Matter of Reconsideration of Initial Motor Vehicle Safety Standard 201,” May 22, 1967, Warren, Michigan, pg. 91, NHTSA Microfilm, Reel 2) 30 National Center for Health Statistics, Weight, Height, and Selected Body Dimensions of Adults, United States, 1960–1962 (Washington: U.S. Dept. of Health, Education and Welfare, Public Health Service, 1965). As I will discuss in Chapter 4, the federal automotive emissions control efforts were originally based in the Public Health Service. Thus, at its inception, automotive regulation was strongly intertwined with “public health” efforts. But perhaps more striking is how auto safety regulation has become disentangled from federal agencies focused on public health.

80 much safer. Though the automakers had essentially lost to the crash safety advocates in the public eye and in the legislature, the companies could still fight through the rule-making process; Standard 201 exposed the most severe moral, technical, and ideological differences between the two sides.

For economy’s sake, I have combined the various parties’ comments from the different stages—petition, written direct testimony, hearings, and rebuttal— into a summary of their positions, noting where the comments came from when it is relevant.

Organizing the Rule-Making Process

NHSB published its original slate of standards on February 3, 1967, and interested parties had until March 6 to file petitions for reconsideration of the standards. Twenty-seven automakers and auto associations from around the world filed petitions challenging the validity of Standard 201 and asking for its reconsideration. On March 29, 1967, Lowell K. Bridwell, who was then the Acting

Undersecretary for Transportation at the Department of Commerce but who would become Federal Highway Administrator later that year, released an order that consolidated these petitions so that they could be dealt with together instead of one by one.31

Because of the strong opposition to Standard 201, Commerce decided to create a “rulemaking proceeding pursuant to U.S.C. 553 (formerly section 4 of the

31 Lowell K. Bridwell, Order of Consolidation, March 29, 1967, NHTSA Microfilm, Reel 1.

Administrative Procedure Act).”32 On April 20, 1967, Bridwell wrote a letter to B. R.

Allen, the Director of the National Transportation Safety Board’s Bureau of Aviation

Safety, requesting that Russell A. Potter, a staff member of the aviation safety agency, be temporarily assigned to the NHSB as presiding officer responsible for the proceedings.33 Little biographical information is known about Potter. He worked at the National Transportation Safety Board’s predecessor agency, the

Civilian Aeronautics Board, since at least 1947 and probably before then because he chaired a hearing in that year, not a task ordinarily left to agency newcomers.34

Potter’s life was cut short in December 1968 when he and two others were lost at sea during a sailing trip to the Virgin Islands.35 It is perhaps for this reason that

Potter did not eventually rise to a high enough level in the bureaucracy to be more visible in the historical and biographical record. Regardless, by 1967, when Potter presided over the hearing on Standard 201, he had accrued at least twenty years of experience regulating complex technological risks—in his case, airline safety.

Each petitioner had until May 1 to file a more detailed explanation of the

“portions of Standard No. 201 to which he objects, the reasons for the position or objections, a statement of proposed issues, the kind of evidence he wishes to tender in support of the position or objection and a statement of what portion, if any, of the evidence he considers it essential to submit orally.” 36 Three days after the deadline, the presiding officer would hold a “prehearing conference” in which he

32 Lowell K. Bridwell, Order in the Matter of the Petitions for Reconsideration of Initial Motor Vehicle Safety Standard No. 201, April 21, 1967, NHTSA Microfilm, Reel 1, 2. 33 Lowell K. Bridwell to B. R. Allen, April 20, 1967, NHTSA Microfilm, Reel 1. 34 “’Breakdown’ Cited in Air-Sea Rescue,” New York Times, October 24, 1947. 35 “3 on 33-Foot Sloop Lost on Trip to Virgin Islands,” New York Times, December 28, 1968. 36 Lowell K. Bridwell to B. R. Allen, April 20, 1967, NHTSA Microfilm, Reel 1.

82 would lump and simplify objections and “seek methods to prevent duplication of efforts by the parties.” The agency would provide anyone interested “copies of written direct testimony” submitted by the parties by May 12, and, one week later, it would furnish “written rebuttal testimony” prepared by agency employees. Then, a hearing allowing for cross-examination of both “direct and rebuttal” testimony would be held on May 23 in Washington, D.C. The parties could, within ten days, file further briefs on what was said at the hearing. Finally, the presiding officer would issue his recommendations as quickly as possible, and interested parties then had ten days to submit further objections. The rulemaking proceeding was a reasoned process that included extensive scientific and technical information, but it moved at break-neck speed. Scientists and engineers, especially those within the agency, were often expected to turn around analyses of complex technical issues within a week. Members of the NHSB tripped up more than once during the process.

The prehearing conference on May 4 bordered on chaos. The agency had received hundreds of pages of submissions and had three days to look them all over. Not only was the standard inherently complex, but the submissions also brought up a plethora of issues. The first objective of the hearing, therefore, was to settle what the issues were. The participants and the hearing officer handled this in an interesting way: they agreed that the submission of the Automobile

Manufacturers Associations (AMA) summarized the major points and would form the basis of future discussions. As Potter wrote, “[The AMA] statement was adopted as the issues in the case, for the 28 petitioners agreed that it accurately

83 sets forth all questions raised by them.”37 This decision strengthened the AMA’s position within the rule-making process. The petitioners had adopted the AMA’s submissions as the basis of their claims, and the hearing officer had chosen the trade association’s materials as a rubric through which he would organize and interpret the proceedings. Thus, the AMA and its products took a central, intermediary role.

The Agency’s Position

Throughout the process, the NHSB held that Standard 201 should remain intact for three reasons.38 It would reduce deaths and injuries. It was “based on existing standards and relevant available motor vehicle safety data,” as was required by the federal law. And it was “reasonable, practical, and appropriate.”

Cast in the language of the Traffic Safety Act, these three words became flash points of disagreement.39

The NHSB had two non-agency auto safety experts, Derwyn M. Severy and

Colonel John Paul Stapp, testify for the agency’s position. Severy had been a colleague of Brenner’s at UCLA and continued to work there at the time of the hearings. Brenner probably recruited Severy to the agency’s efforts, though in

37 Russell Potter, Federal Highway Administration, “Reconsideration of Initial Motor Vehicle Safety Standards: Docket FHA–1: Recommended Findings (Report of Presiding Officer),” June 23, 1967, NHTSA Microfilm, Reel 2, 22. Hereafter, “Report of Presiding Officer.” In all likelihood, Potter co- wrote this document with a number of other people, but since no record exists of these other contributors, I will refer to him as the author. 38 Robert M. O’Mahoney and David Schmeltzer, Statement of Position, In the Matter of Petitions for Reconsideration of Initial Motor Vehicle Safety Standard No. 201, April 29, 1967, NHTSA Microfilm, Reel 1, 1. 39 In chapter 5, I describe how similar disagreement arose over the terms “good faith” and “feasible” in the law that regulated automotive emissions.

84 some ways, Severy’s participation in federal auto safety efforts preceded Brenner’s.

Severy had been a member of a panel that had helped construct the GSA standards, including those upon which Standard 201 was partly based.

Colonel Stapp, on the other hand, was one of the great names in crash safety.

In another example of knowledge “spilling over” from military research to the civilian sector, Stapp’s early research was motivated by ensuring the safety of Air

Force pilots.40 In one of his experiments in which he used himself as a test subject,

Stapp reached 632 miles per hour before coming to a complete stop in less than one second. During this short time stopping interval, he reached 43 Gs, the same force that one would attain if one were to collide with an immovable wall at 70 miles per hour. Stapp reported that his eyeballs nearly popped out in the process. They filled with blood, and for ten minutes, he could not see anything.

But, against these two agency expert witnesses, GM alone submitted written direct testimony from thirteen employees. Additionally, the company submitted a letter from Lawrence M. Patrick, a professor of engineering at Wayne State

University, which is discussed below. The automakers enlisted and demonstrated overwhelming expertise. Although the highway safety movement had been able to persuade a majority in Congress to sign on to its cause, in terms of dealing with detailed technical analysis and argument, it was seriously outgunned.

40 Aeronautical safety played a profoundly important role in the early days of crash safety. As described in the previous chapter, Hugh de Haven was partly motivated to begin studying crash safety after he survived a plane crash. Moreover, the Cornell Aeronautical Laboratory (which became the Calspan Corporation in 1972) was one of the most significant early centers of crash science.

Clear and Objective

The requirement that standards be “clear and objective” primarily involved how the regulations were worded and defined. For several reasons, many of the complaints about Standard 201 related to this requirement: first, the regulation probably did not live up to the exactitude demanded by the auto industry. Second, since Standard 201was cobbled together from several standards, it had added complexities and vague aspects not contained in any of its constitutive regulations.

Third, the standard tried to implement new innovations that were not yet standard in the industry and, thus, had not yet received all the refinements standards typically received in such circles as the SAE’s committees. More particularly, the new parts in the standards had not gained the “enrollment” by engineers and firms that was (and is) typically of the SAE standards-setting process. Finally and most importantly, to the degree that the auto industry’s lawyers’ endeavors were obstructive, obscurant, and sophistical, the standard’s language made a perfect battleground.

In his “Recommended Findings” report, discussed in greater detail below,

Potter pointed to two General Motors employees, Edwin Klove and George Ryder, as being very persuasive about the standard’s lack of clarity and objectivity as well as about other points. George Ryder earned his Bachelors in Engineering at

Michigan State University in 1949, and in 1966, he became the Assistant Chief

Engineer of GM’s Buick Motor Division, the position he held during the 1967

Standard 201 hearings.41 As Assistant Chief Engineer, Ryder’s duties included

“complete responsibility for translation of advanced body designs into a usable production design.”42 Much of his career had been spent moving between models and renderings, producing objects in the automotive and aircraft industries, including the “translation of an aircraft plywood design into sheet metal” and multiple projects on car bodies and “automotive glass development.” Later, Ryder formed an automotive consulting firm in 1981. Less is known about Klove, but he received a number of patents as co-inventor of interior features.43

Most of the criticisms regarding Standard 201’s lack of clarity and objectivity centered on the definition of several terms, including the H-Point (which represented both the “hip” on an adult human and the “hinge” on a manikin or

“crash dummy”) and the various “impact areas”—head, pelvic, and knee and leg.

The automakers further contested the standard’s requirement for the 95th percentile male and 5th percentile female crash manikins. Many of these requirements were part of the GSA’s 1966 revisions of its first automotive procurement standards. But automakers may have felt that they had less recourse to complain about the GSA standards than the nationally-binding NHSB standards,

41 “Direct Testimony of George R. Ryder” in “Direct Written Testimony of General Motors Corporation,” May 12, 1967, NHTSA Microfilm, Reel 1, 1. 42 Ibid. 43 Edwin H. Klove, Jr., Joseph J. Magyar, Marvin A. Packett, Detroit, and Robert E. Meshew, Birmingham, MI, assignors to General Motors Corporation (GMC), “Safety Belt Buckle,” US Patent 3,256,576, filed April 29, 1964, and issued June 21, 1986; Edwin H. Klove, Jr., Warren, and James L. Noll, Livonia, MI, assignors to GMC, “Closure Latch,” US Patent 3,423,117, filed August 17, 1967 and issued January 21, 1969; Edwin H. Klove, Jr., Bloomfield Hills, and James L. Noll, Livonia, MI, assignors to GMC, “Inflatable Occupant Restraint,” US Patent 3,618,977, filed May 11, 1970 and issued November 9, 1971.

87 since the GSA standards only explained what was required in cars it would buy and did not apply to the entire production runs of the auto manufacturers.

Yet, debates about seemingly superficial issues, such as how something was defined, often masked deeper disagreements about the standard’s very justice. For instance, should the standards require the automakers to innovate, not just in terms of the automotive safety technologies but also in terms of the tests used to determine compliance? Grappling over the H-Point was one such case in point.

Industry representatives consistently maintained that that H-Point and the impact areas should be defined in terms of two-dimensional, graphical methods—what was sometimes called “graphic equivalency”—for designing automotive interiors, not in terms of a three-dimensional, “real-life” crash test. As General Motors wrote in its “Statement of Position, ”[the graphic equivalent] is necessary to permit designing compliance into the car from the very beginning of the design process— the drawing board. The use of a graphic equivalent is also necessary to provide a precise and repeatable method for continually checking design at all phases in the automotive design and production cycles.”44 General Motors, like many of the other automakers, used the two-dimensional manikin prescribed in SAE Recommended

Practice J826. Designers first worked out issues, such as spaciousness and the control elements—door handles, window cranks, stereo knobs—being within reach, in two-dimensions. The automakers’ real fear was that the standard would require them to build up every potential vehicle into a working model, only to find out that the car would not meet the space requirements. Such a discovery would

44 “Statement of Position of General Motors Corporation,” NHTSA Microfilm, Reel 1, 3.

88 potentially add huge lags into the car’s lead-time and would create an expensive pile of wrecked cars. Thus, engineers and designers from the automakers wanted two-dimensional methods to solve these issues beforehand. At the same time, automakers may have played up the difference between actual industry practices and the new standards in order to show that regulators were clueless and that their mandates were absurd.

Doubtlessly, some of the difficulties fostered by creating a standard based on crash tests instead of one based on two-dimensional graphical methods was that the scientific and technical network at NHSB was based, in part, on crash tests.

NHSB, after all, composed a large segment of the crash safety community. It crashed cars; it did not design them. This lack of knowledge about the ins and outs of the auto industry would continue to plague federal regulatory efforts. Still, members of the NHSB may have shot back that what they cared about was the real- world object. Two-dimensional design and analysis were valueless unless they translated into a safer car, and the only established way to determine the car’s safety was to run it into a cement wall.

For the most part, industry representatives would not just come out and say that they wanted a two-dimensional standard, however. Most of their rhetorical work was accomplished through attacking the sense of the H-Point and the other terms. For instance, they questioned whether the H-Point was knowable. If a manikin was placed on the driver’s seat, its weight would depress the seat cushion.

The automakers argued that slight variations in the mankin’s and the seat’s

89 tolerances would mean that the H-Point would always be at a slightly different place, rendering the tests unrepeatable.

Practicality

The central question over whether Standard 201 was “practicable” was inextricably tied to whether it set design criteria rather than performance criteria.

From the get-go, the NHSB stated that it would set performance standards while maintaining that some such standards would inevitably affect design in determinable ways. A report on how the agency developed the standards approvingly quoted something Senator Warren Magnuson had said:

“We are also pleased that the House agreed to the restoration of Senate language for the definition of ‘motor vehicle safety,’ recognizing that safety is related to design. Performance standards issued under the act are expected to affect the design of such features, for example, as steering assemblies, instrument panels, seat structures, windshields, seat belts, brakes, and door latch and frame components—all of which will particularly affect the design of these components.”45

As the NHSB report summarized this idea, “There are no precise legal or other criteria for deciding when a performance standard begins to have design overtones.” Therefore, what constituted a design standard and what constituted a performance standard was a matter of interpretation.

The main requirements that the automakers complained about in terms of practicability were the parts that dealt with “protrusions,” “bezels,” and other

45 United States Department of Commerce, National Traffic Safety Agency, “Report on the Development of the Initial Federal Motor Vehicle Safety Standards Issued January 31, 1967,” March 17, 1967, NHTSA Microfilm, Reel On, 6. Original quote in Congressional Record, August 31, 1966, 20600.

90 interior objects, including window cranks, door handles, and support beams. The automakers sent in a number of blue prints and photographs demonstrating the absurd effects Standard 201 would have on interior auto design. For instance, some submitted photos of redesigned window cranks showing that Standard 201 would make the head of the crank so rounded and oblique that it would be almost impossible to grasp. Similarly, they argued that Standard 201’s requirement that these aspects of the vehicle’s interior not be able to withstand more than 90 lbs. of pressure (so that occupants would not be impaled or otherwise injured on these features) was ridiculous because people often used things, such as window cranks, to close vehicle doors. Thus, if someone did this and Standard 201 were in effect, the window crank might snap off in the person’s hand. All of the written submissions, including Ryder’s and Klove’s, mentioned the ridiculous nature of these requirements.

The NHSB’s witnesses pushed back against these criticisms, however.

Derwyn Severy, wrote of Ryder’s assertion that “‘In my [i.e. Ryder’s] opinion, design requirements for protrusions are improper,’ suggests a completely callous attitude toward the injury producing qualities common to protuberances.”46 Stapp claimed that some of the auto industry’s representatives had gone to “grotesque intellectual effort to belabor the definition of protrusion” in order to make it seem incoherent, while “not one engineer suggested a better term.”47 That is, industry witnesses were not trying to be helpful in establishing a better standard; they were trying to destroy the proposed one.

46 “Verified Statement of Derwyn M. Severy,” NHTSA Microfilm, Reel Two, 27. 47 “Affidavit of Colonel John P. Stapp USAP MC,” NHTSA Microfilm, Reel Two, 5–7.

Safety

General Motors and the AMA recruited Lawrence M. Patrick to testify on the proposed Standard 201’s human tolerance requirements—that test manikins not exceed 80 g’s for 1 millisecond during testing. Patrick was a Michigan boy, born in

Highland Park (home to Henry Ford’s famous factory) in 1920. He did all of his secondary schooling at Wayne State, earning a BS in mechanical engineering in

1942 and a BS in aeronautical engineering in 1943 before going on to receive a

Masters in mechanical engineering in 1955.48 When he became a professor at

Wayne State, his work focused on human tolerance studies. General Motors supported much of Patrick’s research in a “joint venture” between Wayne State and the company, though Patrick did not reveal this relationship in any of his testimony.49 In 1974, Patrick won the American Association of Automotive

Medicine’s (AAAM’s) Award of Merit, the association’s award for life-time achievement in the field, which Col. John Stapp, Hugh DeHaven, and William

Haddon also received over the years.50

Patrick’s testimony, along with that of professors from other Michigan universities, such as Donald F. Huelke, Professor of Anatomy at the University of

Michigan, raises the question of networks of knowledge. For obvious reasons, much of the knowledge on automobile design and other prerequisite fields, such as

48 Patrick’s curriculum vitae was submitted along with his testimony and is contained in NHTSA Microfilm, Reel 1. Additional biographical information can be culled from Mary Roach, “I Was a Human Crash-Test Dummy,” Salon.com, November 19, 1999, http://www.salon.com/health/col/roac/1999/11/19/crash_test/print.html (Last accessed April 18, 2011). 49 As Potter wrote, “John P. Danforth . . .testified that the Wayne State research was a cooperative venture with General Motors.” “Report of Presiding Officer,” 32-3. 50 The information about the AAAM’s awards is taken from its webpage: www. Carcrash.org.

Huelke’s studies of human crash tolerances, were centered in and around Detroit.

Indeed, Huelke’s safety research was so well-known and renowned enough that

Ralph Nader had cited him to prove his own (Nader’s) case during the testimony that led to the Traffic Safety Act. Huelke also received the AAAM Award of Merit as well as, in 1980, the association’s first-ever A.J. Mirkin Service Award, which recognized sustained service to the organization. He served as the editor for the

AAAM Quarterly and the association’s Proceedings of Annual Meetings for a number of years. Huelke’s and Patrick’s research was deeply connected. In fact, in an a brief article on human tolerance research being conducted in the Detroit area, Huelke wrote about both Wayne State’s and the University of Michigan’s laboratories as

“facilities available.”51 More to the point, the same volume contained an essay co- written by Patrick and University of Michigan Professor F. Gaynor Evans.52

Most of the other human crash studies studies done in the United States came out of three fields—aeronautical safety (Hugh DeHaven, the Cornell

Aeronautics Lab) and transportation studies (UCLA) as studies conducted within the military (Col. John Stapp). Although UCLA, Cornell, and other safety centers did crash studies, they did far fewer than the various centers in Detroit, including those of the automakers themselves. Moreover, the automakers sponsored much of the research done at the Michigan universities. Together, the automakers and the

Michigan academics formed a phalanx. A previously existing network tightened in

51 Donald F. Huelke, “Biomechanical Studies on the Bones of the Face,” Impact Acceleration Stress, National Academy of Sciences, Space Science Board, Man in Space Committee, ed. (Washington, D.C.: National Academy of Sciences, 1962), 131-133 52 Patrick and Evans, “Impact Damage to Internal Organs” in Impact Acceleration Stress, 159-171. Another Wayne State Professor, E. S. Gurdjian, also had an essay in the volume, “Experiences in Head Injury and Skeletal Research,” 145-157.

93 the face of adversity. The Detroit knowledge-network was stronger in every way than the one the NHSB mustered. It had more people, more money, more technical resources (such as labs, proving grounds, and access to cars for crash studies), and more political clout. As we will see, Russell Potter, the hearing’s presiding officer, found the Detroit network to be extremely convincing.

Wayne State’s Patrick had additional credibility and popularity in traffic safety circles. Like Colonel John Stapp, he acted as a human crash dummy in his own experiments. Of course, self-experimentation had a long history in science and medicine. In this way, his actions were a bit of scientific orthodoxy: Patrick believed that, if he was going to ask experimental subjects to undergo the crash tests, he should too. On the other hand, several extra-rational factors also played roles in making his actions as a research subject important. Chiefly, the masculinity and edginess inherent in these acts lent Patrick an air of authenticity and credibility.

At the time of Patrick’s testimony, the major upshot of his research was something known as the “Wayne curve,” a graph of tolerance of the human head to deceleration over time.53 The statistics underlying the curve came from merging two sets of crash data: cadaver tests and animal tests. Since most of the cadavers were elderly, their heads had a lower tolerance to forces than the “average” human head; since animals had harder skulls than human cadaver heads, their use was

53 Saying “skull” here would be incorrect since, though skull fracture was a real danger, the more likely and imminent danger was brain injury. Thus, I will continue to refer to the “human head,” which is a more inclusive term.

94 thought to balance the equation. Still, because of cadavers’ frailty, most researchers seemed to believe that the Wayne curve was a conservative estimate.54

In his Recommended Findings report, Potter rhetorically combined Patrick’s testimony with that of John P. Danforth, Senior Research Engineer at the General

Motors Research Laboratory’s Electro-Mechanics lab, and Verne L. Roberts, a PhD in Biomechanics Research at University of Michigan, to argue that there was some consensus about the reliability of the “Wayne curve.” It is difficult to gauge from the available historical sources whether the “consensus” about the Wayne curve was real or simply a product of the tactics of the AMA and the Detroit automakers.

Patrick’s article on the topic was not highly cited in the safety literature, but it had been published less than two years before the hearing. (The paper had been presented at the 7th Stapp Car Crash Conference in 1962, but the proceedings were not published until 1965.) The article laying out the Wayne Curve may have been widely discussed informally in crash safety circles, but we have no sure way to verify this. A likely explanation is that the apparent consensus arose only out of the

AMA’s efforts to organize a response to the NHSB’s proposed standards. If the scientists recruited to the AMA’s cause were to march in formation, it was easiest to give them some “boundary object” to coalesce around.55

The Wayne curve was helpful for three reasons. First, it presented a clear, easily grasped representation of the issue, which aided recruitment of non-experts, including Potter. Second, it suggested that the standard—80 g’s for no more than

54 See, for example, Roberts’ statement in Potter, 33. 55 Susan Leigh Star and James R. Griesemer, "Institutional Ecology, 'Translations' and Boundary Objects: Amateurs and Professionals in Berkeley's Museum of Vertebrate Zoology, 1907–39," Social Studies of Science, vol. 19, no. 4 (1989), 387–420.

1.0 millisecond—should be lengthened since the graph demonstrated that the human head could sustain 80 g’s for longer intervals. Third, it indicated that

Standard 201 should allow for brief spikes in pressure. As Patrick summarized, the curve showed that up to 200 g’s could be sustained for up to a millisecond, while up to 150 g’s could be sustained for 2.0 milliseconds. Again, these thresholds were thought to be conservative. The NHSB argued that brief peaks in pressure arose from, in Patrick’s words, “an abnormal condition that may be removed from an instrument panel or seat back by relocating structural members or equipment.”56 In other words, bad design caused peaks. But Patrick rejected this notion. A peak could arise from “discontinuity in the panel characteristics,” or it could “simply indicate the initial inertia[l] reaction of the impacted material to the blow of the head form.”57 Even foam rubber, upon first impact, had some inertial resistance that would cause an infinitesimal peak. Thus, human tolerance studies lay at the extremes of understanding in several fields. The caprice of human biology was not the only problem; material science was also stretched at this frontier. Moreover, many of the scientific questions simply lay beyond the means of the instrumentation of the day and of human-based experiments.

The development of the Wayne curve by the NHSB’s critics caught the agency flat-footed, which probably says something about the lack of preparation and organization on its part. The agency’s two witnesses, Stapp and Severy, sided with the industry over the agency itself on the low g-force standard. Severy

56 AMA Rebuttal Testimony, “Rebuttal Statement of Lawrence M. Patrick,” May 18, 1967, NHTSA Microfilm, Reel 2, 2. 57 Ibid.

96 suggested that research at UCLA supported Patrick’s three-millisecond criteria as a

“reasonable interim compromise,” though he did insist that the one millisecond standard should be applied within a “2-year period.”58 Stapp argued that “the 1.0 millisecond value . . . was reasonable [sic] chosen by the Agency on the basis of the existing scientific evidence and especially in the interests of public safety.”59 He also gave a brief history of human tolerance studies since 1952, however, and claimed that no one, including himself, had yet established with any certainty or scientific rigor how much force the human head could sustain. Ultimately, the agency bent the knee, feeling forced to admit that the limit should be loosened to 80 g’s for 3.0 milliseconds, instead of 1.0 millisecond, for the time being.

Timing

The automakers protested that they could not meet Standard 201 within the allotted time. The matter of production lead time in the auto industry perhaps entered government’s discourse before the passage of Traffic Safety Act of 1966, but after President Johnson signing that bill into law, it would never seriously leave auto regulation discourse again. Cars took a long time to design and build. While politicians and bureaucrats sailed the tides of the Washington calendar, automakers were at pains to explain that they had their own schedules. From one perspective, their remonstrations were essentially attempts at education. The car companies had to explain to Washington how their industry worked and how it was organized. No doubt, to support their cause, automakers highlighted the longest

58 “Verified Statement of Derwyn M. Severy,” 22. 59 “Affidavit of Colonel John P. Stapp USAP MC,” 5–7.

97 cases of new car development. Chrysler claimed, for instance, that it took the company three years to develop a new car. But the NHSB protested that much of this concern over lead time was mounted to obstructing the development of effective safety standards. As the agency wrote, “The intent of the Congress in passing the legislation was precisely to encourage design changes, not to freeze design. At some point, arguments on lead time changes from a legitimate problem of industrial engineering to covert attempts to freeze design to the economic advantage of the manufacturer, but to the detriment of safety objectives.”60

Staging 201: The AMA’s Theatre of the Absurd

While many of the automakers’ claims about Standard 201 can be accounted for by examining their reasoning over the course of the entire process and through all of the materials they submitted, certain aspects cannot be captured without attending to the details of individual interactions. Some things must be dealt with on their own with some sensitivity; their meaning rises from the idiosyncrasy of their expression, not their general content. The automakers’ behavior during the hearings provides a chief example of such a moment. The hearings were held on

May 22 and 23, 1967 in Detroit, Michigan, and May 24 and 25 in Washington, D.C.;

Potter presided in both venues.

60 NHSB, “Brief in Support of Proposed Findings of Fact: Before Presiding Officer Russell A. Potter,” NHTSA Microfilm, Reel Two, 7.

In his book, Science on Stage: Expert Advice as Public Drama, Stephen

Hilgartner describes how scientists and other technical experts use rhetoric and dramatic techniques to bolster and solidify their authority in public.61 The drama that representatives of the auto industry staged at the hearings had the exact opposite intention: it was meant to undermine the NHSB’s authority in the public eye. Perhaps Detroit aimed at a more important goal than influencing the public perception of the NHSB; automakers staged their hearing room drama with the eye of juridical authorities, such as judges and hearing officers, who could exercise power over an agency’s actions firmly in their sights. The automakers’ attorneys and technicians attacked the NHSB’s authority by staging Standard 201’s absurdity.

Lawyers are an absolutely essential part of any regulatory network—as much a part of the network as are the regulatory agencies and regulated firms. Lawyers both defend the current status of a given network and enlist new actors to it, including important decision-makers, such as judges.

The hearings were in many ways a face-off between Lloyd N. Cutler of the law firm Wilmer, Cutler, & Pickering and Robert M. O’Mahoney, the NHSB’s legal council. The two could not have been more different. Half a decade earlier, Cutler cofounded his firm, and it represented the AMA among other auto-related interests on the Hill and on Pennsylvania Avenue in Washington.62 He was not highly regarded in safety and consumer circles; as Ralph Nader once said, "People shouldn't forget that over the years he [Cutler] has represented the crassest

61 Stephen Hilgartner, Science on Stage: Expert Advice as Public Drama (Stanford: Stanford University Press, 2000). 62 For a partial biography of Cutler, see his obituary in the New York Times, “Lloyd N. Cutler, Counselor to Presidents, is Dead at 87,” May 9, 2005.

99 positions of the auto companies, the drug companies and others in their anti- consumer and anti-environmental activities."63 Cutler’s influence and power grew in Wahington; from Jimmy Carter’s term onwards, Cutler consistently acted as a presidential advisor on a host of different matters. As the writer of New York Times obituary wrote, he “shuttled between a lucrative private practice and the White

House.”

O’Mahoney, on the other hand, was an attorney who dedicated his life to public service. Before joining the federal government, he worked as the Deputy

Attorney General of Indiana. Later, O’Mahoney worked for the GSA, where he continued to address the automotive issues that had occupied him since the NHSB.

For instance, he played a role in the GSA’s efforts to purchase natural gas-powered low emissions vehicles, yet another effort designed to use the federal government’s purchasing power to foster innovation.64

The hearing began on a seemingly positive note. Cutler said, “We were pleased to hear counsel for the Bureau state at the . . . conference that a purpose of this proceeding is to produce information.”65 Immediately, however, the NHSB counsel began to suspect that all was not as rosy as the automakers’ lawyers tried initially to portray. O’Mahoney stated for the record that automakers did not treat the agency’s representatives with the deference that Cutler suggested was the case in the opening remarks. They asked to see the materials Cutler brought for the

63 Ibid. 64 “Environmental Currents,” Environmental Science and Technology, Vol. 4, No. 2, 95. 65 Hearing Transcript, 9.

100 hearing, but he allowed them to see one only item, a gray book that held pictures.66

As O’Mahoney exclaimed, “I must say I was incredulous then. I am incredulous now.

I don’t believe that.”67

Reading the transcript of the hearing makes one yearn for a video recording of the event. But none exists. At one point, GM Engineer Edwin Klove, while being questioned by GM’s attorney Edward C. Adkins, stood up and moved to a vehicle that was in the hearing space.68 There, he demonstrated how a test manikin was to be articulated to determine head impact areas in the proposed version of Standard

201. The transcript reads like some kind of unseen vaudeville scene. As Klove struggled awkwardly in the tight space of the passenger compartment with the bulky and ungainly manikin, Adkins asked, “Would you . . . demonstrate its lack of space for doing the things required in the establishment of [the] head impact area found in the standard as a relevant definition?” Apparently, several people had crowded around the car in order to see the demonstration. Potter turned and asked, “Mr. O’Mahoney, would you like to come and look at this?” O’Mahoney asked snarkily, “Is the dummy being used to demonstrate the non-existence of a dummy?

Or the lack of capacity of that particular dummy which doesn’t conform to Standard

201 to do what 201 requires?”69 Potter responded calmly, trying the move the

66 Ibid., 12-13. 67 Ibid., 13. 68 Adkins received his law degree from the University of Michigan and worked for the law firm Arthur and Hadden in Cleveland before becoming corporate counsel for General Motors. At GM, Adkins took part in the successful defense against Ralph Nader’s civil suit. Obituary, Star Beacon, August 18, 2000. 69 Hearing Transcript, 42.

101 matter forward; Adkins tried to continue his questioning; and then Potter called a five-minute, off-the-record recess.

Minutes later, after the hearing resumed, Klove continued to demonstrate the absurdity of the standard and its physical difficulty. Adkins asked, “Are your getting tired yet?” Klove answered, “I am getting out of breath.”70 He continued the play and reached a point where he was not able to manipulate the manikin in the required way with only two hands. The feet of the dummy had to remain on the floor, while someone moved its body. O’Mahoney asked, “Can the gentleman standing on the side be of assistance to him?’ Adkins suggested, “This is not a three- man procedure, is it? It is one operator for the dummy.” O’Mahoney protested,

“There is nothing in the standard about how many people [must] do it.”71 The point of all of Klove’s maladroitly wrestling with the dummy inside of the car, of course, was that the standard was unrealistic, unrepeatable, and impracticable. Samuel

Beckett could hardly have staged a better scene.

Slowly, over the course of the day even a dozy member of the audience could have deduced that the automakers’ testimony and, especially, their introduction of new exhibits were well-choreographed shots at making Standard 201 look foolish.

O’Mahoney’s anger began to build. If he was merely irritable during the dummy wrestling demonstration, than his Irish temper would later flare: “Mr. Potter, if it is helpful to shorten this proceeding, we will stipulate that it is the position of apparently General Motors and apparently AMA that their interpretation of the

70 Ibid., 47. 71 Ibid., 50.

102 standards [is] to design idiotic things they have been showing us. But I don’t see any point in going on with this.”72

GM’s engineers Klove and Ryder brought in redesigned interior features, such as window cranks (which raise and lower windows), which the engineers suggested met the standards objectives. Members of NHSB described the redesigned window regulator handles as “hopped-up.” GM’s outlandish design for such devices clearly betrayed the company’s attempt to make folly of the practicability requirement for the standard. NHSB members argued, in effect, that these artifacts had politics.73 The artifacts were like Rube Goldberg devices, not in terms of complexity, but in terms of its parodic intent to undress the human inclination to over-design the world. If the proposed Standard 201 had a range of design interpretations, engineers and designers at GM took their reading to a logical, comic extreme. There was nothing neutral about the bulbous, dome-like, silly-looking, seemingly ungraspable handles. In their very effort to be free of any sharp, pointed aspect that might wound a passenger, the GM-designed handles made their point. When O’Mahoney challenged this line of criticism, Klove maintained that “he knew no better solution.”74 Representatives of Ford Motor

Company and the executive vice president of Nissan Motor Corp. USA, Soichi

Kowazoe, who testified on behalf of the Japanese Automobile Manufactureres

72 Hearing Transcripts, 135. 73 Langdon Winter, “Do Artifacts Have Politics?”, Daedalus, Vol. 109, No. 1, Modern Technology: Problem or Opportunity? (Winter, 1980),121-136. 74 “Report of the Presiding Officer,” 22.

103

Association, made similar statements.75 Kowazoe declared, “I can assure you, sir, that this (the Japanese design) was strictly done in Japan by our engineers according to their interpretation of the standard. I particularly want to say, especially, [that] I had a big argument with Tokyo about showing this crazy-looking screwhead.”76

These absurdist displays of how the automakers would have to redesign their cars’ interiors finally drove O’Mahoney over the edge:

“I have got to say, we came in here with no advance notice of what these people were going to do. We have a witness up here who gets up and shows us things, he says this is the only way I can comply with this standard. We will get to this in cross-examination. I must say, as a show—and that is what this is—for this assemblage, without any advance notice that this was the sort of thing you were going to do, it seems to me it is terribly violative of the spirit of this proceeding. It is a sort of like a sideshow. I think it is unfortunate.”

Potter responded, “Mr. O’Mahoney, I believe that the demonstrative testimony is necessary and serves a very useful purpose in illustrating what the end result of the standards will be.” O’Mahoney shot back, “Mr. Potter, that is just the problem. They don’t demonstrate what the standards will be. They demonstrate what these people choose to make the end results of the standards by a rather extreme interpretation of them.”77 When Potter would not hear his complaints, O’Mahoney suggested that, at the very least, “I would like an instruction, Mr. Potter, from you, to the parties who are conducting this show, that they should conduct it in a manner befitting the

75 Ralph Kisiel, “Soichi Kawazoe,” Automotive News, May 19, 2008. Kawazoe had earlier received his B.Sc. from Dayton University in 1930 and MS from MIT 1931. He worked for General Motor’s and Ford’s Japanese branches before moving Nissan. 76 “Report of the Presiding Officer,” 24. 77 The preceding three quotations are all from Hearing Transcripts, 136-7.

104 dignity of a serious pursuit of fact involving the safety of the citizens of this country.”78

By the end of the first day, it was clear that O’Mahoney and the NHSB had been run over by the automakers’ traveling theatre troupe. As those testifying for the auto companies continued to discuss how the proposed Standard 201 would hypothetically affect interior design, O’Mahoney pleaded uselessly, “Mr. Potter, I came here ready to listen, to hear, and to see. I am still ready to do that. But I am not interested in hypothetical problems, only real problems. Neither is the administration for which I work. We want real problems, not hypothetical problems, not assumptions but problems.”79 Potter, however, continued to believe that the automakers’ testimony had identified “real” problems, and he would take those into account in his final act as hearing officer, writing the Recommended

Findings report.

Closing Standard 201: Potter’s Recommended Findings and the NHSB’s

Response

On June 22, 1967, Potter issued his Recommended Findings report. It devastated the NHSB’s cause. Potter broke his analysis into four parts, with the first three dealing with the requirements of the law—the need to be clear and objective, the need for practicality, and the need for proved safety. Potter’s fourth point treated the issue of whether it was “reasonable to require compliance by January 1,

78 Ibid., 140–1. 79 Ibid., 149.

105

1968.”80 He argued that it would not be fair to require the automakers to comply by this date because it did not give them enough lead time.

Potter also asserted in his report that the 95th percentile male and 5th percentile female manikin criteria should be jettisoned: “People are a variety of sizes. Long-legged men may be in the 95th percentile group for standing height though their sitting height is below the average.”81 In his view, the size definitions were completely arbitrary, or what some might call “socially constructed.” Potter concluded, mincing no words, “Therefore, the definitions of the head impact area, and the knee and leg impact area based on use of manikins are vague, misleading, inexact, and completely unworkable for design purposes.”82Moreover, Potter was overwhelmingly moved by General Motors’ material demonstration of how automakers’ parts would need to be redesigned if Standard 201 was promulgated and enforced. He called the testimony of Edwin Klove and George Ryder on this point “the most informative and persuasive of all evidence introduced throughout the proceeding.”83 Consequently, Potter declared that the NHSB should give up its requirement on protrusions and other interior features.

The NHSB issued a rebuttal to Potter’s Recommended Findings on July 3.

Potter’s report contained many errors and “erroneous concepts,” the agency claimed, but most of these related to the substance of Standard 201, which the agency had already “conceded to be deficient” in a post-hearing brief dated June

80 “Report of Presiding Officer.” 81 Ibid., 6-7. 82 Ibid., 7. 83 Ibid.,

106

5.84 The NHSB did not want to make exceptions to the issues it had already conceded; rather, it took issue with Potter’s procedural philosophy. Most importantly, the rebuttal claimed, “The [Recommended Findings] Report

Considered the Proceedings to be Adjudicatory and Followed the Quasi-judicial

Requirements of Sections 7 and 8 of the Administrative Procedure Act.”85 The

NHSB staff members were saying that the hearings were meant to create information that the agency would take into account when creating its standards;

Potter had shaped the hearings as an adjudicatory process meant to judge the justice of the agency’s standards. These were two very different things.

At the heart of the matter lay the question of whether the proceedings should have been quasi-legislative or quasi-judicial. The NHSB’s disagreement with

Potter on this point began early in the process. O’Mahoney had called Potter on

May 15, before the hearings, to be clear about how the hearings would proceed.

Potter suggested to them that the support materials (written testimony, engineering studies, etc.) that the NHSB had submitted for the hearings were inadequate. As O’Mahoney reported, Potter suggested that he would not be asking the NHSB any questions during the cross-examination period because it was not his role to “help it out at all.”86 O’Mahoney took this to mean that Potter saw his role as being close to a judge, while he believed that the purpose of the hearings should be to “establish fact,” which the agency would then take into consideration in revising

84 Federal Highway Administration, “Reconsideration of Initial Motor Vehicle Safety Standards: Docket FHA–1: Bureau Counsel Exceptions to Recommended Findings,” NHTSA Microfilm, Reel 2. Hereafter, “Bureau Counsel Exceptions.” 85 Ibid., 1. 86 Memorandum, Robert M. O’Mahoney to “The Files,” “Telephone Conversation with Hearing Officer Potter,” May 15, 1967, NHTSA Microfilm, Reel 2, 1.

107 the standard. Rather, Potter had claimed in the May 15 phone call that one of the questions was whether “the requirements in Standard 201 [were] reasonable, considering costs, appearance of the automobile, and saleability of the automobile.”

That is, he would be judging the standard itself. As O’Mahoney wrote, “It seems clear from this conversation . . . that we will have difficulty with Mr. Potter.”

The NHSB’s rebuttal argued that although “the Standard should have enjoyed a presumption of regularity, instead the report consistently balanced the evidence and found that the weight of the evidence was with the petitioners.”87 The

NHSB claimed that the agency’s administrator had intended the public hearings and submissions as a means to amend and improve the Standard, but Potter had asserted the opposite, viewing the hearings as a review of Standard’s justice. In the agency’s view, “the presiding officer’s complete and puzzling failure to grasp this rather simple instruction” had led him to produce a document that was of “little or no utility in achieving the objective of the Administrator’s hearing order.”88

Moreover, the agency believed that Potter had not come to his basic position on the nature of the hearing on his own:

“Bureau Counsel recognized this misunderstanding on the part of the presiding officer early in the proceeding and made frequent, but obviously unsuccessful attempts to turn the proceeding away . . . from the adversary character to which it was allowed to descend albeit with active encouragement of counsel for the Automobile Manufacturers Association and other United States manufacturers, most particularly counsel for the General Motors Corporation.”89

87 “Bureau Counsel Exceptions to Recommended Findings,” 3. 88 Ibid., 5. 89 Ibid.

108

But O’Mahoney’s argument was largely futile. Earlier, the New York Times—after discussing how Haddon believed that Standard 201 would stand up to scrutiny— had written, “The hearing commissioner [that is, Potter] will be given final authority on whether to grant the manufacturers’ demands for modification, Dr.

Haddon said.”90 It seems that, in the Bureau’s rebuttal, it was trying to back itself out of a deal.

For some time, historians of technology have discussed the notion of

“closure”—the moment when a given technological artifact stops changing in radical ways and becomes relatively stable. Standards also close. The final steps of how Standard 201 came to reach closure are as “closed” to us (in the other sense of the word) as 201’s origins within the NHSB; these last steps are invisible. No archival records document any meetings subsequent to the date the NHSB submitted its rebuttal. Perhaps, Haddon, Brenner, and others met and strategized, hoping to beat Potter’s recommendations. Perhaps, they depressingly resigned themselves to their fate. In any case, on August 13, 1967—a little less than three months after the hearings—the New York Times carried a story—“Safety Standards on ’68 Cars Eased”—on Standard 201, noting that “requirements for extra knee and leg impact protection, and changes of interior handles and knobs have been dropped.”91 The news story was one of the few times that the automotive standards-making process had broken the surface of bureaucratic anonymity, making its way into the public eye. The message was simple: The NHSB had been defeated.

90 “Auto Makers Sue on Safety Rules,” The New York Times, April 1, 1967. 91 “Safety Standards on ’68 Cars Eased,” New York Times, August 13, 1967.

109

Conclusion

The fight over Standard 201 illustrates the nature of rule-making at NHTSA.

As several scholars have argued, in the early 1970s, the agency moved towards a

“technology-forcing” strategy in its attempt to get airbags, a largely untested technology, into vehicles. Though the reasons are highly complex, these early efforts failed. One of the key causes of this failure was that, in many readings of the

Traffic Safety Act, the agency lacked legislative authority to promulgate

“technology-forcing” rules. While the Traffic Safety Act did not require automotive safety standards to be based on “best-available” technology, for all intents and purposes, they were because initially the Act required that NHSB’s standards be based on existing standards.

The success of the TSA lay in getting available, or “off-the-shelf,” technologies into vehicles in an industry that valued cost-cutting over safety and even quality. Later, largely due to the rise of cost-benefit analysis and economic impact studies, NHTSA adopted the conservative policy that it needed to prove that some technology could meet a potential standard before proposing it. The agency did not mandate this technology, which would have violated the law’s requirements that the agency issue standards based on performance, not design; rather, the agency sought to prove that at least one, affordable technology could meet a proposed new standard. Industry was free to use other technologies if it saw fit.

Part of what hindered the NHSB’s power was that its standards were based in rule-making instead of legislation. As I discuss in Chapter Four, legislators wrote

110

the federal automotive emissions standards (which were and are administered by

Environmental Protection Agency) into the law.92 The NHSB, on the other hand,

had to struggle through each step of the rule-making procedure. In this case, the

law had a structuring effect on the struggle between regulators and regulated

firms.93 Another way to consider this is to ask, “If each side of the regulatory

equation amounts to a competing network, to what end is each network aimed?

Most noticeably, the automakers and the lawyers and consultants they employed

poured huge amounts of time, resources, and scientific and technical labor into

undermining the NHSB’s efforts. The agency also tried to fight back, but, as we have

seen, it lacked the resources, organizational capabilities, and bureaucratic know-

how to do so. The economist William Baumol famously distinguished three forms of

entrepreneurial behavior—productive, unproductive, and destructive.94 In

Baumol’s view, productive entrepreneurialism is what we think of as the traditional

entrepreneur: risk-taking, “innovation”-creating, and productivity- and efficiency-

making, the entrepreneur finds new technologies and techniques that unleash what

Joseph Schumpeter called the “winds of creative destruction.” But Baumol realized

92 The economists David Gerard and Lester Lave write about this important division between standards produced through rule-making and standards that are written into the law itself in their insightful essay, “Experiments in Technology Forcing: Comparing the Regulatory Processes of US Automobile Safety and Emissions Regulations,” International Journal of Technology, Policy, and Management, Vol. 7, No. 1 (2007), 1-14. Yet, Gerard and Lave focus only on the formal differences between the power of Congress versus the power of regulatory agencies. I argue that organizational capabilities and leadership within the agencies played an equally important role in shaping outcomes. Moreover, while, as Gerard and Lave note, the NHSB may have attempted a technology- forcing strategy with airbags in the late-1960s and 1970s, the outcome of the Standard 201 hearings largely predetermined the agency’s inability to successfully carry out such a program. 93 Richard John uses the language of structuring in his recent book, Network Nation. John’s phrasing about the idea sometimes approaches the work of Anthony Giddens, whose work is also relevant in this regard. For Giddens on the theory of “structuration,” see Anthony Giddens, Constitution of Society: Outline of a Theory of Structuration (Berkeley: University of California Press, 1986). 94 William J. Baumol, “Entrepreneurship: Productive, Unproductive, and Destructive,” The Journal of Political Economy, Vol. 98, No. 5, Part 1. (Oct., 1990), 893-921.

111 that not all “innovative” behavior fits this image. The classic example of unproductive entrepreneurialism is “innovative” ways to increase “rent-taking.”

According to Baumol, in Ancient Rome, for instance, the three ways elites made income was through “landholding (not infrequently as absentee landlords), ‘usury,’ and what may be described as ‘political payments.’” None of these activities contributed positively to society. Finally, destructive entrepreneurism takes the form of any, but typically criminal, activity that diminishes social welfare.

Examples might include innovative techniques for producing or distributing illegal narcotics.

Of course, the boundaries between these three forms of activity are murky and are highly sensitive to one’s political and moral understanding of the world.

The automakers’ legal efforts to destroy or weaken auto safety standards were at least unproductive.95 The energy, time, money, and scientific and technical resources could have been put toward making the car safer, but instead they went toward legalistic wrangling. Additionally, the automakers employed their design staffs to fashion the grotesque artifacts in the hearings discussed above. If one believes that unsafe auto design was and is immoral and even criminal, as Ralph

Nader and other safety advocates did, then the automakers’ legal efforts would probably appear to be a form of “destructive entrepreneurialism,” since the parties sought to defend immoral and criminal behavior. Whether unproductive or

95 Historians of business of the “Chandlerian” persuasion have expended a great deal of time and energy explaining how large firms either built or contracted out important business functions, such as research and development. So far, legal capabilities within (the legal department) and without (legal retainer) the firm have not received a comprehensive or comparative treatment. Such a study might suggest under what conditions in the business environment, including regulation, firms find it profitable or necessary either to create strong legal departments or to hire external law firms—to “make” or to “buy” legal services.

112 destructive, the automakers “innovations” were rhetorical. To put the matter differently, historians of business have for a long time examined how corporations used branding, advertising, and other forms of marketing to enlist consumers to the companies’ products. Their “innovative” activity was geared towards winning a broad swath of the “market.” But in the case of regulatory rule-making, the automakers put their energy to winning just one person, the decision-maker

(Russell A. Potter in the instance of NHSB’s proposed Standard 201). In the parlance of economics, therefore, regulation via agency-borne rule-making may create incentives for corporations to pour resources into the unproductive means of legalistic entrepreneurialism, or “legal jousting,” instead of into beneficial technical change.

As scholars have noted, NHTSA’s later efforts at technology-forcing foundered. But the roots of this failure—the agency’s inabilities to create tough standards, or, more exactly, to tighten preexisting ones—lay in the construction and closure of Standard 201.

Chapter 3—The Banality of Justice

“[Federal Motor Vehicle Safety Standard] 204 is ambiguous in many respects. . . . A standard that is so ambiguous and indefinite cannot be relied on for the purpose of seeking civil penalties. Louisville & N. R. R. Co. v. Railroad Commission of Tennessee (C.C. Tenn., 1884) 19 Fed. 679; Waters Pierce Oil Company v. Texas (No. 1), 212 U.S. 86, 108-11 (1909); International Harvester Company v. Kentucky 234 U.S. 216 (1914); Champlin Refining Co. v. Corporation Commission of Oklahoma, 286 U.S 210, 243 (1932); Conally v. The General Construction Co., 269 U.S. 385, 391 (1926); Cline v. Fink Dairy Company, 274 U.S. 445, 465 (1927); United States v. L. Cohen Grocery Company, supra; A. B. Small Co. v. American Sugar Refining Co., 267 U.S. 233, 45 S.Ct. 295 (1925); Yu Cong Eng v. Trinidad, 271 U.S. 500, 46 S.Ct 619 (1926); Stromberg v. California, 283 U.S. 359, 51 S.Ct. 532 (1931); Lanzetta v. New Jersey, 306 U.S. 451, 59 S.Ct. 618 (1939); Winters v. New York, 333 U.S. 507, 68 S.Ct. 665 (1948); United States v. Five Gambling Devices, 346 U.S. 441 at 452, 74 S.Ct. 190 (1953); People v. James, 162 N.Y.S. 2d 927 (1957); People v. Zanchelli, 8 Misc. 2d 1069, 169 N.Y.S. 2d. 197 (1957); People v. Sisson, 176 N.Y.S. 2d. 785 (1958); United States v. 174 Cases, etc., (D.N.J. 1961) 195 F.Supp. 376.”

—Edward Garfield, Esq, Counsel for Fiat Motor Company, Inc. to James B. Minor, Esq., General Counsel for the National Highway Safety Bureau, December 1, 1970

On September 24, 1969, four staff members of the National Highway Safety

Bureau (NHSB) traveled to Los Angeles, California, to witness crash tests run by the

Digitek Corporation, a private firm that performed tests for the agency.1 The company was to crash a car, a 1969 Fiat 850 Sedan. Having agency staff members oversee tests carried out by contractors was standard operating procedure. Digitek was also crashing some Volvos that day and doing a “joint post-impact inspection” with a representative from Volvo of America, Inc. Thus, the NHSB may have seen the day as particularly sensitive because the test contactor was meeting with one of

1 Safety Standards Engineers to Chief of Validation Division, “Trip Report, September 24-27, 1969, to Digitek Corporation, Los Angeles, California,” October 7, 1969, 1. The four employees were J. Watts and D. Porter of NHSB’s Los Angeles Regional Office and R. Gardner and J. McDowell of the agency’s Office of Performance Analysis. If historians are now “bringing the state back in” again, for instance, by examining how government has shaped business, they should also pay attention to the inverse relation, the role of private firms in both shaping and executing government practices. Here, I am not referring to regulated firms, which attempt to sway government procedures through lobbying and other means, but other firms that work for the government. Too often, scholars speak as if privatization and government contracts began with Ronald Reagan, but contractors have long played an important role in doing government work. For example, firms have performed the bulk of crash testing for NHSB (and later NHTSA) from the very beginning, since the Traffic Safety Act’s passage in 1966. Each of these firms has its own story, and their internal routines played some part in structuring the outcomes—that is, the success or failure—of regulatory testing.

113 114 the automakers, a meeting that could have fostered collusion and regulatory capture, or at least accusations thereof. Regardless, the NHSB employees showed up as planned, but not everything went as expected. When the employees arrived at

Digitek, the “850 Fiat was inspected and it was found to be an 850 Coupe and not the 850 Sedan with the three-segment steering column.”2 NHSB’s purchasing agent had ordered the wrong car. The NHSB staff members called back to the home office, which gave them permission to go ahead with the test, using the Coupe instead of a

Sedan. With the four staff members watching, Digitek crashed the Coupe. It fared poorly. And so began a two-year process that would end in Fiat’s recalling thousands of vehicles, paying the federal government thousands of dollars in fines, and walking away with more black marks on its already tarnished reputation. The

NHSB staff members flew home and started doing paperwork.

The notion of mundane, or banal, work, like that of the NHSB staff members, received a good deal of attention during the 20th century, little of it positive. In

1961, the philosopher Hannah Arendt attended the trial of Adolph Eichmann, the

Nazi logistics specialist who played a major role in the Holocaust. She reported on the trial for The New Yorker magazine throughout the event.3 Arendt characterized

Eichmann’s psychological and social constitution. He was, she believed, neither anti-

Semitic, nor mentally ill but simply a social climber who sought the approval of his superiors; e had simply been a part of “the banality of evil.” Whereas, the Holocaust was a horrifying, earth-shattering act of genocide, the people who perpetrated it

2 Ibid. 3 Hannah Arendt’s reports were collected in Eichmann in Jerusalem: A Report on the Banality of Evil (New York: The Viking Press, 1963).

115 treated it as if it were as mundane as any other kind of work, as ordinary as pushing paper. Arendt’s phrase about evil’s banality has become part of our common tongue.

Earlier books, such as David Riesman’s The Lonely Crowd, William H. Whyte’s The

Organization Man, and Sloan Wilson’s The Man in the Gray Flannel Suit had decried the conformity—the “lives of quiet desperation”—that humdrum, bureaucratic labor had brought to modern life.4 These works, as well as the equally anti- bureaucratic writings of the Frankfurt School, Max Horkheimer, Theodor Adorno, and, especially, Herbert Marcuse, deeply influenced emerging political thought in the 1960s, including the New Left and the student movement.5 Routine labor had a bad rap.

This development was somewhat ironic. Throughout the 1960s and 1970s, the US federal government passed a number of laws—including the Traffic Safety

Act, the Clean Air Act, and the Occupational Safety and Health Act—and built a number of federal agencies meant to protect workers, consumers, and the environment. The political left heralded and commended these government interventions on the part of ordinary citizens. Yet, to make them work, these laws required a great deal of mundane work, including constant meetings, memo and

4 David Riesman in collaboration with Reuel Denney and Nathan Glazer, The Lonely Crowd: A Study of the Changing American Character (New Haven: The Yale University Press, 1950); William H. Whyte, The Organization Man (New York: Simon and Schuster, 1956); Sloan Wilson, The Man in the Gray Flannel Suit (New York: Simon and Schuster, 1955). Henry David Thoreau coined the phrase “lives of quiet desperation” in his Walden (1854). 5 Banality fits Theodor Adorno’s and Max Horkheimer’s contention that “rationality” would increasingly come to dominate and stultify everyday life. Of course, they drew on older traditions of thought to make this claim, especially Max Weber’s notion of the “cage of reason.” See, Horkheimer and Adorno, Dialectic of Enlightenment: Philosophical Fragments, trans. Edmund Jephcott (1944, Standford: Stanford University Press, 2002). Marcuse believed that rationality, especially its technological varieties, would make humanity into a “one-dimensional man,” lacking in depth and the ability for critical thinking. See, Marcuse, One-Dimensional Man: Studies in the Ideology of Advanced Industrial Society (Boston: Beacon Press, 1964); see also, Marcuse, Eros and Civilization: A Philosophical Inquiry into Freud (Boston: Beacon Press, 1955).

116 report writing, scientific studies, and general paper-pushing, all of which leads one to ask, is evil’s banality also justice’s? Do the hum-drum habits of the “organization man” (and woman) play some part in fulfilling a nation’s hopes, including safety and environmental protection?

By justice, I mean simply that each person receives his or her due. Until the passage of the Traffic Safety Act of 1966, citizens and politicians viewed a car’s safety, or lack thereof, as a matter strictly between automakers and consumers. The ideology of caveat emptor—“let the buyer beware”—ruled the day. Over the course of the 1960s, this mindset slowly shifted, however, into something that scholars have characterized as the development of “risk society.” As the sociologist Anthony

Giddens wrote, society became “increasingly preoccupied with the future (and also with safety).”6 The safety of consumer products, including automobiles, became a matter of justice that demanded government action when producers failed their moral duty.

In this chapter, I will examine an investigation into Fiat cars that the NHSB, which became the National Highway Traffic Safety Administration (NHTSA) in 1970, carried out from 1969–1970. As with the standards-building that I examined in the previous chapter, none of the NHSB’s investigations or recall efforts is representative. Each is unique. But each case arises from the daily labors of regulators, who work to ensure that automobiles built or imported in the United

States meet the criteria that federal standards demand. As I argue, this work constitutes the backbone of federal regulatory enforcement, and, therefore,

6 Anthony Giddens, “Risk and Responsibility,” The Modern Law Review, Vol 62, No 1 (January 1999), 3.

117 historians of business and technology must begin attending in earnest to these quotidian affairs. We must turn to the practices that constitute daily life, including regulation.

Trials of the Ordinary

Historians of business and technology have traditionally focused on novel and successful businesses and technologies, including an emphasis on invention and innovation. This focus is reflected in the history of regulation, where historians have concentrated either on the visions of “policy entrepreneurs” or on the embattled emergence of new regulatory legislation. In both cases, the focus is on

“innovations” in governance. For some time now, critics have called for invigorated attention to historical cases of failure, cases where, for whatever reason, businessmen, inventors, and system-builders fell on their faces.7 But between the heady heights of innovation’s bleeding edge and the nightmarish pits of loserdom lies a wide swath that most businesses, inventors, and ordinary people inhabit. The history of the ordinary, the everyday, or the quotidian has its own, albeit thin, historiography, especially around the history of ordinary technology.

In his 1934 book Technics and Civilization, Lewis Mumford argued that cutting-edge technologies pushed civilization through subsequent stages, from a

7 Some examples include Michel Callon, “Society in the Making: The Study of Technology as a Tool for Sociological Analysis,” The Social Construction of Technological Systems: New Directions in the Sociology and History and Technology, Wiebe E. Bijker, Thomas P. Hughes, and Trevor Pinch (Cambridge, Mass.: The MIT Press, 1987), 83–103; Bruno Latour, Aramis, or the Love of Technology (Cambridge, Mass.: Harvard University Press, 1996). See, also, the special issue of Social Studies of Science edited by Hans-Joachim Braun on “Failed Innovations,” Vol. 22, No. 2 (May 1992). Scott A. Sandage has examined the social and psychological ramifications of business failure in his Born Losers: A History of Failure in America (Cambridge, Mass.: Harvard University Press, 2005).

118 wood-based society, to one of iron, and from to one based on steel to one finally based on aluminum and plastics. Fourteen years later, however, the Swiss historian and critic Siegfried Giedion focused not on the novel but on the prosaic in his classic work, Mechanization Takes Command: A Contribution to Anonymous History. What made the history that Giedion sought to uncover and interpret “anonymous” was his attention to the banal details of pedestrian technologies, details that are so commonplace that their richness often escapes our awareness.8

Giedion’s turn to the “anonymous” shares much in common with two philosophical traditions that were popular in the mid-20th century, namely phenomenology (especially in its variety known as “existentialism”) and ordinary language philosophy, both of which “called people back” to the mundane in some sense. Similarly, each of these traditions believed that there was some power in turning to the everyday. The ordinary language philosophers eschewed jargon in favor of the common tongue, which they believed was simply more powerful and apt than neologisms. Phenomenologists turned away from the abstraction of

Platonic ideas and towards the human being’s (or consciousness’s) rootedness in the present moment. Giedion’s writing had a similar, committed tone of mission. As

8 Giedion, Mechanization Takes Command: A Contribution to Anonymous History (New York: W. W. Norton & Company, Inc., [1948] 1969). Giedion’s sense that we somehow overlook, or neglect, the commonplace, especially the connections between anonymous things, is similar, if in an inverted way, to J. L. Austin’s conception of philosophy. Part of what philosophers do is to create distinctions, distinctions that hopefully show us something about the world. But much of Austin’s work involved showing, as Stanley Cavell writes, “the slovenliness, the grotesque crudity and fatuousness, of the usual distinctions philosophers have traditionally thrown up.” In the rush to cleverness, philosophers cover over the finer, more natural distinctions of plain speaking. In Cavell’s characterization, Austin tried, in a fashion opposite to typical philosophers and more like an “art critic,” to use distinctions to “compare and distinguish” in order to bring “capacities and salience of an individual object in question . . . to attention and focus.” He tried to bring our attention to the power of distinctions that already lay at hand in the everyday. Cavell, “Austin at Criticism” in Must We Mean What We Say?: A Book of Essays (Cambridge: Cambridge University Press, [1968] 2002), 103.

119 he noted, the historian’s “role is to put in order . . . what we experience piecemeal from day to day, so that in place of sporadic experience, the continuity of events becomes visible.”9 Following his instinct that the human tools had “molded the present,” Giedion chose to focus on the simplest of technologies—“humble things, things not usually granted earnest consideration.”10 Drawing an analogy with painting, in which the subject does not matter as much as the artist’s rendering of it, anonymous history’s role was to draw together the connections—“the uncovering of relationships”— between mundane things.11 “The sun,” Giedion wrote, alluding to the age-old symbol of truth and beauty, “is reflected even in a coffee spoon.”12 Yet,

Giedion’s sense of the connectedness of things led him to believe that their total effect was greater, as the saying goes, than the sum of their parts. Making constant reference to history’s and science’s parallel uncoverings of hidden worlds, he claimed, “The slow shaping of daily life is of equal importance to the explosions of history; for, in the anonymous life, the particles accumulate into an explosive force.”13

Giedion spoke before his time. He published Mechanization Takes Command in a United States that was quickly headed into the heady high-tech, high-science world of the Cold War and the Space Race. The cutting-edge was in; the anonymous was out. By the mid-to-late 1960s, however, the shine had worn off technology’s

9 Giedion, 2. 10 Ibid., 2–3. 11 Giedion’s emphasis on the connectedness between things and his focus on mundane objects suggests its likely influence of James Burke’s highly successful television show and book, Connections (Boston: Little, Brown, and Company, 1978). 12 Giedion, 3. 13 Ibid.

120 chrome image.14 The public and the media began to become preoccupied with various technological risks, especially those related to consumer safety and environmental concerns. One of the less obvious ways in which this Zeitgeist was expressed was in a focus on maintenance.15 In opposition to a social predilection to emphasize the novel and the new, thinkers began to examine how we live with the old and the common.

The economist Albert O. Hirschmann had hit upon this idea earlier in his

1958 book, The Strategy of Economic Development.16 During the early-to-mid-1950s,

Hirschmann served in various capacities as an economic advisor for developing nations. After publishing a book in 1955 on economic growth in Colombia,

Hirschmann turned to a broader consideration of the issue of development. One of the many topics that Hirschmann covered in The Strategy of Economic Development was maintenance. In his view, the level of technological improvement in a society did not matter if the technology could not be maintained or kept up. He was

14 Thomas Parke Hughes, ed., Changing Attitudes Toward American Technology (New York: Harper Row Publishers, 1975); idem., American Genesis: A Century of Invention and Technological Enthusiasm, 1870–1970 (Chicago: University of Chicago Press, 2004), ch. 9. 15 The enforcement of auto safety standards, as described in this chapter, is somewhat analogous to maintenance. Of course, in the history of technology, calls to pay attention to maintenance are deeply connected with warnings that we focus too much on innovative technologies (and often “high technologies” at that). Historians of regulation often focus similarly on “innovations.” At the most superficial level, this skewed emphasis takes the shape of focusing too much on “policy entrepreneurs” and legislation, on the moment when a regulation is born instead of on the long, hard, and typically bureaucratic work of making sure it becomes fulfilled in the world. But the problem extends deeper than this. For instance, some, including Jerry L. Mashaw and David L. Harfst, have argued that auto safety enforcement efforts aimed at generating recalls are not nearly as effective as technology-forcing regulations, which induce manufacturers to generate answers to technological problems through R&D. But, in some ways, technology-forcing is to innovations what enforcement is to maintenance. No matter to what degree an agency continues churning out technology-forcing regulations (and there is good reason to believe that the likelihood of this is exactly zero), established regulations will need to be watched, preserved, enforced. Thus, enforcement is the “maintenance” of regulations. Mashaw and Harfst, The Struggle for Auto Safety (Cambridge, Mass.: Harvard University Press, 1990). 16 Albert O. Hirschmann, The Strategy of Economic Development (New Haven: Yale University Press, 1958).

121 especially sensitive to this issue because of the number of Cold War-based technology transfer programs going on around the world: developed countries, but especially the United States and the USSR, sent technologies (and often some sort of minimal training program) to, what were known in the language of the day as,

“Third World” nations, in an effort to win the loyalty of those countries. The battle between capitalism and communism was fought, in part, through technological innovation and the distribution of it. Yet, these technology transfer programs often had perverse results. Broken machines without replacement parts, dead engines without the requisite knowledge to fix them—the technologies inevitably broke down, rusted, and moldered. Hirschmann viewed this phenomenon as a societal failure to develop a “maintenance habit.” He characterized “the lack of proper maintenance” as “one of the most characteristic failings of underdeveloped countries and one that is spread over the whole economic landscape.”17 Unless a society became infected with the “compulsion to maintain,” Hirschmann argued, imported innovations would amount to nothing and even indigenous innovations would die on the vine.

The person who took the most philosophical and, even, spiritual approach to maintenance was Robert Pirsig, who enshrined it in his best-selling 1974 book, Zen and the Art of Motorcycle Maintenance. Pirsig wrote partly in response to those trends in the late-1960s that sought to flee technology. Western culture, he admitted, typically favored what he called the “Classical” mode of rationality, which produced science and technology and which came with benefit of instant

17 Ibid., 141.

122 gratification. This favoring of the Classical had led to problems, including all of the technological risks so present to people’s minds at that time. But Pirsig worried that people were beating a retreat into, what he characterized as, a “Romantic” worldview that rejected science and technology. This dichotomy between the

Classical and the Romantic lay at the heart of Western society, Pirsig argued. He sought to overcome this divide by appealing to a different, broader philosophy—his

“metaphysics of quality”—that was inspired by Eastern thought and that encompassed both the Classical and Romantic modes of thinking. Maintenance played a central role in Pirsig’s proposed solution. By living in the moment with technology in a contemplative fashion and replying with sensitivity to technology’s need for upkeep, Pirsig believed that people could transcend the division between the two modes, incorporating both into a more fulfilling, grounded, and holistic vision. The health of society required attending to technology’s everydayness.

Since this time, historians of technology have tarried in various ways with the role technology plays in our everyday lives and beckoned others to do the same.

As in Giedion’s work, this path remains a prophetic one, a calling back. The message is repeated periodically, but fails to stick for long, only to be repeated later once again. In 1990, the historian Arnold Pacey published Technology in World

Civilization: A Thousand-Year History. By taking a historical view that encompassed the whole world, Pacey hoped to escape the narrower, skewed visions that favored economically developed nations. Part of this change of perspective included moving away from cutting-edge technologies towards old and ordinary ones. Most of the world relies almost completely on technologies that are as far as possible away from

123 the cutting-edge. Even in rich Western nations, most of the technologies that surround us—the incandescent light bulb, the electric fan, the Otto four-stroke engine, etc.—have not fundamentally changed since their first introduction over 100 years ago. Pacey called historians to examine how these most unglamorous things shape our lives. More recently, in his book The Shock of the Old, the historian David

Edgerton has echoed Pacey’s concerns. In Edgerton’s account of the historiography of technology, historians have focused for too long the history of inventions and innovations, a production-centered history that necessarily privileges the new and novel. Edgerton’s strategy is to focus on what he calls “use-centered” history.18 By focusing on how people actually use technology, we will realize that old technologies play a continued, even obdurate, role in daily life.

As with any insight, however, the turn towards the ordinary, anonymous, prosaic, mundane, and everyday is not a cure-all. While our habits often become so mundane that we fail to take notice of them, no single day recurs. Routinization and rule-making can never capture every contingency. The world bucks the human sense of order. This lack of perfect repetition pushes us to think and make choices.

Scholars often describe durable habits within organizations as routines or procedures.19 Bureaucrats attempt to rein in chaos by establishing procedures.

They try foster dependable, repeatable behaviors in their staff members. Yet,

18 Edgerton, The Shock of the Old: Technology and Global History Since 1900 (Oxford: Oxford University Press, 2007), xi–xiv. 19 Herbert Simon did much to initiate the contemporary “behavioral” interpretation of organizations in his Administrative Behavior: A Study of Decision-Making Processes in Administrative Organizations (New York: MacMillan, 1947). Richard R. Nelson and Sidney G. Winter examined how firms developed “routines” through an evolutionary process in their An Evolutionary Theory of Economic Change (Cambridge, Mass.: Belknap Press of Harvard University Press, 1985).

124 ultimately, this “proceduralism” is a fantasy. No matter how many rules, laws, and protocols we create, the world produces cases that fall outside or between our criteria. In these cases, someone must decide.20 In this chapter, I try to describe how this interaction between proceduralism and the need to make decisions played out in the case of enforcing automotive safety standards.

As I argued in Chapter Two, the federal government’s auto safety programs have never centered on “technology-forcing,” pushing companies to develop new, currently unavailable technologies. Yet, the creation of standards, and the tests that embody the standards, outline the state’s potential technological influence in that they inherently cover the contours of technical change that may emerge. But nothing about the mere existence of standards guarantees their fulfillment. Standards, especially those that are contested and resisted, must be brought into being. If federal testing procedures create the technical possibility of regulation, the way in which federal bureaucrats enforce and carry out these tests produces technical reality.21 In federal agencies, as in other organizations, procedures like tests

20 This problem has received more attention in the philosophy of law than it has in studies of bureaucracy and administration, probably because the latter is so imminently practical. Both rightists and leftists have learned a great deal about this phenomenon from Carl Schmitt, the Nazi jurist, whose intellectual position is sometimes known as “decisionism.” For Schmitt, there will always be an exception to the rule, and, in these cases, the “sovereign” must step in to decide one way or the other. Schmitt errs, however, when he believes that exceptions are decided only by the highest authority. (One can see how this view fit neatly with National Socialism.) In reality, exceptions and decisions are spread throughout the ranks, and bureaucrats often find themselves facing so-called “executive decisions.” 21 Of course, firms are free to develop the regulated aspects of their technologies beyond the requirements embodied in government-mandated standards, and in some cases they have, but as a rule, federal performance standards regulate those facets of technology that are seen as socially beneficial but that, for various reasons including lack of “consumer demand” or insufficient threat of civil litigation, companies have chosen not to foster. In these situations, the limits of regulation are the limits of technological change. As I will examine in Chapter Four, federal performance standards create a dynamic in which firms try to innovate just enough to fulfill the standards, or even to innovate around the standards, to circumvent them.

125 become routines. In this chapter, I will argue that everyday routines were as important as standard-setting in creating safe automobiles. By examining these routines, I seek to examine another form of what Giedion called “anonymous” history.

Enforcing Safety: The National Highway Safety Bureau’s Certification

Information Requests

As other scholars have previously emphasized, after setting the initial motor vehicle safety standards, the NHSB quickly became focused on generating recalls of automobiles.22 As Daniel Patrick Moynihan’s oft-cited 1968 Report of the Secretary’s

Advisory Committee on Traffic Safety noted, “In the first 15 months after the new regulations were in force, some 3.6 million cars were recalled. During 1967, for every three vehicles manufactured, one that had been manufactured previously was being recalled.”23 By the end of model year 2008, the agency had, over the course of its history, “influenced” the auto industry to recall nearly 270 million vehicles.24

From its beginning, the agency had to create tools and procedures for pushing industry to recall dangerous and defective cars.

Throughout the history of the auto safety regulation in the United States, the

NHSB [and its successor agency, the National Highway Traffic Safety Administration

22 Jerry L. Mashaw and David L. Harfst, The Struggle for Auto Safety (Cambridge, Mass.: Harvard University Press, 1990). 23 Department of Health, Education, and Welfare, Report of the Secretary’s Advisory Committee on Traffic Safety, February 29, 1968, 5–6. 24 I take this number from a document that was provided to me by a National Highway Traffic Safety Administration staff member.

126

(NHTSA)] have begun investigations for a wide variety of reasons. Crashing cars is expensive, and the government has always relied on the automakers to do most of the upfront testing to ensure that cars comply with standards. After doing this work, the automakers file paperwork with the agency, which then looks it over before certifying each car. The question then becomes when—and how often—the agency should probe the automakers’ testing results, doing crash tests of its own.

The agency would do a certain number of random tests each year to check up on the auto companies. It would also do some tests in order to study the standards and the tests themselves and to broaden scientific, engineering, and medical knowledge about auto safety. Of course, when cars were tested for these non-enforcment-based purposes, the tests would still inevitably produce information on vehicles’ conformity or non-conformity with the standards.

Yet, the agency also carried out a number of investigations because it perceived that something was wrong, that something about a vehicle was unsafe.

Such investigations could arise for several reasons, including consumer complaints, media coverage, and industry self-reporting. Eventually, the agency also built sophisticated, computerized crash statistics systems that identified potential problem vehicles. But, in the early days, auto safety enforcement was not so advanced.

The first step of the investigation process was called a “Certification

Information Request” (CIR), in which the agency asked the auto manufacturer for more detailed information on the certification efforts than was required in the original certification efforts. In many cases, CIRs became a defensive strategy. When

127 the agency felt like others had gotten the jump on its prerogative to keep cars safe, a

CIR would be sent out promptly to ensure that the agency did not appear to be sleeping on the job.

The agency’s first CIR arose from such a case. The NHSB’s first twenty safety standards became effective on January 1, 1968.25 Twenty-two days later, the Wall

Street Journal and New York Times carried articles reporting that the February issue of Consumer Reports would include a piece on unsafe brakes in the 1968 Chevrolet

Chevelle. NHSB responded immediately, sending out a CIR to General Motors, of which Chevrolet was a division, the next day, January 24.26

Brakes were primarily covered by Standard 105, one of the wave of twenty initial standards. Standard 105 was based on two SAE Recommended Practices,

J937, “Service Brake System Performance Requirements—Passenger Car,” and

J843a, “Brake System Road Test Code—Passenger Car.”27 (“Service brakes” was an auto industry term of art for what people generally refer to as “brakes,” that is, of the non-emergency, non-parking variety.) These two standards had a complementary relationship: J843a described the test for brakes, while J937 set the

25 Federal Register, Vol. 32 (February 3, 1967), 2408. 26 Letter, John A. Dunsmoor, Acting Chief, Division of Defects Review, to John C. Bates, Director, Service Section, General Motors, January 24, 1968. 27 Federal Register, Vol. 32 (February 3, 1967), 2410–11; J937, “Service Brake System Performance Requirements—Passenger Car,” (June 1966); J843a, “Brake System Road Test Code—Passenger Car,” June 1966. Unfortunately, the SAE apparently does not have any archival records of the committees that set Recommended Practices. The Society published so many Recommended Practices during 1965 and 1966 that it would almost be unthinkable if the federal government did not somehow spur this institutional prolificacy. Perhaps, the new or updated practices emerged from the SAE’s dialectical relationship with the GAO’s already existing standards. In this scenario, the government’s standards induced research into crash science, thereby, creating new normative ideas about practices. More likely, the SAE committees were taking an activist role, building new recommendations in order to shape the government’s own standards. Since many of the NHSB’s first standards were taken from the SAE—to the degree that the Society’s committees were aiming at shaping policy—they succeeded.

128

(voluntary, industry-accepted) level at which brakes should perform in that test.

The performance tests for Standard 105 were comprised of a series of controlled braking carried out at established minimal speeds, maximum lbs. of pedal force, and minimal decelerations (measured in feet per second per second). During each braking period, the automobile also had to maintain certain handling characteristics, primarily staying within a standard twelve-foot lane of “clean, dry, smooth, Portland cement concrete pavement (or other surface with equivalent coefficient of surface friction).”28 Motor vehicle safety standards thus got quite specific; they had to do so in order to reduce ambiguity and to take into consideration as many contingencies as possible, thus ensuring reproducibility of results. In the case of the first CIR,

Consumer Reports claimed that one Chevelle had “swerved so severely that we could not keep it within a twelve-foot lane” and that another required “heavy pedal effort” and did not immediately respond to pedal depression.29 Swerving outside of the twelve-foot lane certainly violated Standard 105’s handling requirements, and the amount of pressure needed to stop the car and the slow response time, when quantified, may have violated the brake’s performance criteria.

The 1968 Chevelle case provides an example where industry could use federal standards against other interest groups, particularly consumer safety organizations. Harry F. Barr, GM’s Vice President of Engineering, responded to the

CIR on February 1, writing, “Since our results were obtained by conforming to the requirements of Federal Motor Vehicle Safety Standard 105 test procedures, it is

28 Federal Register, Vol. 32 (February 3, 1967), 2411. 29 As quoted in “Consumer Reports Says Chevelle Failed to Pass Road Tests,” Wall Street Journal, January 23, 1968.

129 apparent that Consumer Reports data were obtained in some other test manner.”30

While federal performance standards often come about largely through the efforts of safety groups, the standards can also be used against them. A consumer group may insist that a technological artifact is unsafe, but when federal standards set the definition of “safety,” these groups can lose traction in their efforts. GM was using federal standards against claims of unsafe vehicles within a month of the standards’ taking effect; the strategy was intuitive and extremely appealing.

NHSB staff members believed that GM’s certification material showed that the Chevelle and the company’s other “like-sized vehicles meet or exceed” Standard

105, but they still concluded that “tests should be observed and a visual determination made of the directional stability during 60 m.p.h. stop.”31 Staff members likely believed that the car strictly passed the test but that it may have had some stability problems, and they wanted to check the matter out. On February 14 and 15, 1968, two NHSB staff members witnessed SAE brake tests at General

Motors’ proving grounds in Milford, Michigan, the nation’s oldest automotive testing facility, established in 1924.

Since the publication of Steven Shapin’s and Simon Schaffer’s Leviathan and the Air Pump in 1985, historians of science have discussed how the ideal of a community of eyewitnesses has shaped practices within modern science.32 As historians have noted, science usually falls short of this ideal: audiences of scientific

30 Letter, Harry F. Barr, Vice President, Engineering Staff, General Motors, to John A. Dunsmoor, Acting Chief, Division of Defects, NHSB, “CIR 001—198 Chevelle,” February 1, 1968. 31 Memo, Joseph R. O’Gorman, Acting Director, Office of Performance Analysis, to Edwin Slagle, “1969 Chevelle Brakes,” February 5, 1968. 32 Shapin and Schaffer, Leviathan and the Air-Pump: Hobbes, Boyle, and the Experimental Life (Princeton, NJ: Princeton University Press, 1985).

130 eyewitnesses are rarely composed. If these historians are even close to being right, then it is surprising how quickly such audiences arise in regulatory controversies.

Regulators travel to automakers’ proving grounds. Automakers fly to the labs of private testing firms. All those involved meet and watch together. Each group watches the others doing what they ordinarily do. Additionally, testers often produce high-speed motion picture footage of the tests, which interested parties then review in slow motion. Doubtless, these interactions between the regulators and the regulated are more adversarial than those found in “normal” scientific inquiry, but science itself is certainly not free of competition and rivalries.

At the Milford Proving Grounds, GM’s staff members ran brake tests on a

1968 Chevelle, as the NHSB staff members observed. “Both vehicles, at our direction, had been equipped with new brakes shoes and linings and new drums,” an NHSB staff member later wrote. 33 They also tested a 1968 Oldsmobile F-85 to compare results. To prove their point, after running the SAE tests, GM staff repeated the tests at higher-than-required speeds. “Pedal pressures,” one of the NHSB witnesses noted,

”were well within the limits specified by the SAE standards. All stops were straight and smooth with no control problems evident.” 34 The NHSB staff members believed that GM’s brakes were “more than adequate” for the federal standards. Moreover, they placed the root of the problem in practices at Consumer Reports: “In the view of the foregoing, I can only assume that Consumer Reports made tests on the two cars

33 Memo, John A. Dunsmoor to George C. Nield, “Visit to GM’s Milford Proving Ground to Observe Chevelle Braking Tests,” February 21, 1968. 34 Ibid.

131 without proper conditioning of the brakes. Possibly, too, the brakes could have been partially glazed before the . . . tests were performed.” 35

This conclusion addressed an issue that, to this day, remains fundamentally unresolved in auto regulation. The SAE tests called for new brake shoes, linings, and drums because the goal was to test a car’s braking system when new, not necessarily how it would perform later after it had accumulated substantial mileage.

This stipulation especially made sense with brakes and some other systems, which everyone knew would require maintenance at some point. Moreover, driving styles greatly influenced the way in which many auto components, including brakes, degraded. Someone who “rode” their brakes would wear out parts faster than someone who looked ahead and took their foot off of the accelerator earlier.

Additionally, geography, particularly hills and mountains, played a substantial role.

The SAE tests also specified pre-test “burnishing” stops, which reduced any glaze that might hamper brake performance.

All of this was potentially troubling to auto safety advocates, and with some justification. While all brakes inevitably require maintenance, they usually do so only after accruing tens of thousands of miles, not after only a few thousand miles.

On February 9, Consumer Reports sent a letter to GM, which promptly sent the note on to the NHSB. As the letter stated at one point, “All of our test vehicles are driven a minimum of 2000 miles in normal service before tests commence.”36 A GM or NHSB staff member underlined this statement and jotted a question mark in the margin.

35 Ibid. 36 Robert D. Knoll, Head, Auto Division, Consumers Union of U.S. Inc., to J. G. Musser, Chevrolet Motor Division, General Motors Corporation, February 9, 1968.

132

(Of course, who wrote the question mark is of considerable interest. If GM wrote it, the question mark was a rhetorical endeavor.) The letter also mentioned, “No visual inspection of brake linings is made either before or after the test.” Consumer

Reports had a different and more activist conception of auto safety than the SAE did.

The SAE tests displayed the cars in the best light; Consumer Reports was trying to capture how cars performed in more normal use. For instance, some cars, for various reasons, might foul their brakes more quickly than others. This speedier degradation was, therefore, a safety issue. For this reason, some consumer advocates have claimed that the pre-production certification systems were misguided. What mattered was testing cars that were “in-use.”37

In the end, NHSB staff members believed that the 1968 Chevelle was within the bounds of Standard 105. Over a year later, the Automotive News, a trade newspaper whose editorial slant often sided with the auto industry, ran a story, in which it claimed that the NHSB stated that Consumer Reports made inaccurate claims in 1968 about a few vehicles, including the Chevelle. Consumer Reports replied to the Automotive News, stating that the agency had neither decided against

General Motors nor faulted Consumer Reports.38 Moreover, Consumer Reports claimed that the 1969 Chevelle’s brake systems were substantially revamped,

37 As I will discuss in the Epilogue, after beginning to tour and study automotive emission control programs in other nations—primarily Australia and Sweden—Eric Stork, the then former head of the EPA’s automotive emissions control program, called into question the value of certification programs. In his eyes, it made more sense to instantiate draconian fines for violating standards and then to test cars that had accrued miles “in use.” Producers would be under a greater burden to make sure that their technological systems worked for a considerable period of time, and regulators and producers would have to put fewer resources, both financial and personal, into the upfront certification process, which might not matter in the long haul anyway. 38 Robert J. Smith, Assistant Director, Consumers Union of U.S. Inc., to B. J. Wemboff, Publisher, Automotive News, July 28, 1969.

133 probably because of their bad publicity. Private and non-governmental groups continued to play an important role in both changing and conserving the ways of the auto industry.

If investigations had proceeded beyond where the Chevelle case left off, federal regulators may have attempted to force a recall. Yet, automakers complete the vast majority of recalls without NHSB/NHTSA ever having entered into formal enforcement procedures. For this reason, NHTSA continues to track recalls that it has “influenced,” or had some hand in, since so few recalls are actually forced. Even

Fiat recalled its vehicles under its own volition, though the NHSB would probably have induced a recall if the company had not taken that step. The NHSB’s investigation into Fiat’s vehicles went much deeper than it had in the Chevelle case, however.

The Fiat Case

The NHSB’s investigation into Fiat’s 850 Sedan and Coupe, which ran from late-1969 to early-1971, was, to that date, the agency’s largest defect investigation.

It also won the agency an unprecedented civil penalty against an automaker.

Ironically, the standard that the NHSB claimed Fiat had violated, Standard 204, had been one of the least controversial of all the agency’s standards. Technology existed to meet, and exceed, the standard’s performance requirements, so it should not have challenged the capabilities of any firm in the industry. Yet, the standard seemed to challenge Fiat. Thus, before examining the Fiat case, the history of Standard 204 and the technology that answered its requirements demand our attention.

134

Building Standard 204

Standard 204—Steering Control Rearward Displacement—was one of the first wave of standards promulgated in 1967 along with the braking standard,

Standard 105, and the interior occupant protection standard, Standard 201

(described in the last chapter). Standard 204 limited the distance that the steering wheel could “displace” backwards towards the driver’s body in a front-end collision.

The basic test for Standard 204 involved, using whatever means, driving a car into an immobile wall at thirty miles per hour and measuring the rearward displacement of the steering wheel “relative to an undisturbed point in the vehicle.”39 Yet, there was no established means of making this measurement, nor did Standard 204 specify one. Some companies and groups relied primarily on onboard cameras (that is, ones built into the interior of the vehicle) to examine how the steering column moved in reference to a grid. A number of problems attended this method, including parallax distortion, issues with camera lenses, and movement of the camera’s mounting. Another method, however, was quickly gaining predominance in testing circles. A telescoping rod was mounted to the steering wheel and some other point in the rear of the car. (The NHSB typically preferred the platform between the rear seats and the rear window.) These rods contained a component, known as a scriber, that visibly scratched the interior tube, which was marked like a ruler.

39 Federal Register, Vol. 32 (February 3, 1967), 2414.

135

Both of these testing procedures, which were often used in concert, attempted to address a central issue, namely that the standard demanded a

“dynamic measurement” during the crash test. “Dynamic” here was opposed to

“static,” or stationary. The problem was this: the simplest way to measure rearward displacement would be to measure the steering wheel’s distance from some other part of the car, then to crash the car, then to measure the distance between those two points again, before calculating the difference between the pretest and posttest measurements. Doing this would result in a static measurement, calculating how far the steering column had been displaced when all was said and done. But metal has elastic properties. Therefore, during the crash, when the car was undergoing the most dire forces, the steering wheel projected backwards further than where it ended up. Consequently, engineers had to figure out a way to measure the rearward displacement while the car was crashing. The established practice was to take both the static and dynamic measurements—the latter always being greater than the former.

Standard 204 was based on a single General Services Administration (GSA) standard—No. 515/4a—as was Standard 204’s sister standard, Standard 203—

Impact Protection for Driver from the Steering Control System.40 Neither Standard

203 nor Standard 204 was controversial. The former received 117 pages of submissions from automakers, while the latter received 95 pages; neither

40 As described in Chapter Two, after passage of the “Roberts Act” in 1964 (P. L. 88–515), the General Services Administration began setting safety standards for automobiles that the government purchased.

136 approached Standard 201’s behemoth 302 pages.41 Most of the automakers’ remarks on Standard 204 centered on the agency’s attempt to tighten the rearward displacement limit from its GSA-mandated level of five inches to three inches.

NHTSA employees justified this increased stringency, as they wrote, on the “basis of a consensus that the lesser value would be preferable.”42 If limiting rearward displacement to five inches was good, three inches would be even better.

Automakers protested the more stringent standard not because it was difficult to accomplish (it was not) but because they had already “redesigned the steering columns to meet the original 5-inch requirement” and re-redesigning the column would involve expensive and time-consuming retooling.43

More importantly, the automakers and regulators worried nearly as much about steering columns that displaced too little as they did about ones that displaced too much. Model year 1967, the first year that the GSA standards from the

Roberts Act came into effect, was also the first model year that saw some automakers adopt “energy-absorbing steering columns,” a technology upon which many people hung high hopes. (Indeed, in 1967, automakers, turning regulatory necessities into marketing opportunities, touted a number of “energy-absorbing” features on their cars.) When talking about the prospects of the steering-columns,

NHSB staff even suggested, “Clearly, if the trend continued, a breakthrough was at

41 United States Department of Commerce, National Traffic Safety Agency, “Report on the Development of the Initial Federal Motor Vehicle Safety Standards Issued January 31, 1967,” March 17, 1967, Microfilm copy of selected subject correspondence relating to motor vehicle safety standards, 1966-70, General records, Records of the National Highway Traffic Safety Administration 1966-91, Record Group 416, National Archives at College Park, MD, 58, 65, 73. Cited hereafter as, “Report on the Development of Initial Motor Vehicle Safety Standards.” 42 Ibid., 73. 43 This is some of the best evidence that the federal government’s efforts to shape vehicle design by acting as a consumer—via the Roberts Act—had some influence. Ibid., 74.

137 hand.”44 Although similar steering wheels remain a part of cars today, they did not generate the radical change originally predicted.

Standard 203 addressed the energy-absorbing aspect of steering columns. It relied on SAE Recommended Practice J944, “Steering Wheel Assembly Laboratory

Test Procedure.”45 Unlike the test used in Standard 204 but like the one used in

Standard 201, J944 was based on a lab bench test, not a frontal crash test. Engineers would use a propulsion system to slam a body form (much like Standard 201’s head form) against the steering structure at fifteen miles per hour. The standard limited the acceptable forces that the body form could experience to 2,500 lbs. Probably in order to avoid accusations that it was setting “design standards,” the NHSB stipulated that any car that did not meet the 2,500 lbs. standard could be fitted with a seatbelt of the shoulder harness variety. This way the agency was not mandating a safety system, especially given that General Motors manufactured the primary energy-absorbing steering column on the market.

Regulators likened the energy-absorbing steering columns to a “fire net,” arguing that “there is a limit to the give that can be allowed in a fire net without the person ‘bottoming out’ onto the sidewalk, with fatal or serious consequences.”

Standard 203, therefore, required the steering column to project backwards to some degree. As an NHSB report put it, as evidenced by the “efficacy of the energy absorbing steering system . . . it became obvious that designers should be given the opportunity to design devices to use rearward displacement . . . to provide a longer

44 “Report on the Development of Initial Motor Vehicle Safety Standards,” 69. 45 Federal Register, Vol. 32 (February 3, 1967), 2414.

138 forward decelerative [sic] distance.”46 Similarly, if the steering column collapsed too much in its effort to absorb the energy, the driver would be thrown against the dashboard, a fate just as bad as, if not worse than, colliding with the steering wheel.

Thus, a balance had to be struck between Standard 203 and Standard 204.

While regulators were excited by the prospects of energy-absorbing steering columns, no one had any demonstrable knowledge about the technology’s effectiveness in the “real world,” that is, outside the laboratory. Standards 203 and

204 were two of the standards that were influenced by quickly moving developments in crash science. Since energy-absorbing steering columns had just moved into cars in the 1967 model year, data on their performance from crash investigation sites did not exist. This lack of information would soon be rectified.

In October 1966, in the middle of NHSB’s first standards-making process, a thirty-three year old man in the Detroit area drove his “speeding car,” a 1967

American Motors Corporation (AMC) Ambassador Hardtop, at 50–60 miles per hour into the rear of a “stationary, heavily loaded semitrailer with such great force that the roof of the car was nearly ripped off.”47 The semitrailer contained 32,000 lbs. of cargo, yet the impact threw the trailer more than eight feet forward. The man was admitted to the hospital with “multiple facial lacerations,” but outside of a few bruises, he was unscathed. An AMC staff member who investigated the crash later noted that the hospital staff believed that the man had experienced a “typical” crash.

46 “Report on the Development of Initial Motor Vehicle Safety Standards,” 75.

47 Harlan L. Hanson, “Energy Absorbing Steering Column: A Case History,” Proceedings of the Tenth Stapp Car Crash Conference (November 8–9, 1966), 7. The fact that this paper was presented at a conference that was held at most a month after the crash is a testament to the great interest in this technology.

139

As the AMC staff member wrote, “By typical, I mean perhaps an unbelted front passenger whose head had partially penetrated the windshield of an automobile in a relatively minor collision.”48 In other words, the hospital staff had no inkling of the kind of vicious forces produced in the crash and figured that it had been a relatively minor affair. Although the man required over three hours of facial surgery, a plastic surgeon involved in the operation claimed, “Had the windshield been of the pre-

1966 car model design . . . the lacerations would undoubtedly have been very deep and greatly disfiguring.”49 Since his most prominent bruise was one caused by a seatbelt, at least three safety technologies that the government induced—an energy- absorbing steering column, a seatbelt, and safety glass—played a role in saving the man’s life.

The injured man must have found what followed his release from the hospital to be a truly bizarre experience. Though the hospital staff may have thought that the auto crash and the man’s injuries were typical, staff members at

AMC knew better. When the company heard reports of the crash, staff members quickly realized that the crash was, in effect, a real world test of the new energy- absorbing steering column. Staff members of both AMC and the federal government rapidly descended on the man’s home in suburban Detroit. Scientists and doctors examined him, taking measurements of his wounds. Engineers also examined the crashed car. Other than his facial lacerations, the bruise from his seatbelt, and a few other minor contusions, the man was uninjured. When he complained of chest pains a week later, doctors attributed it to a case of psychosomatic symptoms induced by

48 Ibid. 49 Ibid., 12–13.

140 being surrounded by doctors so often. As the AMC staff member wrote, “It was the opinion of at least two doctors that this chest pain might have been more psychological in nature than actual, since the patient was frequently questioned by not only the doctors but engineers as well with regard to the presence of such pain.”50 The man was a guinea pig, surrounded by experts who poked, prodded, and palpated him.

The AMC staff member who wrote up the company’s report on the crash tried to temper his enthusiasm for the technology’s promises by hedging: a “single accident has no statistical significance.” But his eagerness would not be deterred. As he wrote, “It is apparent that the laboratory performance of the energy-absorbing steering column has been completely verified in the field.”51 In the eyes of both industry and the government, the crash and the man’s lack of serious injury were near-miraculous.

Additionally, as an NHSB report claimed, “By January, a medical research group in Los Angeles, by studying a series of crashes of vehicles so equipped, had developed the needed information and this served as the basis of the performance requirements incorporated in the standard issued January 31.”52 These researchers were primarily NHSB Deputy Director Robert Brenner’s former colleagues at UCLA, the most important one being Dr. Alan M. Nahum, the director of the Vehicle Trauma

Research Group at UCLA’s School of Medicine and Assistant Professor of Surgery. By

January 4, Nahum had studied seventeen crashes that involved the energy-

50 Ibid., 12. 51 Both quotations, Ibid., 13. 52 “Report on the Development of Initial Motor Vehicle Safety Standards,” v.

141 absorbing steering columns and “determined that none of the drivers had sustained significant chest injuries.”53 Brenner’s connections with this community continued to buttress his causes and proposals.

At a deeper level, the energy-absorbing steering columns brought up a long- standing debate in crash science, which related to one of the NHSB’s central questions, “What, in pounds, should [Standard 203] set as the limit to be approached but not exceeded?” Since the 1950s, John Stapp had carried out his human volunteer crash studies at Holloman Air Force Base in Alamogordo, New Mexico.

Stapp claimed that the body could safely survive crashes without grave injury as long as the force it experienced remained below 40 g’s, which amounted to “very transient loads of several thousand pounds.”54 The Wayne State studies carried out by Lawrence M. Patrick and others (described in the last chapter) created a much lower estimate of human tolerance to force, somewhere around 1,000 lbs. The issue remained fundamentally unsettled within the crash science community.

In January, William Haddon told General Motors about Nahum’s studies in

California. The company immediately contacted Nahum and began working and sharing information with him. As I will argue at greater length in Chapter Five, regulation often forged novel connections between previously-unconnected communities of inquiry and, thereby, produced new scientific knowledge.55 The

53 Ibid., 70. 54 Ibid., 67. 55 Margaret Taylor (and her co-authors) have written about how regulation fosters knowledge- production and information-sharing in Margaret R. Taylor, “The Influence of Government Actions on Innovative Activities in the Development of Environmental Technologies to Control Sulfur Dioxide Emissions from Stationary Sources” (PhD diss., Carnegie Mellon University, 2001); Margaret R.

142 chief question that Nahum and GM probed was one that had to be answered before the standard’s limit of how much force a person should experience could be set: they were trying to figure out how much resistance the steering column was absorbing in the first place. Even this datum was unknown. Researchers took what was known about the kinds of wounds people experienced at different levels of force and then examined Nahum’s data on the kinds of wounds people were experiencing at different speeds with the energy-absorbing steering columns. They then estimated how much energy the steering mechanisms were absorbing.

Earlier, in November, Haddon had asked GM to estimate this number, and the company had said that probably between 900 and 1,400 pounds were being absorbed. The NHSB thought that this resistance was “not too high” but “might prove too low to give optimum protection to those in more violent impacts.”56 Thus,

Haddon set an interim standard of 1,800 pounds. In January, however, Nahum’s and

GM’s research had borne more fruit. Nahum and Haddon met at a meeting of the insurance industry in mid-January and sat and discussed what was known. Nahum believed that, in reality, the GM steering assembly was absorbing at least fifty percent more than the company had estimated in November. Haddon and Nahum agreed that a good interim standard would set the “upper limit” at 2,500 pounds.

Ironically, this value was the threshold set by the panel that had created the GSA standard on which Standard 203 was based.

Taylor, Edward S. Rubin, and David A. Hounshell, “Regulation as the Mother of Innovation: The Case of SO2 Control,” Law & Policy, Vol. 27, No. 2 (April 2005), 348-78. 56 “Report on the Development of Initial Motor Vehicle Safety Standards,” 71.

143

At this point, the interplay between Standard 203 and Standard 204 was settled, at least for the time being. Standard 203 would require the car to have a steering column that absorbed at least 2,500 pounds or to have a shoulder harness seatbelt. Standard 204 would restrict the steering column to a rearward displacement of less than five inches, enough to allow automakers room to work with energy-absorption but not so much that the steering wheel would likely crush the driver’s chest. The NHSB, industry, and the UCLA researchers all found this solution amicable.

NHTSA’s regulations were not “technology-forcing” per se; they were not geared at generating new R&D. Instead, the agency often pushed automakers to adopt readily available, “off-the-shelf” technology. Segmented steering columns that prevented rearward displacement were a known technology. Unlike with Standard

201, the automakers did not resist the passing of Standard 204, and all of the companies, except Fiat, passed Standard 204 tests. What NHTSA found again and again was that, for various reasons, the automakers did not adopt best available technology. The agency’s routines, checking random vehicles and investigating reported defects, drove manufacturers to put more current technologies in their cars.

Scholars studying the automobile have long focused on the “productivity dilemma,” the tension between introducing innovations in product and production and limiting change to increase output.57 We understand only incompletely what motivates companies to adopt new processes and technologies. Promises of

57 William J. Abernathy, The Productivity Dilemma: Roadblock to Innovation in the Automobile Industry (Baltimore: Johns Hopkins University Press, 1978).

144 efficiency for the producer do not provide enough explanation. Even with such promises waiting constantly to be harvested, innovation is uneven and lumpy.

Technological change partially explains some developments. For instance, automakers had experimented with and used aluminum engines since 1895, but domestic US manufacturers began to work regularly with aluminum engines only in the late 1950s and 1960s after significant innovations in aluminum production and casting around the time of World War II.58 But as the social constructionists have shown, technology does not change on its own. By pointing to technical change, we simply open more questions about why particular innovations came when they did.

This leads to the question, in failing to adopt safety technologies, were firms simply

“externalizing costs” as the anemic language of economists would put it, forcing consumers to assume the costs and risks of the automobile? Whatever the case, the

NHSB’s standards pushed companies to adopt already-existing safety technologies.

The standards themselves did not “force” automakers to invent, develop, and adopt de novo safety technologies.

Investigating Fiat

The unplanned testing of the Fiat 850 Coupe arose from a long process. In

March 1969, Fiat contacted the NHSB, stating that the company’s 1968 model year

850 Sedan might have a “conformity problem” with Standard 204.59 The NHSB was not immediately concerned because Fiat was no longer importing the vehicles and

58 Abernathy, The Productivity Dilemma, 57-58 59 Memorandum, David E. Wells, Chief Counsel, to F. C. Turner, Federal Highway Administrator, through Douglas W. Toms, Director, NHSB, “Fiat Civil Penalty 850 Sedan and Coupe,” March 3, 1970, 1.

145 the company said that it would begin replacing the non-conforming steering wheels through a recall program. Fiat never said that the steering wheels completely violated Standard 204; rather, its representatives suggested that the steering wheels came too close to the standard’s threshold and infringed upon the company’s own sense of safety margins. Through “subsequent questioning,” the NHSB realized that

Fiat had evidence of non-conformity going back at least a year before the company came forward to NHSB and only two months after the NHSB had formally promulgated its initial safety standards, including Standard 204.60 In March 1968, the company carried out three experimental crash studies, comparing two different steering columns, an older two-segment one and a new three-segment one that was thought to be safer. The company’s test, however, did not conform to Standard 204 and SAE J-850. Yet the test still suggested that the steering column currently being exported to the United States did not conform to the standard. In one test, the steering wheel’s rearward displacement exceeded six inches.61 In October 1968,

Fiat ran a crash test that more closely approached SAE J-850. Here too, however, the company discovered non-conformity. At that point, Fiat stopped exporting the vehicles to the United States. (Fiat indubitably sold the cars in Europe and other markets where safety rules were laxer.) Finally, five months later, in March 1969,

Fiat came to meet the NHSB. At the meeting, Fiat explained that the company had some worries about conformity and that it was going to begin a recall in order to replace the 1968 Sedan’s one-piece steel rod steering column with a safer three-

60 Letter, Lawrence R. Schneider, Acting Chief Counsel, NHSB, to L. Patrick Gray, III, Assistant Attorney General, Civil Division, Department of Justice, January 6, 1971, 4. 61 Ibid.

146 segment one, which was designed to buckle in a frontal crash. The NHSB was not pleased. The agency felt that Fiat had taken far too long to come forward, and the company’s process lacked necessary transparency.

Still, if the NHSB had not run the unplanned test of the Fiat 850 Coupe on

September, 24, 1969, the issue of Fiat’s rearward-displacing steering columns may have remained where it was in March: with the company issuing a recall and the agency slightly peeved. But the unplanned test spurred a wider investigation. Of course, this point suggests the contingency of the whole process. The investigation came about because of an accident. The point of the September test had been to crash a Fiat Sedan with the new three-segment steering column, which Fiat claimed was safer than the older single-segment one. Had NHSB’s purchasing agent sent the right car (the Sedan), NHSB may never have opened the civil penalty suit.

On October 7, 1969, Robert H. Gardner, NHSB’s chief Safety Standards

Engineer, and D. J. McDowell, one of Gardner’s staff members, filed a “trip report” about the crash tests they witnessed in California, describing the tests as well as the

Fiat Coupe’s apparent non-conformity. On the same day, Gardner completed a “CIR

Background Synopsis,” the first step of a Certification Information Request. In the synopsis, Gardner summarized the relevant information, including that the steering wheel “displaced horizontally rearward in excess of 6.75 inches.”62 The CIR was given the number 162. Both of these bureaucratic procedures, which contained a fair amount of redundant information, formed the beginning of the mundane

62 Robert Gardner, “CIR Background Synopsis,” October 7, 1969.

147 paperwork that made up the backbone of the agency’s regulatory enforcement effort.

Throughout the investigation, much of the work on the agency’s side fell to

Gardner and his staff. Regulatory investigations were and are a war of analysis and counter-analysis. On the one hand, Fiat would analyze the agency’s (and, thereby, the contractor’s) reports and produce its own account countering the agency’s findings. The agency would retort, while simultaneously tearing into Fiat’s earlier studies. Each side would make claims about proper scientific practice and what such practice demonstrated about Fiat’s cars, and each side would accuse the other of covering over difficulties to make a too-pretty picture of how engineering practice functioned. Seen from above, the torrents of paperwork resemble a kind of dance.

Playing the agency’s part in this contradance, Gardner’s role as the Safety Standards

Engineer was to examine all of the data coming in from both the contractors and

Fiat,.

Hypothetically, Gardner did not step into the realm of legal analysis, though this was easier said than done. The NHSB had its own Office of Chief Counsel, which was separate from any other agency within or outside the Department of

Transportation (DOT). NHSB required its own Chief Counsel in part because of the complexity of automotive regulation; the topic and law were simply too specialized to hand off to a general legal office in the DOT. In an investigation into standards violations, the Chief Counsel was to examine the legal aspects of the case, while

Gardner’s office handled the engineering. In reality, however, in auto regulation,

148 law and engineering were coiled together like strands of a rope.63 Statutory questions, such as whether an automaker had exercised “due care” in its certification efforts, simply had little meaning outside of established engineering practice.

Gardner also composed a draft of a CIR letter to Fiat on October 7, and from the list of concurrences (i.e., the “buck slip”) in the archive, one can see that the draft letter marched its way around the agency for the next few weeks as different offices examined it. Finally, the letter was tweaked ever so slightly and sent to Fiat on

October 29.64 This was paper-pushing at its finest.

Also on October 29, Gardner and McDowell flew out west again.65 That day they witnessed a crash test of Ford Maverick at another contractor, Dynamic

Science, in Phoenix, Arizona. Then they moved on to California, back to Digitek, where, on October 31, they crashed another Fiat Coupe. This time the Coupe’s static displacement was 7.53 inches. Its maximum dynamic displacement was 9.21 inches, nearly double the standard’s maximum distance. The agency had a real case building against Fiat. While on the trip, Gardner and McDowell did additional legwork, trying to locate a new or used Fiat 850 Sedan, the car Digitek was originally

63 The notion of the co-evolution, or co-production, between policy on the one hand and science and technology on the other, as put forward by Sheila Jasanoff and others, has strong resonance with this point. In this Fiat case, policy, engineering practices, and the technological possibilities of the day all mutually affected each other. For other applications of the idea of co-production, see the essays in Jasanoff, ed., States of Knowledge: The Co-Production of Science and the Social Order (New York: Routledge, 2004). 64 Letter, Robert Brenner, Acting Director, to V. A. Garibaldi, President, Fiat Motor Company, “Subject: Fiat Model 850 Coupe,” October 29, 1969. 65 Robert H. Gardner, Safety Standards Engineer, to Chief, Validation Division and Chief, Vehicles Branch, “Trip Report, October 29-31, 1969, to Dynamic Science, Phoenix, Arizona, and Digitek Corporation, Los Angeles, California,” November 24, 1969, 1.

149 supposed to crash.66 They found a used model with a little over 14,000 miles on the odometer and with the requisite one-piece steering column at a dealership in Mesa,

Arizona. Dynamic Science purchased and picked up the vehicle to crash test it later.

(At some point, Dynamic Science crashed tested this Fiat Sedan, and it too was found to be in violation of Standard 204. Yet, for whatever reasons, Dynamic Science’s crash test did not make its way into the archival record of the Fiat case.)

On November 5, Digitek produced its final report on the first Coupe test

(September 24), though Gardner did not sign off on it until November 20.67 The report testifies to the wide variety of technological and organizational resources needed to pull off any single crash test. Though Digitek carried out the tests, another company, Chrondek Electronics, was responsible for calibrating the instruments, while a third firm weighed the cars on an industrial scale. The tests required the work of many different people with many different skills.

Digitek’s test procedure for Standard 204 had sixty-five steps divided into preparation, impact, and “post-impact inspection and data collection” phases.

Preparation involved such tasks as removing seats and other features and installing the camera rig and other testing equipment. Pieces were removed in order to keep the cars as close to their delivered weight as possible, and the effort was largely successful. Crash cars weighed less than 150 lbs. more than their original weight, a nominal difference considering that any car crashing in “reality” would have a driver

66 Ibid., 2. 67 Digitek Corporation, “Vehicle Test Report FMVSS 204 1969 Fiat Model 850 VIN 0242771,” November 5, 1969. It’s unclear what sort of bureaucratic rigmarole the report had to go through before Gardner could sign off on it, but he likely had to ensure that the report was complete, in proper order, and that it generally fulfilled Digitek’s contractual obligations.

150 and possibly passengers. Yet Fiat would later point to this weight difference as something that made Digitek’s tests harsher than the company’s own.

Digitek’s first report was the first quasi-official document that chronicled

Fiat’s non-conformity for what would become the public record. To be a democratic institution, the NHSB had to do its work transparently, and Digitek’s report was an exercise in the rhetoric of transparency and an expression of the older tradition of documenting scientific experiments in order to create “virtual witnesses.”68 The

Digitek report reconstructed each stage of the test and found, ultimately, that Fiat’s

850 Coupe had a static displacement of 7.05 inches and a dynamic displacement of

8.47 inches.69

All of Digitek’s measurements were “adjusted,” meaning that the company used mathematical formulas to transform the test’s raw data. Fiat also made mathematical adjustments, though for completely different reasons than Digitek.

These various methods for “adjusting” data lay at the heart of the disagreement between Fiat and the NHSB. Digitek actually devised its adjustment methods

68 The term “virtual witnesses” comes from Steven Shapin’s and Simon Schaffer’s Leviathan and the Air Pump: Hobbes, Boyle, and the Experimental Life (Princeton: Princeton University Press), esp. 60– 65. In this work, Shapin and Schaffer describe how scientists used scientific texts as “literary technologies” to make “virtual witnesses” out of people who could not be immediately present at the experiment. As they note, “The technology of virtual witnessing involves the production in the reader’s mind of such an image of an experimental scene as obviates the necessity for either direct witness or replication.” (60) Of course, the question of whether crash tests could be replicated, or repeated, was the very issue at hand in the Fiat investigation. Digitek also included photographs of the test instruments and pre- and post-crash automobile to increase the evocativeness and credibility of the text. On the role of transparent science in democracy, see Sheila Jasanoff, Designs on Nature: Science and Democracy in Europe and the United States (Princeton: Princeton University Press, 2007); Yaron Ezrahi, The Descent of Icarus: Science and the Transformation of Contemporary Democracy (Cambridge, Mass.: Harvard University Press, 1990); Jonathan Kahn, Budgeting Democracy: State Building and Citizenship in America, 1890–1920 (Ithaca: Cornell University Press, 1997); Mark B. Brown, Science in Democracy: Expertise, Institutions, and Representation (Cambridge, Mass.: The MIT Press, 2009). 69 Both measurements had margins of error of plus or minus .04 inches. Digitek Corporation, “Vehicle Test Report FMVSS 204 1969 Fiat Model 850 VIN 0242771,” November 5, 1969, page 2-1.

151 because of how poorly the Fiat Coupe faired in the frontal crash test. When most

American cars were crashed, the steering wheel moved backwards along a horizontal trajectory. This wasn’t the case with the Fiat cars, whose steering wheels also moved upwards at a gruesome angle, not just backwards. Part of this difference arose from the cars’ rear engine configuration. In front engine cars, a good deal of material sat between the passenger compartment and the crash wall. Fiat’s cars had considerably less such substance. Therefore, in a crash test, the Fiat steering column itself quickly struck the wall, propelling the steering wheel—and, potentially, the driver’s head—skyward. Indeed, the steering wheel moved so far upwards that it left the frame of the onboard camera during the dynamic displacement. It also rendered the ruled telescoping scratch tube data inaccurate.

Digitek created a trigonometric method for adjusting these data, in principle compensating for the upward displacement.

On December 3, 1969, Vincent Garibaldi, Fiat’s U.S. Representative, responded to the CIR in a letter to Robert Brenner, who by that time had become

NHSB’s Acting Director. Garibaldi led Fiat’s response to the NHSB throughout the investigation.70 Witnessing the success of the VW Beetle in the late 1950s, Fiat had decided that its cars might succeed in the US and made a major investment in building retail capacity there.71 Garibaldi directed Fiat’s move into the US from the beginning. Upon his arrival in the United States, Garibaldi quickly became a mover

70 Letter, Vincent Garibaldi, US Representative, Fiat Motor Company, to Robert Brenner, Acting Director, “Subject: Fiat Model 850 Coupe,” December 3, 1969. 71 Federico Barbiellini Amidei, “Italian Investments in the United States—Contributions to a History,” available at www.unige.ch/ses/istec/EBHA2007/papers/BarbielliniGoldstein.pdf (Last accessed March 10, 2011).

152 and shaker in “foreign” car circles. A picture from the early 1960s shows him shaking hands with John F. Kennedy, handing him a $100,000 check for his presidential library.

The CIR had posed seven questions, or requests for information, to Fiat, and

Garibaldi’s letter responded to the each of the questions in turn, before attaching some supporting materials. Right away, there was a wide divergence between the kind of information the NHSB wanted and the kind of information that Fiat provided. For instance, in the first of seven information requests, the NHSB asked for all “test reports” related to the Fiat 850 Coupe from model years 1968, 1969, and

1970.72 Garibaldi, however, only gave a synopsis of the tests in his letter and then attached materials that summarized them in more detail. What NHSB wanted was the raw data from the tests. Throughout the investigation, the divide between the

NHSB’s desires and Fiat’s actions—and inactions—would persistently irritate the agency. From today’s vantage point, it is difficult to determine how much of the discrepancy between NHSB’s requests and Fiat’s actions arose from language barriers or from selective ignorance on Fiat’s part. As time went on and the NHSB continued asking for raw data, agency staff members began to wonder whether Fiat simply did not keep organized test data or, more likely, whether the company was withholding raw test data because it believed the information would be incriminating.

Garibaldi’s letter introduced a number of themes that would comprise Fiat’s defense throughout the period. The NHSB staff members wondered why Fiat had

72 Letter, Robert Brenner, Acting Director, to V. A. Garibaldi, President, Fiat Motor Company, “Subject: Fiat Model 850 Coupe,” October 29, 1969, 1.

153 not contacted the agency earlier. Garibaldi said that the company decided not to contact the agency after the earlier, troubling tests because the company believed that cars were still within the acceptable limit and because the company was in the process of moving to the three-segment steering column anyway. In July 1969, Fiat ran a crash test using, for the first time, the telescopic rod method that the NHSB endorsed on the pre-three-segment columns.73 The telescopic rod-based test produced a rearward displacement of 4.52 inches, which made Fiat uncomfortable because the results were too close to the legal limit but which the company attributed to a higher than required crashing speed.

Here, Garibaldi announced another theme: though Fiat was uncomfortable with the test results, it still believed that the cars were in conformity. As Garibaldi wrote, “As explained, however, we are not entirely satisfied with the margin of safety inherent” in the columns made before the company moved to three-segment ones. So, the company was beginning a recall of 1968 and 1969 models to install the new steering columns. But Garibaldi emphasized, “This recall is voluntary, and it is not an admission that the cars are not in conformity with Standard 204.”74

Rhetorically, such claims worked towards two goals, both of which revolved around casting Fiat as a moral actor. First, the company claimed that it believed the cars were in conformity, which meant that it had not purposely hidden the cars’ dangers.

This claim was particularly important because the Traffic Safety Act contained a

“due care” clause. If a company had exercised “due care” in certifying its cars but

73 Letter, Vincent Garibaldi, US Representative, Fiat Motor Company, to Robert Brenner, Acting Director, “Subject: Fiat Model 850 Coupe,” December 3, 1969, 2. 74 Ibid., 7.

154 had made an error inadvertently, the NHSB was to make allowances. Thus, Fiat was claiming that it had exercised “due care” through its tests. Second, Fiat claimed that—though its cars conformed to the standards—the company was initiating a recall anyway because, as one recall letter put it, “under our policy of constantly developing new equipment and devices for improving the safety of our automobiles, we have developed a new steering shaft and column.”75

Of course, the agency believed that red flags should have been going off after the Fiat’s tests, and Fiat knew that the agency felt this way. Garibaldi attempted to explain the company’s inaction: “We had intended to make immediate additional tests for confirmation, but these had to be suspended during the month of August due to the closing of the plant and from September on because of the grave difficulties arising from the strikes, that are still in progress.”76 Fiat’s strikes were real and very disruptive to all of the company’s activities; they also became a rhetorical crutch upon which the firm repeatedly fell back. The strikes explained— and excused—nearly everything.

Garibaldi’s letter also tipped the company’s hand on what would become the central issue of the investigation—the company’s testing grounds were very different than the one called for in Standard 204’s central document, SAE J850. In the process of the investigation, NHTSA employees toured Fiat’s “makeshift” testing grounds in Italy. They were makeshift because the testing grounds were in actuality the private Fiat airport that housed the company’s jets. The company would plan

75 The letter, dated February 16, 1970, which was sent to the Federal Highway Administration, was addressed “Dear FIAT Customer.” 76 Ibid., 2.

155 crash tests and set up a temporary crash wall using large concrete blocks. After the tests, it would take the wall back down. Yet, it would be difficult, if not impossible, to make the wall called for by Standard 204 temporary. The SAE Recommended

Practice J 850, on which Standard 204 was partly based, called for the crash barrier to be a 10 x 5 x 2 ft. concrete wall, which was to be back by 200,000 lbs. of compacted earth.77 Fiat’s lawyers argued that, although the airport “is an excellent test area for all kinds of crash tests” . . . “the use of a permanently fixed barrier is precluded because a nearby runway must be used for aircraft.”78 Instead, Fiat built a temporary wall out of a number of 2.63 ft concrete cubes. NHTSA employees quickly hypothesized that the Fiat Coupe had passed the company’s own crash tests because the concrete cubes were shifting, thereby, absorbing some of the car’s kinetic energy. In fact, some of the pictures that Fiat sent of the crash tests showed that the cubes had moved.

In the letter, Garibaldi also initiated the opening moves of Fiat’s attack on

Standard 204’s rationality. Garibaldi did not go into great detail here. Nor did he attack the standard with any venom. Rather, he began arranging the chessboard, moving the rhetorical pieces into place, to anticipate and structure the more heated and antagonistic interactions that would follow. Fiat’s objections to Standard 204 can be boiled down to a few points. The company argued, for instance, that the onboard camera increased the car’s weight, which, thereby, increased the steering wheel’s rearward displacement. But most of the criticisms revolved around a single

77 This information is recounted, among other places, in a letter from Fiat’s American lawyers to NHSB, December 3, 1969, Office of Standards Enforcement Files, 1974-1983. 78 Ibid.

156 point: the company claimed that, during the crash, nearly every point of the car’s body deformed, or changed, in reference to every other point. If this was true, the company argued, the car lacked any “undisturbed point” from which one could accurately measure rearward displacement. Fiat’s staff members put the most work into developing this line of attack, and it formed a central battle in the investigation.

The lack of any “undisturbed point,” Fiat asserted, undermined any measurement system that was inside the car. So, for instance, regulators could not rely on the onboard camera in the test because the camera’s anchor point would inevitably change. This problem could not be solved in any way. As Garibaldi argued,

“[Camera] supports that are too weak cause oscillation of the camera at the moment of impact, while supports that are too strong reduce the normal deformation of the

[car’s] body.”79 According to Fiat, the same line of criticism cast doubt on the telescoping tube. If no point in the car remained undistorted, then the tube’s anchor points would undo the scriber’s accuracy. More specifically, since the rear point of the tube was anchored to the panel between the back seats and the rear window, the test would produce “excessive values . . . by the deformation of the body which in turn causes a forward rotation of the panel.”80 This rear panel moved upward and forward, Fiat claimed, which meant that the panel was moving towards the steering wheel as the steering wheel was moving towards it. The end effect was an artificially high “rearward displacement.” This phenomenon, Fiat argued, was probably even greater in rear engine cars, like Fiat’s, because such cars had more weight in the rear.

79 Ibid., 3. 80 Ibid., 3–4.

157

Although the point should not be belabored or exaggerated, the notion of an

“undisturbed point” was related to the history of 20th century science, which tried, in vain, to find an objective perspective after the rise of relativity with Einstein’s theory of special relativity (1905).81 Philosophers have described this objective standpoint as the “view from nowhere” or as a God-like perspective. In questioning the availability of an undisturbed point, Fiat was saying there was no such “view from nowhere”: cameras that were outside of the car did a poor job of measuring the steering wheel’s movements. But, Fiat argued, every point in the car moved in relativity to every other point in the vehicle, so neither the onboard camera, nor the telescoping rod could be trusted to provide a sound measurement. Fiat tried to turn the debate into one over objectivity as an idealized form. The company’s claim: all is relative. And most philosophers and theorizing scientists would have ultimately agreed with Fiat—objectivity did not exist; it was a fantasy. But from the standpoint of engineering practice, the question was whether the SAE, engineers at

NHSB, and others could find a satisfactory and repeatable method, not a perfectly ideal one.82 As noted in last chapter’s discussion of the automakers’ lawyers attempts to deconstruct Standard 201, when pressed, automakers and their

81 One has to be awfully careful when discussing this phenomenon, since scientific, mathematical, and philosophical works that questioned the foundations of science, such as Schroedinger’s cat thought experiment, Einstein’s special relativity, Heisenberg’s uncertainty principle, and Godel’s incompleteness theorem, have been used to justify all kinds of dubious endeavors, including New Age spirituality, moral relativism, “energy medicine,” and experimentation with hallucinogenic drugs. All I mean to suggest is that Fiat’s arguments about the “undisturbed point” fit inside a larger historical frame. The misuse of these theories richly deserves a history of its own, though some works on “pseudo-science,” such as Alan Sokal’s, get us part of the way there. Alan Sokal and Jean Bricmont, Fashionable Nonsense: Postmodern Intellectuals’ Abuse of Science (New York: Picador, 1999); Alan Sokal, Beyond the Hoax: Science, Philosophy, and Culture (Oxford: Oxford University Press, 2008). 82 Herbert Simon originally formulated the idea of satisficing in “Rational Choice and the Structure of the Environment,” Psychological Review, Vol. 63, No. 2 (1956), 129–138.

158 representatives often endorse philosophical and scientific skepticism. In Fiat’s case, this skeptical strategy took the form of a false opposition: either knowledge is perfect, or it is nothing.

On December 9, the NHSB held an “informal meeting” with Fiat at the

Donohoe Building in Washington, D.C. Brenner, the NHSB’s acting administrator, was in Brussels, Belgium, attending at NATO conference on auto safety. In Brenner’s stead, Lawrence Schneider, the Acting Chief Counsel for Regulation, chaired the meeting. Schneider was a striking young lawyer who impressed the auto safety community by continually pushing the auto industry and getting important standards into the federal code. Before coming to the NHSB, Schneider had worked for a Washington, D.C., law firm after being employed briefly at the U.S. Department of Justice.83 Though his Bachelors and Masters degrees were awarded by Ohio institutions—Kenyon College and Western Reserve University respectively—he received his law degree from Georgetown. From the start, Schneider’s career was deeply rooted in the antagonistic legal environment of the nation’s capitol. Soon after the NHSB was reorganized into the NHTSA in 1970, Schneider became the agency’s Chief Counsel.84 Frank Armstrong, the Director of NHSB’s Office of

Performance Analysis, co-chaired the meeting with Schneider.

During this meeting, Fiat’s side of the table was led by Garibaldi and Edward

Garfield, an American lawyer, based in New York, who represented Fiat. Garfield

83 Insurance Institute for Highway Safety, “NHTSA Advocate Dead at 37,” Status Report, Vol. 9, No. 14 (July 26, 1974), 6. The vast majority of the Insurance Institute’s Status Report newsletters are available online: http://www.iihs.org/sr/default.aspx (Last accessed December 20, 2010) 84 Sadly, Schneider died at the age of only 37 in 1974. An award was created to memorialize his name not long after his death.

159 would become a major player in Fiat’s resistance to the investigation. These two men were also accompanied by three engineers from Fiat’s Italian headquarters:

Oscar Montabone, Enzo Franchini, and Alberto Negro. Montabone was the head of

Fiat’s engineering division, while Franchini was in charge of much of Fiat’s auto safety work. In 1966, Franchini had written a paper, “Crash Testing Evolution at

Fiat,” which he presented at the Automotive Engineering Congress in Detroit in

January of that year.85

Schneider opened the meeting by announcing, “the scope of our discussion has been narrowed” because “Fiat intends recalling the 850 Coupe and replacing the steering column.”86 If Fiat had not announced to the NHSB that the company would recall these vehicles, the bulk of the meeting would have doubtlessly been dedicated to this issue. Instead, the participants focused on clarifying where each group stood.

Of paramount importance to the NHSB was getting recall plans from Fiat. Brenner had instructed Schneider to get detailed recall procedures from Fiat, including target dates, logistical plans, and the number of vehicles involved. After beginning the meeting and making introductions, Schneider immediately began asking Garfield and Garibaldi about the recall.

Garibaldi, who acted as the translator for the Italians throughout the meeting, answered this question by conferring with Montabone. Montabone claimed that about 9,000 vehicles were involved and that the company hoped to get the recall underway in mid-January 1970. Fiat had stopped production of the

85 Enzo Franchini, “Crash Testing Evolution at Fiat,” SAE Paper 660165. 86 Federal Highway Administration, “Transcript of Proceedings: Informal Meeting with Fiat,” December 9, 1969, 2. Hereon, “Informal Meeting with Fiat.”

160 original steering columns in July 1969. But because of delays in the ports and Fiat’s strikes, the cars with them had probably continued being shipped to the United

States through August. All plans, however, were contingent on the strikes that had brought the company’s work to a standstill. Schneider said that the agency had read about the strikes in the news. “One thing is to read,” Franchini said in broken, proverb-like English, “and one thing is to see.”87 The strikes were a desperate situation for the company, which believed it would shed at least 3,000 cars in lost production. Yet, Garibaldi said that he hoped—and that there were some signs— that the strikes would be resolved by Christmas. If this were true, the kits necessary to complete the recall repairs would begin shipping in the middle of January 1970.

“Not all of them, of course,” Garibaldi hedged. Schneider reassured him, “We understand that.” The recall was an thoroughly negotiated process. Montabone had a few questions. Most importantly, he wanted to know if the company had to wait for all of the recall kits to arrive before the recall began. No, it would not be necessary the NHSB replied. The company could carry out recalls as kits became available.

Then, Armstrong asked Garibaldi when the company planned on informing consumers about the recall, “Would you propose, prior to [mid-January], to issue a defect notification?” Garibaldi responded offhandedly, “Somewhere around the middle of January we will start the recall campaign.” Then he seemed confused.

Schneider began saying, “The bureau’s position is of course that . . .” But Garibald cut him off and backtracked, “Did you say ‘defect notice’? No, we cannot call it a

87 “Informal Meeting with Fiat,” 4.

161 defect.”88 Schneider wondered what Garibaldi meant. If the company was instantiating a recall, would it do so “under the provisions of the Act?” Would it

“send certified letters” to consumers? “Yes,” Garfield insisted, “we’re going to do that, but we are not going to characterize the cars as defective in any way. We don’t believe they are defective. And we are not going to so characterize them in any notice we send out. We are going to characterize it for what it is . . . in conjunction with our desire to maintain the safety factors at as high a level as possible, we have decided that the compliance may be marginal [i.e., “sufficient”] on this particular model.”89

Perhaps we should puzzle over what exactly “defective” meant to the members of Fiat and the NHSB in the room that afternoon. But more important were the moral underpinnings of this discussion. If the Traffic Safety Act was a statement that automakers were morally responsible for making safe cars, then violating the standards was an ethical failure. Yet, Fiat could not own up to this truth. Fiat would continue with this strategy. It issued recall letters that eloquently evaded any culpability. Rather, by replacing the Coupes’ and Sedans’ steering columns, Fiat wanted to do the consumer a favor—in keeping with the company’s pursuit of excellence, of course.

At the meeting, the NHSB let pass the matter of whether the steering columns should be labeled “defective.” Instead, the agency’s staff members pushed Fiat for actual lab results—that is, the raw data—from the company’s certifying crash tests, not the summaries that the company had sent along with Garibaldi’s first response

88 “Informal Meeting with Fiat,” 4–5. 89 “Informal Meeting with Fiat,” 5.

162 letter. The NHSB’s Schneider lobbed a quip heavily dosed with sarcasm, which was steadily growing on the agency’s side of the table, at Garfield: “This [our request for raw data], of course, operates, Mr. Garfield, on the premise that it exists.”90

Montabone, via Garibaldi, replied that Franchini made all of the crash studies but that Montabone did not require Franchini to go into detail on the data or the instrumentation, but only to provide summaries, which is what the agency had seen.

Garibaldi, acting as an interpreter, stated, “All [Montabone] expects from Mr.

Franchini is to give him assurance that he follows the right procedures.” Montabone may have trusted Franchini implicity, but the agency didn’t, nor could it. Armstrong replied, “I would just interject here that we require the same assurance. And [we] require substantiation of that assurance by the submission of written data.”91

Ultimately, it didn’t matter whether the NHSB trusted Franchini; it trusted numbers, and it wanted to compare Fiat’s test procedures and the data produced therein to those arising from SAE’s recommended practices.

At one point, Montabone invited the NHSB staff to come tour Fiat’s labs in

Italy at the company’s expense. As Garibaldi related, Montabone wanted the agency to “follow our test to see if we are doing everything properly. We think they are. But he would like to have this clear in his mind, to check out what they are doing. This is unofficial, you understand; just as a friendly invitation if it is at all possible.”92

Armstrong had strong reservations about the idea, no doubt because it would look very bad if anyone believed that agency staff members were taking a vacation in the

90 “Informal Meeting with Fiat,” 7. 91 “Informal Meeting with Fiat,” 8. 92 “Informal Meeting with Fiat,” 11.

163 sunny Mediterranean on Fiat’s buck. He asked Fiat to put the request in writing and reassured everyone present that the agency had every intention of touring foreign manufacturers’ testing facilities when there was time. Later in the meeting, the

NHSB also warned Fiat about contacting Digitek directly. Any queries should be sent through the agency to the contractor, Armstrong and Schneider insisted. “Surely,” said Garibaldi. “Oh, sure,” said Garfield. But neither of them mentioned that Alberto

Negro had already tried to call Digitek (directly) about a month earlier.93 Negro, who was sitting seemingly mute in the meeting, didn’t bother to say anything either.

Much of the meeting was dedicated to three issues: First, Fiat laid some further groundwork for criticizing Standard 204. Montabone, who had flown in the night before and only had a few hours to look over Digitek’s report, called into question any measurements taken during the crash using the onboard camera as its mount had indubitably shifted. He characterized his criticism as “more gossip than .

. . a report,” however, as he had not had adequate time to rake over Digitek’s account in detail. Second, the NHSB questioned the timing of Fiat’s decisions. The company seemed to have delayed at several key points. Most importantly, the NHSB noted that some of the blueprints that Fiat had sent to the NHSB were dated after

November 1, 1969. This late date suggested that Fiat had not finalized its design of the new, supposedly safer three-segment steering column until much later than the company claimed. Garibaldi assured the agency’s representatives that this was not

93 Memo, Gardner to the file, “Report on Contact of Fiat S.P.A. with Compliance Testing Laboratory, Digitek Corporation,” November 21, 1969. In the memo, Gardner reported a conversation he had with Bill McKibben, the project manager at Digitek on the Fiat tests, where the issue came up. Apparently, Negro quizzed McKibben on several issues, including the test procedures and data, and “queried Mr. McKibben as to his confidence in the accuracy of the test results.” McKibben had replied that he was very confident in the results.

164 the case. Fiat’s midstream design changes, he claimed, were often urgent affairs.

Emergency drawings would be jotted down, rushed to the production floor, and finalized only later. The strikes held up this finalization process. Thus, the late date on the blueprints was merely a formality and did not relate in anyway to presence or absence of the new steering column.

Finally, the NHSB began openly criticizing Fiat’s crash barrier. This discussion was the first exchange over this issue. While the lion’s share of evidence lay on the agency’s side in the barrier issue, Fiat was equally trying to position itself strategically on the matter. Armstrong opened the discussion, “Now I would just indicate at this time that from review of your data your barrier does not meet the requirements.”94 The company was ready. “Even if it’s immovable? Even though on impact it does not move even one millimeter?” Garfield asked. Armstrong protested,

“Gentlemen, we have photographs supplied by your own company which clearly show movement.” Garibaldi and Garfield tried to argue against Armstrong, but

Armstrong tried to avoid a debate, stating that he just wanted them to know where the agency stood on the matter. Fiat wanted to state that any tests where the barriers moved were thrown out. A later document showed that there may have been some truth in this matter.95 Yet, of the 28 tests that Fiat carried out, 17 had to be discarded because of barrier moment. In the case of a different model, the Fiat

124 Spider, all four of the tests showed evidence of some movement. Still, the NHSB tried to quell the discussion for the time being.

94 “Informal Meeting with Fiat,” 11. 95 These test summaries were attached as appendices to Memo, Robert H. Gardner to Chief of Validation Division, “Review and Analysis of Fiat’s Response to CIR-162, Steering Control Rearward Displacement, FMVSS 204,” December 11, 1969.

165

Later, however, the topic again bared its face; it refused to go away. The

NHSB staff pointed out that Fiat continued increasing the barrier’s weight over the last few years and that the agency believed that the company had done this because the barrier kept moving during crashes. Yet, the company never created a barrier that even came close to weighing 200,000 lbs. as required in SAE J850. O’Gorman went even further, “It’s [been] indicated to me that you did not [create a barrier that weighed 200,000 lbs.] because you were afraid you would exceed five inches.”96

Garfield erupted, “Excuse me. If something is immovable, it’s immovable. I’m not an engineer, and I’m at a disadvantage here. I’m only a lawyer. So you have to forgive me. What you have to determine is that’s absolutely immovable.” O’Gorman rebuffed

Garfield sardonically, “Okay, I’m not a lawyer, I’m an engineer. My question to you is this: if the requirements required a specific compliance of 200,000, why didn’t you use it?” Garfield retorted, “All right. I will answer your question by asking a question: why didn’t your laboratory [that is, Digitek] use it?”97

Garfield had a point. In some ways, Fiat, via Garfield, was indicating that the artist’s rendering and the textual description of the crash barrier in SAE J 850 were works of idealism—Platonic like forms that could be approached but not captured.

Just as Fiat protested, Digitek itself did not use the exact barrier described in SAE J

850. Rather, the contractor used a reinforced concrete wall, whose face met the SAE

J 850 description, ten feet wide, five feet tall, and two feet thick. But Digitek backed the face “with additional reinforced concrete, designed into a solid buttress

96 “Informal Meeting with Fiat,” 37. 97 “Informal Meeting with Fiat, ”37.

166 structure. The total mass of the concrete structure is approximately 65,000 lbs.”98

NHSB representatives argued that Digitek’s crash barrier was “equivalent” to the

SAE barrier, while Fiat’s wasn’t. But what did “equivalent” mean? Fiat’s argument was that the SAE barrier was meant to describe a wall that would not move and, thus, did not absorb any energy. As Garfield put it, “If something is immovable, it’s immovable.”99 The question was, whose barrier and crash procedures would count?

Which banality—which set of mundane practices—would become hegemonic in the world of auto testing? Each side pushed a set of practices.

Moreover, much to the NHSB’s consternation and embarrassment, Standard

204 was not fully settled, or “closed.” It was in a state of flux, although the band of flux was rather narrow. After first publishing Standard 204, the NHSB had received a number of “inquiries” about how the standard should be properly understood, given certain contingencies and ambiguities. On June 21, 1967, only a few months after the first standards were published and a full half-year before they took hold, the NHSB published an “interpretation” of Standard 204.100 The interpretation dealt with two issues: first, the original 203 and 204 standards described a crash dummy being involved in the test, which would have been used to measure the energy experienced by the driver. The interpretation ruled that, if a manikin was involved, the testers did not have to take account of the forces that were absorbed in the dummy’s chest, which would necessarily reduce the rearward displacement. Second, the interpretation gave a mathematical means of dealing with cases where the car

98 Digitek Corporation, “Vehicle Test Report FMVSS 204 1969 Fiat Model 850 VIN 0242771,” November 5, 1969, page 4-1. 99 “Informal Meeting with Fiat,” 37. 100 Federal Register, Vol. 32 (November 5, 1969), 8808.

167 crashed into the barrier at a slightly faster speed than the requirement. If the car crashed at more than thirty miles an hour but less than thirty-three miles an hour, the subsequent results could be (mathematically) adjusted.

Later, the NHSB revised Standard 204 so that the dummy was either dropped from the test or testers would have to account for how much energy the manikin had absorbed.101 The agency also did away with the equation for adjusting results from higher crash speeds. Laboratories, both within and outside the auto industry, had attained enough control that correcting for speed deviations was no longer necessary. Moreover, Fiat had already tried to use the part of the interpretation that dealt with equations for adjusting results to justify much higher crash speeds (even though the interpretation specified that the equation was only to be used for crashes that happened between 30 and 33 miles per hour). The equation was a liability.

Finally, the agency revised the notion of the undisturbed point: the undisturbed point was defined as a point on the exterior of the automobile that was comprised by a line that ran vertically through the center of the car’s rear left wheel and a line that ran horizontally one inch below the left rear window. The NHSB decided to define this point because of the controversy with Fiat and because other automakers had made inquiries. Thus, the NHSB was in the embarrassing position of trying to enforce, and potentially seek civil penalties for, a standard that had not reached closure.

Yet, despite their differences, the NHSB and Fiat officials seemed to have left the meeting on friendly terms. Schneider, on behalf of Brenner, thanked the Fiat

101 Ibid.

168 representatives for coming to the meeting, and, as Garibaldi related, Montabone expressed his hope that he could drop into the NHSB to say hello whenever he was in the United States. “I’m sure,” Schneider said, “Mr. Armstrong and our engineers would be delighted to see you.”102 Relations would grow much frostier over the coming weeks, however.

Meanwhile, Gardner continued pushing paper. On December 11, two days after the meeting with Fiat, he sent around a memo that analyzed Garibaldi’s

December 3 letter in detail. Gardner was working quickly under considerable pressure. As Gardner noted, much of Garibaldi’s letter focused on questioning

Standard 204’s validity and, more specifically, of Digitek’s work. Gardner wrote,

“Fiat interjects numerous speculative shortcomings of . . . measurement techniques presumably to cast doubt on the validity of data” obtained through Digitek’s tests.103

He also furthered the agency’s thinking on Fiat’s crash barrier. “Fiat did not describe any means of anchoring or lashing together of their concrete cubes to obtain an integral mass,” he noted. With this notion of “integral mass,” Gardner identified something that would become a central platform of the agency’s case against Fiat. “If the cubes are not tightly secured into an integral mass, movement can take place on one or more cubes and, therefore, energy is expended through work,” which reduced the car’s deceleration rate “with consequently less destructive results.”104 As he continued, “A slight separation between the cubes

102 “Informal Meeting with Fiat,” 52. 103 Memo, Robert H. Gardner to Chief of Validation Division, “Review and Analysis of Fiat’s Response to CIR-162, Steering Control Rearward Displacement, FMVSS 204,” December 11, 1969, 2. Hereafter, “Review and Analysis of Fiat’s Response.” 104 “Review and Analysis of Fiat’s Response,” 4.

169 precludes the total mass of all the cubes being effective until work is expended in moving all of the cubes into a tight, compact mass.”105 Gardner reiterated NHSB’s basic belief that the pictures that Fiat submitted showed unacceptable movement of the blocks. He also examined how Fiat’s barrier became heavier over time

(probably because it kept moving) and how, as the barrier increased in mass, the steering columns of crash cars experienced concomitant increases in rearward projection. Finally, Gardner went further than anyone else had to that point and claimed that Fiat’s cars—that is, its entire fleet—were probably out of compliance.

Fiat’s other models, including the “124 Sedan, Coupe, and Station Wagon,” had worse test results than the 850 car line in Fiat’s original certification submission.

Thus, if Digitek’s test demonstrated nonconformity in the Fiat 850 cars, it would probably produce even worse failure in the company’s other vehicles, given the

“known wide divergence between the Bureau’s tests and those of Fiat’s.” In conclusion, Gardner made two recommendations: The NHSB should begin

“appropriate action to enforce” the Traffic Safety Act, meaning it should prepare to bring civil suit against Fiat, and it should begin crashing Fiat’s other cars.

On December 22, the Federal Highway Administration, the NHSB’s parent agency, published a press release in an attempt to shape the narrative about Fiat’s cars, a narrative that Fiat itself would try to control through its coming recall notices. As a rule, NHSB did a poor job of what would later become known as

“messaging.” This particular press release, however, was an important exception to the rule. The NHSB knew that Fiat was going to stress that its cars were not

105 “Review and Analysis of Fiat’s Response,” 4–5.

170 defective and that the company was commencing a recall because of its high safety standards. To counter this framework, F. C. Turner, the Federal Highway

Administrator, announced that Fiat’s cars had failed two tests. Moreover, Brenner wrote that the NHSB “is anxious that the company initiate the recall campaign as rapidly as circumstances permit, to lessen the chance that drivers may be injured in crashes involving these vehicles.”106 Fitting the agency’s general weakness at messaging, however, it failed to prime the interest any major daily newspapers, and none of the newspapers picked up on the press release.

After the winter holidays, on January 7, 1970, Garibaldi sent further blueprints along with additional supporting material, most of which had been prepared by Montabone and his staff members. Four annexes, or appendices, were attached to the letter—a description of the sixty-two steps involved in Fiat’s “radio- guided” crash test procedure, a letter that explained the “calibration” done at the firm’s labs, a document that outlined Fiat’s “Verification of Crash Test Performance,” and Enzo Franchini’s 1966 SAE paper, “Crash Testing Evolution at Fiat.” Unlike most automakers who used a pulley, or catapult, system to launch cars toward the crash barrier, Fiat had settled on a radio-controlled method of propulsion. The main drawback, in Franchini’s eyes, to catapult systems was that they worked best at relatively low speeds. At higher speeds, it was tough to control the car’s directionality, leading to either a glancing blow or even a wholly missed pass instead of a head-on collision. In most cases, low crash speeds were acceptable as,

106 Federal Highway Administration, Press Release, December 22, 1969.

171 studies showed, most actual crashes happened at around 25 miles per hour.107 But high-speed crashes were increasingly regular. Franchini attributed this trend to a mixture of four factors: the highway system had expanded and speed limits had increased, while the roads were increasingly filled with inexperienced, unskilled drivers, especially young people and “Sunday drivers.” In this context, crash testing cars at higher speeds became increasingly important.

In Fiat’s system, engineers affixed the test car with a number of radio- controlled devices that would manage steering and acceleration. During the crash test, the radio-guided car would be followed either by a control car or by a helicopter, in which the remote controlling technician would sit. The helicopter method was preferred when the company crashed cars into one another, because there was no risk that, if the test cars missed each other, the control cars would be struck. Fiat’s methodology, which depended on some rather primitive remote control systems, may have partly explained why the company so often crashed cars at higher-than-required speeds and then “adjusted” the results. The technology was simply too blunt to finely control the impact speed. Though Franchini tried to root

Fiat’s crash studies in changing social trends—more highway, higher speed limits, unskilled drivers, etc.— the faster than required crashes may have been a symptom of Fiat’s test system. Franchini’s paper also mentioned that SAE had been recommending standardization of practices, for instance, through the publication of

J850, in order to create comparable results across testing facilities. But the SAE had yet to recommend any standard means of propulsion, allowing groups to pick the

107 Franchini, “Crash Testing Evolution at Fiat,” 3–4.

172 method that best fit their facilities. (In reality, the means of propulsion mattered little; the real issue was ensuring that the car crashed into the barrier near the required speed.)

Yet, the subtext of Franchini’s essay was clear: Fiat’s radio-operated mode of propulsion was superior. Again, Franchini’s paper had hegemonic goals. The question was, whose propulsion methods would become standard if the industry decided to move in that direction? Two benefits fell to the firm (and its engineers) if its procedures were picked up as the industry standards. First was prestige. The firm would be seen as an industry leader, and the company’s engineers could take personal satisfaction in these laurels, especially within engineering societies like the

SAE. The second benefit was more defensive. Often, a company’s idiosyncratic engineering practices had become part of its habits, its organizational culture, its genetic code. If the company had to adopt another firm’s practices because those practices had become standard, it would interrupt the adopting firm’s routines.

Having others adopt your practices was simply much more organizationally efficient. Even as something as mundane as an SAE publication on the development of propulsion systems for crash tests had politics.108

More important than Franchini’s essay or the document that listed the sixty- two steps in the company’s crash procedure was the document that explained in detail how Fiat verified that its crash barrier blocks had not moved. All of Fiat’s submissions stressed rhetorically the company’s commitment to scientific rigor and fidelity to the standard’s spirit. The company claimed that its “test method” allowed

108 Langdon Winner, “Do Artifacts Have Politics?” Daedalus, Vol. 109, No. 1 (Winter 1980), 121–136.

173 it to “reduce the minimum risk of deviating from the requirement.” Fiat emphasized its system of checks, dry runs, and re-checks that would ensure conformity.109 But it was the document, “Verification of Crash Test Barrier Performance,” that sought to put the agency’s most pressing concern—that Fiat’s crash barrier was woefully inadequate—to rest.

Unsurprisingly, this document described Fiat’s methods for ensuring non- movement of the concrete cubes that made up the crash barrier. The process began when Fiat put the concrete cubes on the runway. Using two 10-ton cranes, the blocks were “set” in place by light impacts in order to “eliminate any gaps between block faces.”110 Technicians then placed a wooden face on the side of the cubes that would be impacted, and they applied a layer of paint at the point where the blocks met the runway surface. If the blocks moved, Fiat argued,” it will be evidenced by cracking of the paint film.”111 The paint line was not the only measure of movement, however. Around and over the stack of concrete cubes, Fiat built a metal framework, which was not attached to the cubes and (supposedly) did not shift during the crash studies. Before and after the test, Fiat ran plum lines down from this metal frame and used Vernier calipers to measure the distance of certain reference points on the barrier face and on the cubes. Ideally, there would be no difference between the pre- and post-crash caliper measurements, but this never happened. When the cubes were displaced, Fiat developed two different formulas

109 Fiat—Safety Research Laboratories, “Calibrations,” undated, accompanied Garibaldi’s letter of January 7, 1970. 110 Fiat Safety Research Laboratory, “Verification of Crash Barrier Performance,” undated, also accompanied Garibaldi’s letter of January 7, 1970. 111 Verification of Crash Barrier Performance,” 1.

174 to determine how much energy the barrier had absorbed. The first quantified the movement by examining how much energy was required for the blocks to overcome friction. Thus, if the barrier was three cubes high, the bottom cubes bore not only their own weight but also the weight of the cubes stacked on top of them. The formula then posited a friction coefficient of 1 and examined how much energy it would take to shift the cubes the distance they had moved. This methodology had several problems. Foremost, it simply posited a friction coefficient, a coefficient that was questionable. Second, as Gardner later pointed out, Fiat’s submission pictures showed that the cubes were often stacked on turf, but Fiat’s formulae took no consideration of this. Turf would have much more friction than a concrete runway, and, thus, the barriers would absorb more energy when they were set on the grass.

Moreover, vertical torque might absorb additional energy. Using this method, the company found that the barrier absorbed 148 ft.-lbs. of energy. Fiat’s second method for measuring energy absorption involved examining how much work it would take to create the displacements. The method led to a more conservative estimate. In this case, the displacement-work method suggested that the barrier had absorbed 215 ft.-lbs. of energy. Since the car weighed about 2,000 lbs and produced about 58,660 ft.-lbs. of energy in the crash, 215 ft lbs amounted to .04% energy absorption, a trivial amount, in Fiat’s view. 112 The company concluded, “We believe the barrier of 74 one-ton blocks to be more than adequate for a satisfactory evaluation within the requirements of the safety standards.”113

112 Ibid., 3–4. 113 Ibid., 4.

175

As soon as NHSB received these new materials, Gardner set to work examining them. But in the intervening time, on January 21, Garibaldi wrote the

NHSB, explaining the logistics of the company’s recall repairs. By January 10, 2,500 recall kits had arrived in the United States, and Fiat had shipped another 2,500 kits from Italy on January 21 and expected to send at least another 2,000 in early

February.114 A “technical specialist” had also arrived from Italy. Training for the repairs would work in a tiered system. The technical specialist had begun teaching

Fiat’s U.S. Service Manager and Regional Service Managers on January 13. The

Regional Service Managers were to then begin training the company’s Service

Representatives on January 22, who would in turn train the dealerships in how to complete the recall repairs on January 26. The company estimated that all training would be complete by February 10. On March 5, Fiat sent a letter to the NHSB that it had also sent to its dealerships on February 9, detailing both the paperwork, primarily the need for dealerships to get vehicles owners to sign a paper saying that they’d had the recall modifications done, and the repairs necessary for the recall.115

Fiat estimated that the modification procedures, which involved replacing both the steering column and steering wheel as well as installing two steel reinforcement plates and which took exactly thirty-four steps and twenty-six parts, would require around four hours to complete. Thus, if all of the nearly 9,000 recalled vehicles were brought in for modification, the company would put about 36,000 man-hours into the effort. Of course, this number does not include the amount of time needed to

114 Letter, Garibaldi to Francis Armstrong, “Re: Fiat 850 Coupe Steering Column Recall Campaign,” January 21, 2010. 115 Fiat, “850 Coupe Modification,” attached with additional materials to a letter from Garibaldi to Armstrong, March 5, 1970.

176 make the parts and put together the recall kits in Italy. Fiat sent regular updates to the NHSB on how the recalls were going (the company kept separate counts of the

Coupe and Sedan recalls). For instance, by March 25, 1970, the company had sent out 5,073 letters to Coupe owners, 353 of which had brought in their cars for modification, while eight Coupes had either been stolen or totally destroyed and thirteen owners were unable to be located.116

On February 6, Gardner sent out a three-page analysis of the additional materials that Fiat had submitted. In his view, this new material changed nothing whatsoever. Most of the submissions, Gardner found, described test procedures but not at the level that the NHSB needed to make any “evaluating completeness and validity.”117 Gardner also pointed out that, though the submitted materials emphasized the amount of care that went into making sure that the cubes were placed directly against each other, including using the two ten-ton cranes to tap the cubes into position, at the December 9 meeting, Alberto Negro had said that technicians used a forklift to set the cubes. Fiat may not have been even following its own protocol, regardless of how much this protocol did or did not approximate the SAE procedures. Gardner reiterated that the photos that Fiat had submitted seemed to show obvious movement.

He went on to criticize a number of other factors. For instance, the bob line procedure was not a “very accurate method.” Additionally, Fiat made erroneous assumptions about the coefficient of friction. Finally, he went after the methods

116 Letter, F. Barbacci, Assistant Technical Manager, Fiat Motor Company, Inc. to Armstrong, “Re: Fiat 850 Coupe Steering Column Recall Campaign,” March 25, 1970. 117 Gardner to Chief of the Compliance Testing Division, “Review of Additional Items Submitted by Fiat Motor Company,” February 6, 1970.

177 used to measure energy absorption. The biggest problem, Gardner believed, was that the formula took into account only the movement of the front blocks, since displacement was based on using the bob lines to measure movement of the face.

This method seemed to treat the blocks behind the front row as if it were a solid mass, but those cubes certainly moved too, meaning that they also absorbed energy.

Gardner characterized Fiat’s formulae as a “simulated barrier verification process,” an attempt to mathematically model a complex physical reaction.118 In Gardner’s view, such simulation amounted to a hill of beans. As he wrote, “Due to the many indeterminate variables, it is concluded that the only realistic approach to a possible verification of the concrete cube barrier, would be a correlation of data from similar tests conducted with the cube barrier, and of a barrier having the SAE characteristics.”119

By the beginning of March, the NHSB had heard enough. No one there was swayed by Fiat’s perseverations. On March 3, David E. Wells, the NHSB’s Chief

Counsel, sent a memo to Federal Highway Administrator F. C. Turner through

Douglas Toms, the NHSB’s newly appointed director. Toms would lead the NHSB through an important reorganization. Part of his agreement to taking charge of the agency included moving it out of the Federal Highway Administration into a position that reported directly to the Secretary of Transportation. This change would give the agency more prominence and power—prominence and power that Toms demanded. The agency also left most of its highway functions behind in the Federal

Highway Administration, though, to this day, it still retains some highway-safety-

118 Ibid., 2. 119 Ibid., 3.

178 related functions. Thus, the agency began to focus evermore solely on the safety of the automobile itself, a change that Haddon doubtless would have relished four years earlier.120 These changes, however, still lay several months off when Wells sent his memo regarding Fiat.

Wells wrote his memo with one purpose in mind—to lay out the “pro and con arguments” for bringing civil penalties against Fiat. In his memo, Wells pointed to several problems with Fiat’s certification and subsequent behavior. For instance, he criticized Fiat’s practice of crashing cars at speeds higher than those recommended in the standards and then mathematically “correcting” for higher speeds.

But then Wells pointed out that the Fiat case went to the heart of an unresolved policy issue at NHSB, namely whether the agency would seek penalties against firms that effectively turned themselves in for non-conformity. Some argued that penalizing such behavior would discourage automakers from being open with the agency, which would increase the burden on NHTSA’s own testing programs.

Others argued that “manufacturers will strive harder to assure themselves that their products meet the standards if the failure to comply will subject them to penalty.”121

Some on this side of the debate believed that voluntary reporting should be taken into account when deciding the severity of the punishment but not preclude punishment itself. Others went further. As Wells noted, “Those favoring seeking civil penalties argue further that, were the policy otherwise, manufacturers might

120 During the reorganization process, the agency received a name changed as well. No longer the National Highway Safety Bureau, it became the National Highway Traffic Safety Administration (NHTSA), the name it retains to this day. For simplicity’s and clarity’s sake, however, I will continue to refer to the agency as the NHSB for the rest of this chapter. 121 David E.Wells, Chief Counsel, NHSB, to F. C. Turner, Federal Highway Administrator, through Douglas W. Toms, Director, NHSB, “Fiat Civil Penalty 850 Sedan and Coupe,” 2.

179 design their products to barely meet the standards’ requirements and, if the items produced turned out to fail the standard, simply come in and admit the nonconformity, without fear of a civil penalty being imposed.”122

This concern brought Wells to the central moment in his argument:

It is the opinion of [the Office of Chief Council] that the Bureau should adopt a policy which neither commits it to automatically initiating civil penalty proceedings, nor bars it from initiating such proceedings, whenever a manufacturer discloses a noncompliance to it. In other words, the Bureau should treat each case on an ad hoc basis. Such a policy, in our view, would leave the Bureau with complete flexibility in noncompliance cases.123

Wells’s point here ran counter to the organizational culture of NHTSA, which typically preferred establishing procedures and rules to leaving choices up to the whims of decision-makers. But, Wells argued, conserving the decision-making element in cases of voluntary reporting of noncompliance would keep manufacturers on guard and encourage them to produce automobiles that fell well within the standards.

At the same time, the Fiat case simply went beyond this internal debate about the issue of self-reporting. As Wells wrote, “The factual conclusion appears inescapable that Fiat’s testing of these vehicles was grossly insufficient in quantity, in quality, and in timing. In both cases [i.e., with the 850 Sedan and Coupe], the company was using steering columns that were designed and first produced well before the standards were issued. Even the few tests that were conducted, both before and after the standards were effective, gave distorted results because of the

122 Ibid., 2-3. 123 Ibid., 3.

180 movement of the barrier employed.”124 Wells recommended unequivocally that the agency proceed with civil penalties against Fiat for both the 850 Sedan and Coupe.

And, on April 14, 1970, the Department of Transportation sent a letter to Vincent

Garibaldi, letting Fiat’s representative in the United States know that the NHSB was formally “considering seeking the imposition of civil penalty” and that the agency would be happy to hear anything that Fiat had to say about the matter.125 A little over a week later, on April 23, the Federal Highway Administration published a press release announcing that Fiat may have violated Standard 204. Still, no major daily newspapers picked up the story. The press release remained largely unreported.

On April 6 and 7, the NHSB’s O’Gorman and Gardner toured Digitek’s testing grounds with members of Fiat, including Enzo Franchini and Alberto Negro. Fiat had wanted to make such a visit since the investigation had begun, in part to see how the American contractors did their work. But the visit also had a critical intention. While at the testing grounds, Franchini, as Gardner recalled, “emphasized the performance variations in barrier impacts of the smaller rear engine automobiles, like the Fiat 850, and the larger American front engine frame type body.”126 Thus, one of Fiat’s many arguments about Standard 204 was that it favored cars built by manufacturers in the United States.

124 Ibid., 2. 125 Letter, James B. Minor, Acting Assistant General Counsel, Department of Transportation, to Vincent Garibaldi, April 14, 1970. 126 Memo, Safety Standards Engineer to Chief of the Testing Division, “Visitation to Compliance Testing Laboratories, Digitek Corporation and Dynamic Science – April 6 and 7, 1970,” April 25, 1970.

181

After exchanges of correspondence between Fiat and the NHSB in April, contact slowed down considerably. Statements and responses were marked by weeks and months, not days. Each side had its reason for this change. For the NHSB,

April marked a move to a new phase—building a case for a civil penalty. The civil penalty was not only largely uncharted water for the NHSB, but it also involved the arduous and careful work of building a legal case against Fiat in the off chance that the company chose to challenge the agency in court. Although Gardner’s analyses, which formed the backbone of the NHSB’s arguments against Fiat but which were meant primarily for internal consumption, involved a great deal of time, mundane labor, and calculation, they did not require the tedious, line-by-line argumentation demanded in a legal proceeding, which would be viewed by outsiders. Indeed, this difference partly explains the sarcasm and glib, biting comments in Gardner’s analyses. For Fiat’s part, the slowdown in correspondence arose, first, from its careful attempts to counter the NHSB and, second, from the company’s organizational foot-dragging that comes with such resistance.

On June 29, 1970, over two months after NHSB’s initial letter warning Fiat that the agency was considering civil penalty, Garfield replied with a twelve page letter, which dealt with a number of key issues. First, Garfield sought to clear the air, once again, over the idea that Fiat had recalled its cars because they violated the federal safety standards. He insisted on maintaining the company line, “Fiat had developed through long and costly research a steering control that . . . was far superior to any other in use . . . [and] Fiat wished to give to Fiat owners the benefit

182 of its technological and safety advances.”127 Moreover, “this was abundantly clear,”

Garfield argued, because the company had chosen to recall 1967 models, which weren’t covered by federal safety standards, as well as 1968 models, which were.

Therefore, for these and other reasons, the NHSB could not use the fact of Fiat’s recalls against the company. Second, on the one hand, the two Digitek tests could not prove that all 10,000 Fiat cars imported to the United States were outside of conformity; two tests were statistically insignificant for a population of 10,000. On the other hand, it was perfectly reasonable that “Fiat itself certified on the basis of only two tests” since “those tests were made on prototypes . . . that were used for the production cars that followed.” Since “neither the Act nor the standard prescribes the number of tests that must be made to constitute the basis for certification,” the company showed “due care” when it “tested the cars in good faith.” Finally, the Dynamic Sciences test of the 850 Sedan had to be thrown out because the contractor had used a second-hand car. (Of course, it would have been hard for the NHSB to find a new 1968 Fiat Sedan in 1969, but Garfield passed over this issue.)

On July 20, Gardner sent a memorandum examining Garfield’s letter around the NHSB. He covered Garfield’s letter point-by-point, and in the end, saw little in

Garfield’s words to threaten the NHSB’s case. For example, although the 850 Sedan that Dynamic Science crashed was not “new,” the car had always been owned by a dealership, never a consumer. Therefore, Garfield’s statement that the non-newness

127 Letter, Garfield to Minor, June 29, 1970.

183 of the car somehow rendered the test invalid seemed to be a stretch. If words could shrug, Gardner’s did.

Earlier, on July 1, Gardner had sent around a memo that analyzed, in detail,

Fiat’s two mathematical methods for estimating energy dissipation in crash tests— the first based on work used to overcome friction and the second based on the conservation of energy. Gardner had already given cursory attention to Fiat’s mathematical models in other reports, but, as the agency headed into the civil penalty phase of the investigation, administrators wanted him to dot the i’s and cross the t’s. That Fiat had sent the NHSB both methods of computation, while the firm, in reality, only used the second method (conservation of energy), Gardner saw as a rhetorical ploy. By showing that the company had opted for the tougher of the two methods, Fiat hoped to demonstrate the company’s scientific ethos and to create an image of its careful engineering methods.

In the parlance of the Traffic Safety Act, this ploy was aimed at establishing that the firm had exercised “due care.” As Gardner wrote, “It is presumed that Fiat’s intention in these submitted computations is to prove that their barrier movements did not significantly alter their steering control displacement test data.” If this indeed was Fiat’s intention, Gardner remianed unmoved. The computation based on friction was quite bad in his view, especially its specious assumptions about the coefficient of friction. Since the blocks “were placed upon sodded earth,” which

“certainly resulted in a depression in the ground,” the cubes would definitely “cause a greater resistance to movement and more energy expenditure” than if the coefficient of friction were 1 as the Fiat engineers assumed.

184

Moreover—and perhaps even more importantly—the formula presupposed that the cubes moved backwards in a straight line. In reality, they would have twisted, creating “rotational energy” and absorbing forces when the cubes impacted each other while twisting. Repeating the near-constant mantra of these analyses,

Gardner stressed that Fiat’s formula considered only the energy absorbed through the barrier’s front eight blocks overcoming friction. Cubes behind the front line moved as well, absorbing energy, though the bob line method did not account for such movements.

Fiat’s second formula was better in Gardner’s view, “more technically empirical.” But it also made spurious assumptions. Most essential, Gardner argued, the formula made too simplistic assumptions about what the “load buildup” curve looked like. In a crash test, the forces would increase from zero (just before the impact) to a maximum peak after all of the parts of the car that would crumple had done so, and the car’s rigid steel frame connected with the (ostensibly) immovable barrier. Fiat assumed that the load built up in a linear fashion and represented the forces as a “triangular pattern.” In this way, Gardner argued, “Fiat has reduced an extremely complicated force-time history curve to a simple rectangular graph.”128

This assumption would have been fine if it had been empirically derived through some study, but it wasn’t. Furthermore, in both formulae, Fiat created an aggregate measure called the “average displacement value,” which somehow (the method was not clear) added up the distance each block had moved and divided that sum by the number of blocks moved (according to the bob line measurements). These values

128 Ibid., 3.

185 were then fed into the formula to create the average displacement. As Gardner demonstrated, if slightly different but plausible numbers were substituted for Fiat’s values, the cars would have violated the 5-inch displacement standard.

The fact that Fiat used values created from these formulae to then adjust the measured rearward displacement of the steering column only proliferated the complexities and potential shortcomings of this method. In the end, Gardner summarized his point: “The mathematical computations of the barrier movement energy losses use several values which are unsupported and questionable.”129

Moreover, “The quantity of unknowns renders this dynamic action too complicated for [obtaining] rational answers to energy losses.”130 His conclusion was biting:

“The calculated kinetic energy of a moving vehicle appears to be the only unquestionable data in the computations.”131 A high school student learning basic

Newtonian physics could make this calculation. Fiat could not explain, through its computations or through any other means, why its results diverged so extremely from Digitek’s. “The shortcomings of Fiat’s cube type barrier should be apparent from these widely divergent differences,” Gardner wrote.132

On July 7, Garibaldi wrote to the Department of Transportation, forwarding

Montabone’s analyses of Digitek’s tests and making some criticisms of his own. At this point, Fiat’s technical arguments against the validity of Standard 204 were quickly centering on the issue of the “undisturbed point.” This narrowing of focus signaled to the NHSB some apparent desperation on Fiat’s part. The only (technical)

129 Ibid., 4. 130 Ibid., 5. 131 Ibid., 4. 132 Ibid., 5.

186 leg the company had to stand on was that its cars contorted so egregiously during crash tests that all measurements involving the interior of the car, including those made by the scratch tube and the onboard camera, were rendered null and void. In his letter, Garibaldi argued that the lack of an undisturbed point in Fiat’s crashed cars rendered the test measurements “in fact fictitious” and in “serious fictitious excess” at that.133 It was unclear whether Garibaldi’s arguments about the undisturbed points were meant to be a universal damnation of the incoherence of

Standard 204’s test procedures, an assertion about the relative difference between small European and large American cars, a comment on Digitek’s testing procedures, or a statement about the ultimate flimsiness of Fiat’s vehicles. Garibalid belabored the issue of the undisturbed reference point for nine pages, and at times, he breached the inflection point at which bending logic and language becomes lying.

After outlining the numerous reasons why Digitek’s and Dynamic Science’s tests were inaccurate and assigning a numerical value to each of these factors (which, he argued, should be subtracted from each tests’ rearward displacement measurement), Garibaldi concluded that the Fiat cars in these tests had not violated

Standard 204 after all.

Also in July, at the NHSB’s request, Digitek produced its own analysis of Fiat’s critique of the contractor’s early reports. On the issue of the undisturbed point,

Digitek argued that Fiat’s objection was “a broad comment on test philosophy, and as such is not necessarily a technical objection to the test methodology.”134 The

NHSB had to make a determination of the issue, Digitek argued, “since it involves the

133 Letter, Garibaldi to James B. Minor, July 7, 1970, 2. 134 Letter, Jon S. McKibben to Francis Armstrong, July 22, 1970, 1.

187 fundamental interpretation of the standard.” Lying unspoken in the background of these discussions was the point that cars made by other manufacturers had passed

Standard 204 tests in spite of the fact that no point in any car remained completely undisturbed during a crash test. If these other cars passed, why were the Fiat vehicles so different? And, even if Fiat were right that the difference rested in the way its cars were constructed, was this difference value neutral? Or did it mean that

Fiat’s cars, which deformed so grossly during crash tests, were dangerous?

In mid-July, Schneider, the NHSB’s Assistant Chief Counsel for Regulation, sent out a memo eliciting technical comments on three questions: had the three tests carried out by Digitek and Dynamic Science conclusively established non- conformity? Had Fiat “used due care to ensure compliance?” Did Fiat have a

“reasonable basis for certification?”135 That is, did Fiat have good reason to think that its tests conformed to the demands of Standard 204? Schneider was cinching up the agency’s case, carefully, thoroughly. On August 3, Francis Armstrong, the

Director of NHSB’s Office of Performance Analysis, responded. Although he concluded, “It is hoped that the above opinions will assist our in the decision for proceeding with the penalty question issue,” his foregoing three pages made clear that Armstrong and the rest of the technical staff wanted the NHSB to go full-speed ahead.136

The tinge of joy that marked these later technical memos suggests that the

NHSB staff had begun to taste blood in the water, and they liked it. The NHSB’s

135 Memo, Lawrence R. Schneider to Francis Armstrong, “Compliance of Fiat 850 Sedan and Coupe with Standard 204,” July 15, 1970. 136 Memo, Francis Armstrong to Lawrence Schneider, “Compliance of Fiat 850 Sedans and Coupes with Federal Motor Vehicle Safety Standard No. 204,” August 3, 1970, 3.

188 technical staff, Armstrong noted, all agreed that the contractors’ tests were adequate proof that Fiat’s cars failed Standard 204. They deemed Fiat’s objections to the lack of an undisturbed point to be largely insignificant. Moreover, if Fiat thought that the undisturbed point was so critical, the company should have said something about it during the federal rulemaking process on the standard. (This point about the meaning of Fiat’s failure to comment on the original establishment of Standard 204 would become contentious in Garfield’s eleventh hour legal protests against the civil penalty.) On the issue of “due care,” Armstrong echoed the NHSB Chief Counsel

David Wells’ earlier point that, “The factual conclusion appears inescapable that

Fiat’s testing of these vehicles was grossly insufficient in quantity, in quality, and in timing.”137 Finally, on the issue of whether Fiat had a “reasonable basis for certification,” Armstrong believed that the movements of Fiat’s crash barrier went beyond the pale of any semblance of rationality or credibility. That barrier movement would absorb energy was so “technically fundamental” that any competently trained engineer would have to take it into serious consideration. As

Armstrong wrote, “It is inconceivable that some individual in Fiat’s engineering group did not logically deduce that the barrier was not creating the proper reaction for the test vehicle and therefore an improper basis for certification.”138 At this point of inconceivability, the NHSB’s technical staff believed, Fiat’s engineers had moved from poor engineering practice to unethical behavior.

137 Memo, David E.Wells, Chief Counsel, NHSB, to F. C. Turner, Federal Highway Administrator, through Douglas W. Toms, Director, NHSB, “Fiat Civil Penalty 850 Sedan and Coupe,” 2. 138 Memo, Francis Armstrong to Lawrence Schneider, “Compliance of Fiat 850 Sedans and Coupes with Federal Motor Vehicle Safety Standard No. 204,” August 3, 1970, 4.

189

On November 25, 1970, Schneider wrote a memo to Douglas Toms about the strength of the agency’s case against Fiat. “The Bureau’s case,” he argued, “appears to be a strong one from both the legal and technical points of view. Indeed, it is perhaps the strongest case the Bureau has ever developed and it is exactly the kind of case needed to establish, through litigation, the legal precedent which would greatly enhance the credibility of our enforcement program.”139 Schneider’s view may have been at least partly motivated by the fact that Fiat was a foreign carmaker.

Fiat gave the NHSB the opportunity to pursue a serious civil penalty without the risk of calling the agency’s patriotism into question. “Fiat,” Schneider wrote, ”has procrastinated and engaged in dilatory tactics in an apparent attempt to delay a resolution of the matter as long as possibly.”140

On December 1, 1970, Fiat’s counsel Garfield sent a letter to the NHSB’s

Office of General Counsel. In it, he pulled out all of the legal stops. The letter boiled down to two points. First, he claimed that Standard 204 was ambiguous. His absurdly legalistic paragraph on this point appears as the epigraph of this chapter.

It is a study in legal overkill. Left with nothing else, Garfield poured on seventeen cases of precedent, proving supposedly that ambiguous standards could not be enforced. Second, the NHSB had asserted that, if Fiat had a problem with Standard

204, the company should have made its case during the standard’s proposal process.

Fiat disagreed. It could question a standard at any point. Yet, Fiat’s counsel could find no precedent case supporting this logic. Grasping for straws, he turned to a

139 Memo, Lawrence Schneider, Acting Chief Counsel, to Douglas W. Toms, Director, NHSB, “Fiat Civil Penalty, 850 Sedan and Coupe,” November 25,1970. 140 Ibid.

190 case with the Federal Communications Commission; Garfield argued that under the

FCC case “the validity of administrative rules may be questioned at any time.”141

Garfield built pretty legal castles, but everyone involved could see that they were made out of sand. The NHSB’s attorneys would have none of his sophistry.

On December 9, 1970, Lawrence Schneider sent a letter to Garibaldi, informing him that, after considering all of the evidence at hand, including all of

Fiat’s many submissions and Garfield’s legal analyses, the NHSB would proceed with a civil penalty. The agency would ask “the Attorney General to institute a civil action against Fiat to recover the maximum civil penalties authorized by law.”142

Proverbially, the writing was on the wall. Actually, the writing lay in a trail of paperwork amounting to hundreds of pages. File folders held the records of a largely unseen battle. Inscribed by the mundane, ordinary labors of federal regulators who slowly, patiently, and arduously worked to enforce the law, the piles of paper were a testament to the banality of justice—and at least one company’s efforts to thwart it.143

141 Letter, Edward Garfield to James B. Minor, “Re: Fiat Steering Control Assemblies,” December 1, 1970. 142 Letter, Lawrence Schneider to Vincent Garibaldi, December 9, 1970. 143 Interestingly, while the NHSB was in settlement negotiations with Fiat, the agency wrote a letter to L. Patrick Gray, III, the Assistant Attorney General of the Department of Justice’s Civil Division, and requested that he make up the formal complaints that would begin an enforcement lawsuit if negotiations broke down. In November of 1970, the NHSB had sought to bring a smaller civil penalty against General Motors, but when those negotiations broke down, General Motors counter-sued the NHSB before the agency had a chance to file its own suit, effectively locking up the agency’s efforts. Having learned its lesson, the NHSB wanted the necessary paperwork to be ready so that the lawsuit could be filed immediately pending the breakdown of negotiations with Fiat. This way Fiat could not repeat GM’s ruse. A few days later, the Attorney General’s office sent over the formal complaints so that the NHSB was fully prepared. The agency had learned from experience and changed its strategy because of it. Letter, Lawrence Schneider to L. Patrick Gray, III, January 6, 1970; Letter and enclosed complaints, L. Patrick Gray, III, to the NHSB, January 8, 1971.

191

On February 22, 1971, Guido Foggini, the new president of Fiat Motor

Company, Inc., sent a letter to Lawrence Schneider.144 The envelope included a

$100,000 check—the sum of the settlement—made out to the US federal government. Fiat folded. The NHSB had won its largest civil penalty to that date. By this time, Foggini, who had worked in sales in the Fiat subsidiary in the United

States for fifteen years, had replaced Garibaldi, who returned to Italy.145 On

February 23, the Wall Street Journal ran one of the only stories to cover the Fiat investigation in the major press. The article was subtitled “Italian Auto Firm

Allegedly Didn’t Meet Federal Steering-Column Standards; Fine is Highest Yet,” but it also contained a qualification from Fiat. As the article noted, “Fiat issued a statement yesterday in New York disputing the findings of the outside laboratories

[Digitek and Dynamic Science], saying that Fiat’s own tests showed it to be in compliance with government regulations. Fiat said it settled the claim to avoid lengthy litigation.”146 The case did little to tarnish Fiat’s reputation in the eyes of the media.147 (Anecdotally, Fiat never seemed to have a great reputation among consumers, becoming known by the “backronym” of “Fix It Again, Tony.”) A New

York Times article from June of 1971—only four months after the settlement—made

144 Letter, G. Foggini to Lawrence R. Schneider, February 22, 1971. 145 Herbert Koshetz, “Fiat Spurs Trade,” New York Times, June 27, 1971. 146 “Fiat Will Pay U.S. $100,00 in Settlement of Civil-Penalty Claim,” New York Times, February 23, 1971. 147 The media’s silence on the civil penalty was, at least partly, the NHSB’s fault. Savvier administrators would have played media sources off of one another until the story brewed into something more noteworthy. (See my discussion of how EPA Office of Mobile Sources head Eric Stork did this in Chapter 5.) Yet, none of NHSB’s/NHTSA’s first three administrators—Haddon, Brenner (acting), or Toms—seemed to have any instinct about playing the media for their benefit. Though Toms was in the media quite a bit in the early 1970s because of controversies over the air bag, this coverage worked against him more often than it worked for him. See, for instance, “Nixon’s Car Safety Chief Ready to Quit,” New York Times, July 9, 1970.

192 no mention of the civil penalty. Instead, it stated that the company hoped to sell

55,000 cars in the United States that year. Quoting a Fiat representative, “I think we can make it. When it comes to price and quality we have nothing to fear from our foreign competitors.”148 Still, by that time, thousands of used Fiat cars on US roads contained safer steering columns, and the NHSB took some pride in its accomplishments. “Dear Mr. Foggini,” wrote Schneider on February 25, 1971, three days after receiving Foggini’s letter, ”We have received your check in the amount of

$100,000.00 in full payment of the settlement concerning the Fiat 850 Sedan and

Coupe. We are therefore closing our file in this case.”

Conclusion

No case is representative of the NHSB/NHTSA’s investigations into potentially defective vehicles. Each one is unique. Yet, as I have described in the case of the

Fiat investigation and recall of 1969–1971, every case is founded on the mundane, everyday, and uncelebrated labor that comprises regulation. On the one hand, during the investigation, Robert Gardner and his staff members spent the bulk of their effort examining scientific and engineering reports and letters and producing memos, reports, and analyses. On the other hand, Lawrence Schneider and the

Office of General Counsel did the equally necessary and complementary work of examining the legal grounding of the NHSB’s case against the firm. Meanwhile, Fiat tried to undo Standard 204. A large part of the company’s case involved questioning the practices that the NHSB was trying to make standard, or hegemonic, or

148 Koshetz, “Fiat Spurs Trade.”

193 everyday, across the industry and its associated laboratories. Fiat argued that its crash tests, which were considerably different from those recommended by the

Society of Automotive Engineers, produced comparable results. The NHSB disagreed. The two sides battled over competing ideas of what constituted a satisfactory test of regulatory compliance. Fiat lost. Interestingly, however, the agency was not willing to turn all of its work into mundane procedures. When the agency was faced with making a policy about how to treat firms that self-reported non-conformity with the federal standards, it eschewed creating a rule, deciding that each case should be dealt with on its own and that an ambiguous policy might actually spur automakers to innovate more than would a clear-cut one. Though the justice of regulation is largely founded on banality, from an agency’s perspective, the interplay between the ordinary and the extraordinary—between mundane circumstances that are covered by the group’s routines and unique cases that require deliberation and debate—never stops.

Chapter Four—Organizing Federal Automotive Emissions Control: Bureaucratic Change as a Strategic Enterprise

In the early 1970s, Honda Motors announced that it had created a stratified charge engine, termed the CVCC [Compound Vortex Controlled Combustion], that had the potential for meeting the stringent emission standards promulgated by the

Clean Air Amendments without losing performance or fuel efficiency. Honda Motors built much of its public relations campaign around the CVCC engine during that time. Arising from relative obscurity only a two earlier, the company now possessed a piece of technology that outpaced the American auto firms. So when the CEOs of

GM and Ford pooh-poohed the CVCC engine, claiming that it was a fine motor for small Japanese cars but would never work for a genuine, full-blooded American car, company president Soichiro Honda was understandably upset, so upset, in fact, that after the Detroit auto show, he purchased a Chevy Vega and a Chevy Impala and took them back with him to Japan. Once the cars arrived in Japan, Honda’s engineers fitted them with CVCC engine systems to prove that they would work in any vehicle.1

Not long after that time, Eric Stork, the administrator of the Office of Mobile

Source Air Pollution Control (hereafter, Office of Mobile Sources), a division of the

Environmental Protection Agency (EPA), traveled to Japan to visit the country’s auto producers. While at Honda, the company’s engineers asked him if he would like to test drive the Chevy Impala that had been fitted with CVCC technology. “Of course,” he said. Stork drove the Impala around Honda’s proving grounds, getting a feel for

1 On Honda’s putting stratified charge on Chevrolet vehicles, “Do Auto Catalysts Meet the Safety Test?” Business Week, October 27. 1973, 116.

194 195 how the modified car performed. When he was finished, one of the engineers asked him what he had thought about it. Stork said that the car was fine, but he noticed a bit of a delay after the “tip-in,” the moment when the accelerator is depressed. “Oh,” the engineer replied, “we aren’t going to make Impala’s so we didn’t modify the carburetor. We only changed the heads.”2

A month later, Stork was giving a speech at a conference held by GM.3 At the appointed time, Stork rose and walked to the lectern with the text of his printed speech in hand. Inspired by his experience in Japan, he decided to preface his speech with an impromptu version of a traditional tale. First, he told the gathered audience about his experiences in Japan and about the Impala with a changed heads, and then he told the story of Old Blue Suit, which went like this:

Once there was an old man named Burns, a lively drunken fellow who spent most of his time in pubs. To distinguish himself in this anonymous world, Burns wore a blue suit everyday. Overtime, people took to calling him Blue Suit. So, it was

Old Blue Suit who caroused and kidded away his time in the bars and taverns of his town. Well, as with many who spend their days drinking, Old Blue Suit died relatively young from various complications relating to alcohol consumption. His body was taken away and prepared for the wake. On the first day the body was to be displayed, Mrs. Burns arrived at the funeral home to view her husband one last time before the crowds arrived. To her horror, when the funeral director brought her to the body of her husband, Old Blue Suit was dressed up in a brown suit. She

2 Interview with Eric Stork, Arlington, Virginia, April 10, 2010. 3 Eric O. Stork, “Living with Regulation,” Automotive Emissions Seminar, Environmental Activities Week, General Motors Corporation, Warren, Michigan, June 19, 1973. The joke that follows was delivered extemporaneously and is not in the text.

196 cried and protested that this was no way for her husband to be seen. Since everyone knew her husband in his blue suit, they would barely recognize him in this brown getup. The funeral director reassured Mrs. Burns that her husband looked fine in the brown suit and that, besides, there was not enough time to change suits, but Mrs.

Blue Suit would not be denied. Her husband would not be seen in a brown suit. She protested so emphatically that eventually the funeral director gave in. Well, the viewings and the funeral were a success, and, true to his nickname, Burns was displayed in a fine blue suit. After the funeral, Mrs. Burns returned to the funeral home to thank the funeral director. “It must have been such a hassle,” she said, “to undo your work and change his suit.”

“No. No, Mrs. Burns,” said the funeral director, “It was no trouble.”

“But he looked so perfect in the brown suit, and you must have worked so hard,” she continued. “It must have been a great pain to have to take him out, to change him around, to re-fix his makeup, and to place him once again in the casket.”

“No, no, Mrs. Burns,” the funeral director reassured her. “It was no trouble at all.

We just changed the heads.”

Many people in the audience laughed at Stork’s tale, but he recalls hearing only silence from the front row, where the CEOs, CFOs, presidents, and the other

Detroit VIPs sat. They were silent because this story worked on many levels. Stork was discussing not only Old Blue Suit’s head and the heads of the Chevrolet cars but also the heads of companies. In that setting, where the story was told by a government employee who maintained an adversarial relationship with the auto companies, these many meanings would not have been missed. The story of Old

197

Blue Suit, the account of Stork’s telling of it, and the tale of why he came to tell it— all point to how attempts to change the automobile’s technological systems were inextricably tied to an immensely personal dimension of who was in charge, who was making efforts to change technology, and who was trying to conserve the car’s current state.

The story of Old Blue Suit also speaks to the changes that needed to happen within the federal government itself. Federal efforts to regulate automotive emissions during the 1960s had largely failed. But legislative and organizational changes in 1970 quickly gale-force winds that forced changed on the automakers and their cars. Part of this change involved new leadership—a change of the head— when a man named Eric Stork took control of automotive emission control at the new Environmental Protection Agency. Stork considerably revised the organizational structure and priorities of the agency. Regulating automotive emissions took new thinking, and Stork viewed the field of play—much as the business world does—as a place of adversarial strategy. Throughout his tenure,

Stork and his staff members moved the agency in new directions. The agency took on a new “regulatory style,” a style that played an important role in determining the

EPA’s overall effectiveness in combating automotive emissions.

This chapter proceeds through six sections. First, I present some theoretical

“variables” that can be used to track organizational changes. Second, I describe the history of the National Air Pollution Control Administration (NAPCA), the Office of

Mobile Sources’ (by most accounts) largely ineffectual predecessor agency. Third, I recount the birth of the Office of Mobile Sources, which strongly relates to the

198 arrival of Eric Stork, the office’s director from 1970 to 1978. Here, I emphasize how

Stork changed the agency from being one with a strong research and development

(R&D) wing to one being almost wholly focused on technological assessment and how he and his staff built the agency’s capacity around this new organizational vision. I also describe how a number of young experts came to work for the agency.

In the next three sections, I examine how the agency reacted to various instances of adversity, each of which arose from a different sector. In section three, I examine how the Office of Mobile Sources dealt with Ford Motor Company when it discovered that Ford had falsified some certification records. Stork’s immediate reaction was to shut down production at Ford. I once again examine how the agency related to the automakers, but I also look at how William Ruckelshaus, the

Administrator of the EPA, tempered Stork’s fury and found a way to keep Ford open.

Thus, I explore how superiors can limit a bureaucracy’s practices. In the fourth section, I describe how the agency reacted to the automakers’ use of “defeat devices,” technological systems that turned emission control systems off under certain conditions. Here, I point to how Stork used “advisory circulars” to manage ambiguities in the agency’s founding legislation and in the federal code. The section also focuses on how the agency dealt with resistance from the automakers. Finally, I examine how the Office of Mobile Sources dealt with the so-called “sulfate flap,” when some suggested that catalytic converters were producing a by-product that was more toxic than the chemicals the EPA was supposed to regulate. Concerns over sulfate were generated by another office within the EPA, and, therefore, in

199 studying how the EPA and the automakers overcame these concerns, I examine how the Office of Mobile Sources handled intra-agency rivalry.

The EPA’s Office of Mobile Sources is sadly and sorely lacking materials in the

National Archives and Records Administration. It is unclear why more records were not deposited in the federal archive system. That the agency’s main office in Ann

Arbor, Michigan is relatively isolated from Washington, D.C., and the bureaucratic culture that fosters archival collection perhaps explains the situation. The office’s engineering culture provides yet another reason. At some points, I have had to rely solely on oral histories, but wherever possible I have attempted to triangulate these interviews both with government documents (including published reports) and with sources drawn from news and trade publications.

The Strength of an Agency: Four Variables and Iron Triangles

Federal automotive emissions control efforts in the United States were famously weak during the 1960s. Much of this weakness can be attributed to the emission controls laws of the period, but the National Air Pollution Control

Administration (NAPCA), the agency charged with managing all the nation’s air pollution, also had several shortcomings. These twin, intertwined defects—legal and administrative—fed upon one another, yielding a downward spiral of regulatory ineptitude. With the coming of the Clean Air Act and the creation of the

EPA—and even before these developments—the administration of automotive air pollution controls began to change considerably. Therefore, developing some simple markers to examine these changes can be helpful, before turning to the history itself.

200

The political scientist Charles O. Jones, one of the earliest scholars to study

NAPCA, used a four-variable model of administrative strength developed by Francis

E. Rourke to examine the agency.4 Though Rourke’s and Jones’ respective works are somewhat dated, their analyses and concepts remain helpful heuristics for evaluating organizational change at the NAPCA and the EPA. Rourke employed four variables—constituency support, expertise, organizational vitality, and leadership— to capture the relative power of different agencies. In Rourke’s view, agencies that lacked these factors existed as “stepchildren of the executive branch.”5 Jones’ point in applying Rourke’s schema was simple: NAPCA had none of these characteristics; it definitely was a stepchild.

By “constituency support,” Rourke referred to “all groups that regard themselves or that are regarded by an agency as benefiting from its work.”6 In political and academic parlance, organized constituents are known as interest groups. The importance of a federal agency having a congressional committee in its corner is conspicuously absent from Rourke’s formulation. Other writers have emphasized, with some justice, the critical role of the congressional committee.

Indeed, historians and political scientists have theorized that successful federal programs require an “iron triangle” of an interest group, a congressional committee,

4 In many ways, Jones’ essay remains the best work that treats NAPCA in a semi-historical fashion. Charles O. Jones, “The Limits of Public Support: Air Pollution Agency Development,” Public Administration Review, Vol. 32, No. 5 (September/October 1972), 502. Rourke proposed these four factors in Bureaucracy, Politics, and Public Policy (Boston: Little Brown, 1969). 5 As quoted by Jones, 502. 6 As quoted in Jones, 502.

201 and a federal bureaucracy.7 Considered from an agency’s perspective, both the interest group and the congressional committee could be considered constituents.

The inclusion of congress in this broader perspective does not present a problem for

Jones’ analysis, however.

Rourke’s focus on “expertise” and “leadership” may seem more obvious but are no less important. An essential feature of expertise in a regulatory setting is that an agency must find a way to internalize external knowledge and produce new knowledge, and it cannot do so without having capable staff members.8

By “organizational vitality,” Rourke and Jones meant morale, or

“organizational verve.” This variable is doubtlessly dependent upon the others.

Sustaining morale is difficult when an organization is either lacking good leadership or is getting outpaced by external experts.

Inwardly, these four variables can be experienced as an organization’s

“culture,” the feel of the place, the sense that one has when one is inside an institution, something that can be hard to put one’s finger on, a feeling that has a certain je ne sais quoi. Outwardly, we experience these variables as an organization’s

“style,” the path that it cuts in the world. Because of the intertwined factors of enabling legislation and organizational cultures, the NAPCA and the EPA had palpably different styles.

7 On the Iron Triangle, see Brian Balogh’s helpful summary in Chain Reaction: Expert Debate and Public Participation in American Commercial Nuclear Power, 1945-1975 (Cambridge: Cambridge University Press, 1991), 62-3. 8 Wesley M. Cohen and Daniel A. Levinthal, "Absorptive capacity: A new perspective on learning and innovation", Administrative Science Quarterly, Vol. 35, No.1 (1990), 128-152.

202

Before the Clean Air Act Amendments of 1970: The National Air Pollution

Control Administration

A number of historical accounts allow for us to peer behind the walls of the

NAPCA. Jones’ work is one case, but there are others as well. Importantly, one of

Ralph Nader’s “Study Groups” published an examination—and damnation—of federal automotive pollution control efforts during the 1960s, entitled Vanishing Air:

Ralph Nader’s Study Group Report on Air Pollution.9 As with auto safety, Ralph Nader played an important, if less crucial, role in building momentum for the passage of the Clean Air Act amendments of 1970. Between the mid-1960s and mid-1970s,

Nader brought together a number of Task Forces, also known as Study Groups.

These temporary assemblies worked on specific issues. Staffed primarily by volunteer student labor, the Study Groups eventually published many reports.10 By the mid-70s, Nader had grown disillusioned with the Study Group model and came to prefer direct lawsuits against instigators of hazards over scholarly work. John

Esposito, who earned a law degree from Harvard University in 1967 and was a member of the District of Columbia Bar, led the Nader air pollution study, published the resulting book under his name. James E. Krier and Edmund Ursin’s 1977 book,

Pollution and Policy, remains the best overall history of early automotive emissions control efforts in California and bythe federal government.11 Finally and also

9 John C. Esposito, Vanishing Air: Ralph Nader’s Study Group Report on Air Pollution (New York: Grossman Publishers, 1970). 10 The wide variety of these reports truly demonstrates the breadth and power of Nader’s mind and the fact that he had his finger on the pulse of technological risk. 11 James E. Krier and Edmund Ursin, Pollution and Policy: A Case Essay on California and Federal Experience with Motor Vehicle Air Pollution, 1940–1975 (Berkeley: University of California Press, 1977).

203 important, three former NAPCA employees, who went on to work (at least briefly) at the EPA, wrote an essay recalling early federal automotive pollution control work.12

Federal air pollution control efforts can be traced to the, tamely named, Act to Provide Research and Technical Assistance Related to Air Pollution Control of

1955. This law furnished research funds to the Public Health Service (PHS), which, two years earlier, had become part of the then-new Department of Health,

Education, and Welfare. Federal work on automotive air pollution was centered at the Robert A. Taft Sanitary Engineering Center in Cincinnati, Ohio, which had been completed in 1954.13 This early research was focused on “characterizing” vehicle emissions, that is, describing what chemical compounds cars emitted and in what amounts.

The science of pollution from automobiles—and much of air pollution more generally—was in its infancy. A. J. Haagen-Smit, the “father of air pollution control,” had only recently isolated the cause of “smog” as a photochemical reaction between sunlight, hydrocarbon, and nitrogen oxide.14 Haagen-Smit had pointed to the automobile as one source of hydrocarbons in the ambient air. The work at the Taft

Sanitary Engineering Center focused on refining the science behind these conclusions. Researchers built a smog chamber at the center, which was used for studying the creation of photochemical smog under controlled conditions. The

12 Ralph C. Stahman, Kenneth D. Mills, and Merrill W. Korth, “Federal Air Pollution Efforts, the Early Years,” SAE Paper 890814. 13 Stahman et al, “Federal Air Pollution Efforts,” 1. 14 A. J. Haagen-Smit, “The Air Pollution Problem in Los Angeles,” Engineering and Science, Vol. 14 (December 1950), 1–7. On Haagen-Smit’s work and the story of air pollution control in California, see also Scott Hamilton Dewey, Don’t Breathe the Air: Air Pollution and U.S. Environmental Politics, 1945–1970 (College Stattion: Texas A&M Press, 2000).

204 science was in such an immature state that much of the center’s work focused on developing pollution measurement procedures, both for motor vehicle tailpipe emissions and for the ambient air.15

In 1963, federal auto emissions efforts moved to a new PHS laboratory in

Fairfax, Ohio. There, the PHS continued to do its emissions characterization work. It also began testing automotive emission control technologies created by independent inventors, usually “due to Congressional pressure or at the request of other higher level officials.”16 That same year, Congress passed the first Clean Air

Act, which gave the PHS the authority to set maximum levels for pollutants in ambient air (“air quality criteria”) and limited abatement powers.17 The Clean Air

15 Stahman et al, “Federal Air Pollution Efforts,” 1–3. 16 Ibid., 4. The EPA continued doing tests on independent inventions. Nearly all of them made insignificant reductions of auto emissions or increases to automotive fuel economy, which was diminished by some emission controls. Eric Stork, who I will discuss below and who led EPA automotive air pollution efforts during much of the 1970s, said that one significant invention came through the lab during his time at the agency. The technology cut back on the car’s air conditioning when the driver was accelerating, thereby, improving performance. Stork had one of the devices put on his car. (Phone interview with Eric Stork, February 25, 2011) Richard Lawrence, who later headed the EPA’s auto emissions control laboratory, told me that getting reduced emissions or increased fuel efficiency out of a highly engineered piece of technology, like the engine of an automobile, was simply very difficult. If any significant change was such low-hanging fruit, the automakers themselves would have taken advantage of it. (Interview with Richard Lawrence, August 18, 2008, Ann Arbor, Michigan.) The most controversial technology that the EPA tested was the so- called LaForce Engine, created Edward LaForce and his brother, Robert. Congress took interest in the engine, which the LaForce Brothers claimed was far more efficient than the average Detroit vehicle, and Stork was called to testify before his staff members had finished their testing. The EPA staff members had to do a rush job. With information in hand, Stork lambasted the LaForce engine in a Senate hearing, claiming that it contained no new or real contributions and that it had severely failed federal emissions tests. See EPA, “Testing of the LaForce Engine by the Environmental Protection Agency,” Staff Report Prepared for the Committee on Commerce, U.S. Senate, March, 1975; Environmental Protection Agency, “Evaluation of the LaForce-Modified AMC Hornet,” December 1974; “EPA Tests Refute LaForce Engine Claim,” Chemical Engineering News, Vol. 52, No. 51 (1974), 5. The Securities and Exchange Commission later found Edward LaForce guilty of criminal contempt after he continued issuing stock for his efforts after the SEC ordered him to stop doing so. See “Edward P. LaForce Convicted,” SEC News Digest, November 28, 1978. 17 The best treatment of the NAPCA’s abatement efforts are in Esposito, Vanishing Air, ch. 6. Briefly, under conditions of an ever-dwindling budget, the NAPCA’s Division of Abatement had to hold an arduous and time-consuming series of conferences, hearings, and discussions for each and every attempt to enforce federal ambient air standards. After a public hearing, the NAPCA had to begin by

205

Act also encouraged the auto industry to begin developing and implementing automotive emission controls, but such efforts were strictly voluntary.

Soon, however, members of Congress began to believe that the voluntary approach was not going to work. In 1964, Senator Edmund Muskie, a Democrat from Maine, began holding public hearings on the issue; the first hearing took place in California. Muskie grew interested in the topic of automotive air pollution during the original Clean Air Act hearings and debates.18 Over the course of the next decade, he would become the most formidable proponent of automotive emissions control in Congress. This process eventually led to Congress’s eschewing the voluntary approach and its passage of the Motor Vehicle Pollution Control Act in

October 1965. This law was the first to allow the government to set national automotive pollution control standards and, therefore, marked a milestone in the history of governance in the United States. Still, it was weak in several ways. Muskie had proposed that the standards be set in the legislation itself, but other forces prevailed, ensuring that a new administrative body within the PHS, the National Air

Pollution Control Administration (NAPCA), would set the standards through administrative procedures.19 Most important, the law required the NAPCA to prove that technology to meet standards was “feasible” before promulgating any such rules. This system differed from the one the National Highway Safety Bureau

making recommendations to a state government about possible abatements steps, and only after the state failed to take steps towards this goal could the agency begin federal intervention. In the process, state governments and potentially regulated businesses swamped the NAPCA’s capabilities, both in terms of financial resources and expertise. 18 For Muskie’s role in the hearings process and his turn towards pollution control, see Krier and Ursin, Pollution and Policy, 173–4. 19 Ibid., 174.

206

(NHSB) faced when setting automotive safety standards (as discussed in chapter 2), wherein, the NHSB had to base its tests and criteria on “previously existing” standards. The NAPCA had to prove that a technology already existed—these are sometimes called “off-the-shelf” technologies—before making a rule. For this reason, the NAPCA started putting increasing amounts of funding into R&D efforts.

The most important technology to come out of this research was exhaust gas recirculation (EGR), which the PHS developed in concert with the Esso Corporation and which was a key emissions control technology before the advent of the catalytic converter.20 (The case of EGR is important to keep in mind as a possible benefit of federal research as later federal efforts moved further and further away from R&D and more towards technology assessment.) The Motor Vehicle Pollution Control Act limited federal standards to levels that had already been set within California by an earlier state law (also titled Motor Vehicle Pollution Control Act, passed in 1960).

Additionally, the PHS adopted a test, known as the 7-mode cycle, that was used to determine whether a car met the standards. The federal standards were to take effect in the model year 1968 vehicles.

In 1967, John T. Middleton was appointed to head NAPCA. Middleton was a botanist by training, and before coming to the NAPCA, he headed the California

Motor Vehicle Control Board.21 (Haagen-Smit’s early research on smog was focused on its negative impacts on plants, especially those in California’s large and important agricultural sector, so botanists and others focused on plant health played important role in the early science of air pollution.) By most accounts,

20 Stahman et al, “Federal Air Pollution Efforts,” 4. 21 Esposito, Vanishing Air, 23.

207

Middleton was a kind and charming man.22 Professorial in his tastes, favoring tweed suits and wearing a handlebar mustache, Middleton was known for inviting colleagues to a glass of sherry in his office at the end of the day. Similarly unanimously, however, people viewed him as an ineffective leader. The Nader

Study Group criticized Middleton for not fighting harder for more attention for his program within the PHS and for more money from Congress. As Esposito pointed out, even Congressman Paul Rogers, a Democrat from Florida, asked Middleton in a hearing why he was not more forthcoming with budget requests, pointing out that

Rogers and other legislators wanted to help him in any way they could. 23 Similarly, as Charles Jones wrote after interviewing several people at NAPCA, Middeton was

“generally acknowledged and praised in interviews for his knowledge in the field, but rather uniformly criticized for his lack of administrative and political savvy.”24

The NAPCA lacked a true leader.

It also lacked expertise. In this regard, Jones simply pointed to the agency’s lack of both budget and manpower. In 1963, the agency had 407 employees, “a figure roughly equal to the number employed in air pollution control by Los Angeles

County at that time.”25 Though the NAPCA’s budget increased over ten times after

1963—going from an annual budget of less than $5 million to one of about $50 million—and the staff grew to over 1000, the agency still lacked the capacity to deal with the manifold issues in air pollution control, including carrying out studies on ambient air, examining the causes of pollution, and regulating air polluters of both

22 Ibid. 23 Ibid., 24. 24 Jones, “The Limits of Public Support,” 504. 25 Ibid., 503.

208 the stationary (such as factories and electricity generation plants) and mobile

(vehicular) variety.26 Things were even worse in terms of automotive pollution control. Esposito reported damningly, “Whether fraudulent or farcical, there is no doubt that NAPCA’s automotive emissions program is a smashing failure. . . . Though

[the automobile] accounts for 60 percent of the total problem, the program for control of automotive emissions has never accounted for more than 5 percent of

NAPCA’s budget.”27 Even to the degree that the agency had capable experts, few of their efforts were dedicated to the automobile.

The NAPCA had very few constituents during the 1960s. Jones pointed out that regulatory agencies tended to have less a less active public than “service agencies,” which provided some direct, perceivable benefit to the members of the public. Successful regulation tends to express itself as a lack of something negative, rather than by creating any positive reinforcement. Moreover, as Jones points out, state pollution agencies, which may have acted as constituents for federal efforts, were themselves just beginning to develop during the 1960s, so there was no organized group that would support the NAPCA. Rachel Carson’s Silent Spring had been published in 1960, and the modern environmental movement was emerging, but it had not fully emerged. Citizens groups had played an important role in

California, but they were primarily focused on pollution control efforts at the state level. They had not yet formed a national network. Just as the NAPCA lacked

“constituency support” from an interest group, so too did it lack a committed congressional committee. Muskie had become a firm ally to the movement, but his

26 Ibid. 27 Esposito, Vanishing Air, 52.

209 interest was just developing. He had not yet formed the personal fury that would mark his later salvos in auto emissions control.

The “organizational verve” of NAPCA ran low throughout the 1960s. The agency was always cash strapped. Moreover, it was constantly reorganizing and moving offices. The bureaucratic structure shifted and found new stability just before it shifted again. Jones recalled an old bureaucratic joke: a man walks into his office, hangs up his coat, and says to his secretary, “If the boss calls, get his name.”

The agency was lost in the hulking PHS, and the experience of being a stepchild while holding onto a seemingly important issue was very demoralizing.

Jones and Esposito were somewhat off base about their criticism of expertise at the NAPCA, however. The agency’s work was not fruitless. Woefully understaffed for the part it was supposed to play, NAPCA’s budget paled in comparison to the budgets and manpower of the automakers themselves. But done in the agency’s ranks, there was a good deal of institutional learning going on, learning that was eventually folded into the EPA. In terms of institutional knowledge about automotive emissions, the NAPCA had to begin largely from scratch. Furthermore, some of this learning eventually led to important findings.

For example, in 1970, Dr. Delbert Barth, Director of the NAPCA’s Bureau of

Criteria and Standards based in North Carolina, presented a paper at the annual meeting of Air Pollution Control Association in St. Louis in 1970.28 Barth had a long,

28 The paper was eventually published as Delbert S. Barth, “Federal Motor Vehicle Emission Goals for CO, HC, and NOx Based on Desired Air Quality Levels,” Air Pollution Control Administration Journal, Vol. 20, No. 8 (August 1970), 519–523. Much of the following biographical description comes from a website that aggregates the biographies of important engineers at the PHS. For Barth’s entry, see http://www.usphsengineers.org/History/Bio/D_Barth.htm (Last accessed March 2, 2011). For other

210 storied career before he joined NAPCA. He received his Bachelors in Military

Engineering at West Point in 1946 and spent the next two years training at the Army

Chemical Corps at Oak Ridge National Laboratory. Barth then went on to get a

Masters in Nuclear Physics at Ohio State University in 1952, and, in 1960, received another Masters, this one in Solid State Physics, from Stevens Technical Institute.

Barth transferred from the Army to the PHS Commission Corps in 1960, completed his Ph.D. in Biophysics from Ohio State in 1962, and during the next decade became increasingly involved in environmental issues. He was very successful in the PHS.

In 1970, by which time he had entered the NAPCA, Barth was promoted to Assistant

Surgeon General for Engineering (the equivalent of a Rear Admiral). When NAPCA became part of the EPA, Barth stayed on, becoming the head of the National

Environmental Research Center in Triangle Park. In his 1970 paper, Barth examined the role reductions in auto emissions would have to play if overall ambient emissions were to be curtailed. After explaining his methodology, Barth proposed the percentage by which automotive-generated carbon monoxide (CO), nitrogen oxide (NOx), and hydrocarbons (HC) would need to be reduced: 92.5 percent, 93.6 percent, and 99.0 percent respectively. These numbers would entail extremely tough standards.

Barth’s paper has a peculiar place in historical memory. It was clearly important, but exactly what was important about it is less clear. One person who worked for both the NAPCA and the EPA told me that he had always heard that

Barth had posited the reductions needed to make the automobile insignificant in biographical information on Barth, see Who's Who in Science and Engineering - 2000-2001, 5th Edition (1999).

211 ambient air pollution and that these numbers had gotten picked up in the Clean Air

Act Amendments of 1970.29 But another person, who also worked for both agencies, told me that he was sure that Leon Billings, Edmund Muskie’s mercurial chief of staff and a key author of the Clean Air Act Amendments, had, in fact, set the reduction levels.30 In this version of the story, Barth then wrote the paper retroactively in order to buttress the Billings’ numbers.

The auto industry often protested that one major problem with the standards was that they were not formulated by someone with experience in automotive engineering.31 To some degree, the automakers had a point. Barth was trained in many disciplines, but he was not an automotive engineer. While he knew a great deal about air pollution, he had little grasp on the nature of technological change in the auto industry. If Billings set the numbers, the auto industry’s criticism is even more accurate, because Billings was a political staffer with a Bachelor’s degree from the University of Montana, and was definitely not an expert in automotive technology. The level for the standards was not set by anyone with deep knowledge of the industry or its organization. Whether the standards were too strict is a question that cannot be settled. The answer depends on a number of ethical assumptions. The automakers said they could not meet the standards. But were they being honest? Did they put as much effort into research as they could and

29 Interview with Eric Stork, Arlington, Virginia, April 10, 2010. 30 Phone interview with Edwin Tuerk, April 22, 2010. On Billings’, see Who's Who in American Politics, 22nd Edition (2009). 31 See, for instance, Agis Salpukass, “Court Fines Ford $7-Million in Suit on Pollution Test,” New York Times, Feb 14, 1973: “The industry has maintained that the standards were drawn up by people who had little [experience] come up [sic] with the technology and the production process to meet the 1975 levels.”

212 should have? Answers to these questions relate more to our moral intuitions than to our ability to know definitively.

Developing the Federal Test Procedure

The NAPCA left another important legacy—the development of the test at the heart of all federal automotive emissions control. Unlike the National Highway

Safety Bureau/National Highway Traffic Safety Administration, which has a host of tests that it uses to judge the safety of automotive systems, federal automotive emissions standards have always been based on one central test, which has evolved over time. The test, known as the Federal Test Procedure (FTP), has two basic components: a procedure for collecting and measuring tailpipe emissions and a

“driving schedule,” or a series of accelerations, static driving speeds, decelerations, and idle times. The FTP is run in a laboratory on a chassis dynamometer, a device that allows the cars wheels to spin during the test in order to simulate road driving.

Like much of the federal emissions control program, the FTP can be traced back to earlier work in California. In 1956, the Automobile Manufacturers

Association (AMA), the Coordinating Research Council (a research organization primarily funded by automakers and oil companies), and staff members of Los

Angeles County carried out a series of studies used to characterize the driving habits of the “average” driver.32 Subsequent reports recommended and then carried out

32 See Ronald E. Kruse and Thomas A. Huls, “Development of the Federal Urban Driving Schedule,” SAE Paper 730553, 1; Thomas A. Huls, “Evolution of Federal Light-Duty Mass Emission Regulations,” SAE Paper 730554, 1–2. In 1972, the Automobile Manufacturers Association was renamed the Motor Vehicle Manufacturers Association. In 1999, this trade association was replaced by the Alliance of Automobile Manufacturers.

213 studies of driving habits using “11 speed-defined modes,” including accelerating, decelerating, idling, and various speeds of urban, suburban, and highway driving.33

These modes were then translated into “speed-manifold pressure categories,” which were recorded using a “driving habits recorder” designed by DuPont.34 Clocks kept track of how long the driver spent in each mode, but in the process, the chronology of the modes and the movement between them was lost. Los Angeles County air pollution control staff members then built a test that ran a car through each of these modes, spending the requisite amount of time in each engine state.35

From 1961 to 1964, California authorities transitioned from this full 11- mode test to a simpler 7-mode one.36 At first both tests were run within a larger test procedure, but, by 1964, the test procedure consisted of the 7-mode test run seven times, thereby, becoming known as the “7-mode, 7-cycle” test. This procedure was called an “open” test since emissions were only captured at particular times, not during the test’s entirety. (The term “open” makes more sense when one considers that during a “closed” test all of the tailpipe emissions are captured.) In the 7-mode,

7-cycle test, testors did not collect emissions during the fifth cycle, which was used

33 During the cycle, the car was supposed to spend 15% of the time in idle, 16% of the time in a cruise (constant speed) mode, 36.7% of the time accelerating, and 32.3% of the time decelerating. The cruise, acceleration, and deceleration modes were further subdivided by speeds. Taken altogether, the eleven modes were as follows: idle; cruise at 20 mph, 30 mph, 40 mph, and 50 mph; acceleration between 0–60 mph, 0–25 mph, and 15–30 mph; and deceleration from 50–20 mph, 30– 15 mph, and 30–0 mph. D. M. Teague, “Los Angeles Traffic Pattern Survey,” Paper No. 171 presented August 1957 at the SAE National West Coast Meeting, collected in SAE, Vehicle Emissions (Selected SAE Papers), (New York: SAE, 1964), 31. Teague, an engineer at Chrysler, was the head of the Traffic Survey Panel of the Automobile Manufacturers Association, on which he was accompanied by members from General Motors, American Motors Corporation, Studebaker-Packard Corporation, and another engineer from Chrysler. 34 Kruse and Huls, “Development,” 2. 35 Ibid. 36 Huls, “Evolution,” 2.

214 instead as a transition from the “warm up” period to “hot” operation. From early on, experts in automotive emissions realized that cars produced more problematic emissions when they were “cold,” for instance, when they were first started after sitting overnight. To recognize this disparity and to attempt to best represent average driving habits, test designers tried to balance the test “weight” of “cold,” or

“warm up,” phases with “hot” ones. California used the 7-mode, 7-cycle test to begin certifying cars in 1966. The criteria for the test were set on a concentration basis: hydrocarbons were expressed as parts per million; carbon monoxide, as a simple percentage by volume.37

When the U.S. Congress passed the Motor Vehicle Pollution Control Act of

1965, the NAPCA adopted the California 7-mode test nearly completely. The only significant difference was that the NAPCA set different emissions criteria for varying engine displacements. Cars with smaller engine displacements had correspondingly

“lower exhaust volumes,” so they were allowed looser standards because regulated emissions made up a larger percentage of their total exhaust volumes.38 These federal standards took effect with the 1968 model year.

In 1968, the NAPCA announced that it was tightening standards, which would take effect in 1970 model year. For the first time, the NAPCA expressed the standards in mass units (grams per mile) instead of concentration (parts per million or a percentage). Yet, since the 7-mode test produced results in terms of concentrations, testers used an “empirically derived formula,” which took into account the vehicle’s “test weight,” to translate the original results into grams per

37 Ibid. 38 Ibid., 2–3.

215 mile. Similar to having different emissions criteria for different engine displacements, factoring the vehicle’s test weight made allowances for smaller cars.

NAPCA staff members believed that the “change [to mass units] provided improved equitability between vehicles of various sizes and types.”39

Also in 1968, in its aim to improve test accuracy, the NAPCA announced that it was devising a new test procedure for the 1972 model year that would measure the “true mass” of vehicle emissions, rather than relying on the formula to translate between concentrations and mass units. The new procedure, known as the

“constant-volume sampler” (CVS), was a “closed” test that collected emissions for the entire duration of the procedure. One reason that the NAPCA moved in this direction was that, since the 7-mode method only collected tailpipe emissions at given points and times, agency staff members began to believe that the test might too easily allow automakers to design vehicles that met emissions standards under those particular conditions but that otherwise failed the emissions standards. Tests based on concentrations also became problematic because automakers were increasingly using air pumps in their engines for a variety of reasons. These air pumps increased the overall amount of air in the exhaust, thereby, reducing the percentage of regulated emissions in the mixture. Since the CVS procedure gave results in grams per mile, testers no longer had to worry about the relationship between concentrations and exhaust volumes, so they could do away with different

39 Ibid., 3.

216 criteria for different engine classes. The CVS procedure was more accurate, and it was harder to pass.40

“Closing” the test procedure created some problems, however. Most importantly, the 7-mode, 7-cycle test had been carefully designed to “represent” the average driving patterns in Los Angeles County. By closing the test, it became much more difficult, and the automakers argued that it was no longer representative.41

After some initial efforts, the NAPCA decided to abandon the 7-mode test in favor of another driving schedule. The driving schedule to which the agency moved also had an older history. In 1964, California had decided to move away from the 7-mode cycle, which was based on driving patterns throughout Los Angeles County, to an urban driving pattern, since, it was argued, urban driving was the root of the city’s smog problem.42 The driving schedule that eventually emerged from this study, known as LA-4 (it was the fourth Los Angeles urban schedule developed), was modeled on the “morning home-to-work trip.”43 In 1966, the NAPCA created a road route similar to LA-4 in Cincinnati, Ohio. The new road route was dubbed CLA-4, and the agency used a “driving mode analyzer,” which the PHS had co-created with the Ethyl Corporation, to examine the “driving mode data” in a fashion similar to

40 In the first EPA emission control suspension hearings (examined in depth in chapter 5), which occurred in April 1972, the automakers complained that the CVS test procedure, which had only recently taken effect, made it more difficult for the companies to make emission controls that passed the exam, even though the standard (the grams per mile criteria) had not changed. The test had just become tougher. 41 Huls, “Evolution,” 3. 42 G. C. Hass, M. P. Sweeney, and J. N. Pattison, “Laboratory Simulation of Driving Conditions in the Los Angeles Area,” SAE Paper 660546, presented at SAE West Coast Meeting, Los Angeles, August 1966. 43 Kruse and Huls, “Development,” 2.

217 what had been done with the 7-mode test.44 In 1968, the NAPCA devised another road route in Ypsilanti, Michigan (YCLA-4). While the agency continued using mode analyzers that relied on engine manifold pressure, it also began to use a fifth wheel to record actual road speed. The fifth wheel also had the benefit of maintaining the chronology of idling time, accelerations, cruising speeds, and decelerations that mode analyzers lost. The data generated through the fifth wheel led to a paradigm shift. As EPA employees later put it,

“Once it was realized that the average speed profile chart contained not just the major modes of the route but also included all of the little mode changes which characterize real driving, consideration was given to how this information might be transformed to dynamometer driving. Of course, the simplest and most straightforward method would be to use the route speed profile chart as a driving aid and drive the car to match its speed to the previously recorded speed profile. This was tried, and it was found that a driver could indeed follow the speed profile with reasonable precision and without greater difficulty or more fatigue than was associated with driving an equivalent number of 7-mode cycles.”45

The mode analyzers had always suffered from data simplification because, no matter how many engine manifold pressures were captured, small changes in driving speed were lost. By using the actual driving speed as the basis of the dynamometer test, these small changes were retained. (In the early years, technicians in the EPA’s Ann Arbor Lab would look at the driving speed profile and

“re-trace” it by applying and releasing pressure on the accelerator. Later, staff members built a “robot” that did the dynamometer driving.46) The NAPCA eventually sent staff members to Los Angeles to examine average speed profiles on

44 Ibid. 45 Ibid., 3. 46 Interview with Karl Hellman, August 19, 2008, Ann Arbor, Michigan.

218 the original LA-4 test route. They used a number of different drivers to create a

“representative” profile. Since the LA-4 test route was 12 miles long but nationally people drove an average of 7.5 miles a day, the NAPCA created a shorter test route, known as the Urban Dynamometer Driving Schedule (UDDS).47 Together, the CVS and UDDS comprised the Federal Test Procedure.

Early on, many questioned whether the UDDS, which was based on driving patterns in Los Angeles, could really be taken to represent the average American driver. But a 1971 survey, which was carried out by the Coordinating Research

Council, found that LA-4 was broadly representative.48 Since then, scientists and engineers both within the EPA and outside of it have constantly studied the FTP, taking stock of its strengths and weaknesses.

In 1968, the NAPCA moved its automotive branch office from Fairfax, Ohio, to near Detroit, originally renting an empty hanger at the Willow Run Airport in

Ypsilanti, Michigan, that acted both as a makeshift office and as a lab for certification and other tests. 49 The ambient air and most other research operations that were

47 Kruse and Huls, “Development,” 3-4. 48 Coordinating Research Council, “A Survey of Average Driving Patterns in the in the Six Urban Areas of the United States: Summary,” CRC Study CAPE-10, in the author’s personal collection. Soon after this study was finished, the EPA ceased working with the Coordinating Research Council because high-level officials worried that this collaborative research might lead to collusion. Interview with Eric Stork, April 10, 2010. 49 Consider that the NHSB/NHTSA never had a major office or laboratory based near Detroit. The NHTSA lab was built in Ohio. This may have had some influence on the different effectiveness of the two agencies. With the exception of certification tests in which only the automaker being tested was allowed to attend, the Office of Mobile Sources always had an open door policy with the auto companies. The agency may have been able to build stronger, more effective relationships through its nearness to Detroit. Moreover, economic geographers have argued that “spin off” companies tend not to move far from their originating company. Thus, industries tend to build technical hubs. Detroit, Silicon Valley, Pittsburgh, North Carolina’s Research Triangle, and Boston’s Route 128 all fit

219 centered in Cincinnati moved to Research Triangle Park in North Carolina. With the the Clean Air Act Amendments of 1970 on the horizon and the new laboratory in

Michigan, the organization was ready to make a change. In terms of organizational strengths and weaknesses, the NAPCA makes an interesting and compelling counter-point to the history of the agency that followed, the EPA Office of Mobile

Sources, and it is to that agency’s history that I turn now.

Technical and Organizational Change at NAPCA and the EPA

In June of 1970, the US House of Representatives passed the Clean Air Act

Amendments by a vote of 374-1. The House version was based on a bill put forward by the Nixon Administration that lacked the stringent deadlines that would come define the law’s reputation. In September, the Senate passed the tougher Muskie version of the bill by a vote of 73-0. Federal automotive pollution control had found its constituency in the energy generated by the first Earth Day and other expressions of the early modern environmental movement, and in Muskie, who was disgusted that earlier pollution control efforts had come to naught. The Clean Air Act

Amendments had additionally been helped along, as one news article noted, when

“massive air inversions had blanketed the East Coast, including Washington, with hot, sticky, dirty air several times during the summer.”50 As Senator Eugene

this model. If nearness is important for industry, it could equally play a crucial role in business- government relations. 50 Donald A. Colburn, “Washinton’s RX: Will the Cure Be Worse than the Disease?”, Automotive Industries, February 15, 1974, 72.

220

McCarthy claimed, “It’s better than motherhood. At least a few people are against motherhood. Nobody can be against clean air.”51

Unlike with the earlier Motor Vehicle Pollution Control Act, Muskie got his wish with the Clean Air Act Amendments: he built the emission standards into the law itself. The law required that automakers reduce carbon monoxide and hydrocarbon emissions by ninety percent by the 1975 model year and nitrogen oxide by the same amount for the 1976 model year. Legislators and regulators knew that the automakers did not currently have technology available to meet these standards. In this way, the law was “technology-forcing”: it was meant to induce technological change in the auto industry. The Clean Air Act Amendments did contain a “safety valve,” however. If automakers believed that they could not meet the standard, they could petition the EPA to suspend the standards for one year.52 (I deal with this petition and suspension process in the next chapter.) Still, the Clean

Air Act Amendments comprised an extremely tough law; it was perhaps the most stringent regulation of technological risk in the history of the United States. In an interview with Automotive Industries, Eric Stork, who would come to helm these regulatory efforts at the EPA, said, “The fine art of regulatory management involves

51 Ibid. 52 Like previous federal automotive emissions laws, the Clean Air Act Amendments of 1970 allowed for the state of California to set stronger automotive emissions standards than the national ones. This policy created a “two car system,” in which cars manufactured for California would have more effective emission control systems than automobiles made for the rest of the country. Technically, the EPA had to hold public hearings before giving the okay to California’s tighter standards. But a former EPA employee told me that the Office of Mobile Sources saw these hearings as a necessary but an ultimately empty, even boring, activity. As the law was written, the agency had to have a very good reason for denying California’s requests, and no such reasons existed. The EPA allowed California’s tighter automotive emission control programs to go unchallenged until 2007, when the George W. Bush administration struck down the state’s attempt to regulate green house gas emissions, particularly carbon dioxide (CO2), from cars. On the Bush administration’s refusal of California’s request, see Bryan Walsh, “California’s Clean-Air Slapdown,” Time, December 20, 2007.

221 the ability to identify, get hold of, and hang onto the short hair. And the Clean Air Act provided EPA the best grip on the short hair of the industry that any government agency has ever had.”53 The law, one journalist noted, might induce the automakers to “recall Dr. Johnson’s comment about the man who is told he will be hanged in a fortnight: it concentrates the mind wonderfully.”54

Then Came a Stork: Leadership and Strategy in the Latter Days of NAPCA and at the

Office of Mobile Sources

By 1970, it was clear that the automotive portion of NAPCA was in need of repair. The office was disorganized and lacked a clear institutional vision. John T.

Middleton, NAPCA’s head, had taken a beating in front of Congress when he admitted that around eighty-five percent of cars were between fifteen and twenty percent over the emissions standards promulgated under the Motor Vehicle

Pollution Control Act.55 Moreover, the agency knew that the Clean Air Act

Amendments were on the horizon, and its leaders were looking for talent. Ed Tuerk, who had been an Associate Commissioner at the Food and Drug and Administration through most of the 1960s but had become a Deputy Commissioner at the NAPCA, recruited one of his previous employees, Eric Stork, to NAPCA. Stork was a career bureaucrat. From 1962 to 1966, he was an Executive Officer of the Airports Service of the Federal Aviation Agency, and then he became a Deputy Director of the Food

53 Donald A. Colburn, “Washinton’s RX: Will the Cure Be Worse than the Disease?”, 73. Stork often made statements similar to this one, and news sources enjoyed quoting his philosophical reflections on the “short hair.” 54 Ibid. 55 Esposito, Vanishing Air, 60–62.

222 and Drug Administration’s Bureau of Regulatory Compliance. Thus, Stork had worked in regulation for nearly a decade when he came to NAPCA, and he had a keen regulatory mind. Unlike the early heads of the National Highway Safety

Bureau, Stork understood the adversarial relationship inherent in regulation, and he embraced it.

When Stork arrived at NAPCA, the intelligent men and women who worked there were, to some degree, spinning their wheels. From its beginning, the agency had a confused mandate. On the one hand, it was a certifying regulatory agency that carried out day-to-day tests on automobile emissions. On the other hand, from the federal government’s earlier studies on vehicle emissions, NAPCA had a research and development wing.

Nothing in particular was pressing when Stork arrived, and he spent his first few weeks looking at the how to take lead out of gasoline. Later, leaded, or ethyl, gasoline, which Charles “Boss” Kettering partly invented in 1921, became a major factor in debates over and success of the catalytic converter, but in 1970 NAPCA already realized that it was an important issue. After those first few weeks, NAPCA administrators asked Stork to complete a management review, one of his skills, on the automotive branch of the agency. As he went around asking staff members what their function was, he found that most of the employees had no idea what he was asking them. It was a very arduous exercise. Stork’s central finding in the review was that the auto program needed an executive level administrator to manage and reform it. His superiors agreed in principle, but the agency was hemmed in financially, and they knew that the Clean Air Act would almost certainly generate

223 reorganization—it seemed unlikely and silly to hire a new director at that moment.

Instead, Tuerk asked Stork to fill the role, at least temporarily.

Various lower-level programs dealing with the automobile were reporting to different parts of the agency. For instance, the branch that certified whether cars met pollution standards reported to NAPCA’s Division of Abatement, while any branch working on research reported to the research arm of the agency. Stork’s first recommendation was that all of the automobile functions should be centralized into one office with one director. This process occurred fairly quickly. (A year later, the EPA carried out a similar process with non-mobile sources, creating the Office of

Stationary Sources.)

Though the names of the divisions changed over time, the Office of Mobile

Sources retained the same basic organizational structure. Stork’s office was in the

Washington, D.C. area. The Ann Arbor lab then had four divisions: Certification,

Emissions Control Technology, Administration, and Advanced Automotive Power

Systems (AAPS). The Certification Division was in charge of not only doing all of the certification work but also carrying out various surveillance programs, including studies of in-use vehicles. The Emissions Control Technology Division oversaw the various research programs within the agency. The Emissions Control Technology

Division had some rather stable functions, including a branch dedicated to technology assessment, another that examined devices created by external independent inventors, one branch that drafted the technical side of regulations, and, finally, a branch that kept tabs on emerging science issues, such as unregulated but potentially hazardous auto emissions. In the Administration Division, Stork

224 housed not only functions related to personnel issues but also any physical resources, such as the laboratories and computers, over which other branches might fight. The Administration Division was charged with distributing lab and computer time equitably—two things over which engineers and scientists might battle if a disinterested party was not brokering them. Finally, the AAPS program examined alternative engine systems, such as rotary engines, hybrid engines, electric motors, and stratified charge engines.

Reconstructing the biographies of the division chiefs is extremely difficult.

Louis Brownlow, a civil servant and political scientist who played a leading role in reorganizing the federal bureaucracy during the Franklin D Roosevelt

Administration, titled his autobiography, A Passion for Anonymity.56 Brownlow believed that this passion was an essential character trait of the ideal civil servant.

Unfortunately, the Office of Mobile Source’s division chiefs had this characteristic in spades, and their biographies are not available in traditional biographical sources. A man named Ed Brune led the certification branch for most of the 1970s. I discuss some of his work later in the chapter. David Alexander led the administration division, and John Brogan led the Advanced Automotive Power Systems division.

Brogan went on to have a very successful career at the Department of Energy. For the purposes of this chapter, I will focus, for a moment, on the leaders of the

Emissions Control Technology Division.

When Stork first moved to the agency a man named George Kittredge was in charge of research efforts on automotive emissions control. Stork found that

56 Brownlow, The Passion for Anonymity: The Autobiography of Louis Brownlow (Chicago: University of Chicago Press, 1955).

225

Kittredge had an excellent technical mind but believed he was not as tough as Stork believed was necessary. A deal was struck, and Kittredge moved to Washington,

D.C., as a Senior Technical Advisor to Stork.57 In this position, Kittredge was primarily responsible for two duties—regulating aircraft emissions and gaseous fuels. Stork wanted nothing to do with either of these issues, partly because larger and more powerful agencies had primary control over them. He knew the agency stood no chance against the Federal Aviation Administration, which had ultimate authority over aircraft emissions. And the General Services Administration was entangled with gaseous fuels.

Stork then began to hunt for a new branch head and found a man named John

DeKany.58 Before coming to the EPA, DeKany, a chemical engineer, worked for several years in research at Westinghouse under the renowned scientist John C. R.

Kelly. DeKany worked at the EPA’s Office of Research and Development until he heard of the division chief opening at Mobile Sources, to which he applied. DeKany put together the technology assessment team of Thomas Austin and Karl Hellman, who I will discuss below. He was also responsible for providing staff for and backing a number of important regulations that came out of the Office of Mobile

Sources. Stork described DeKany as a “volatile Hungarian” (DeKany was either a first or second generation immigrant from Hungary), but Stork appreciated that

DeKany was “willing to go to the mat if necessary.”59 He needed someone tough in

57 Phone interview with Eric Stork, February 25, 2011. 58 I have reconstructed the following biography of John DeKany from discussions and emails from Eric Stork and from an email from an anonymous source. 59 Phone interview with Eric Stork, February 25, 2011.

226 that role, since the Emissions Control Technology Division would provide the agency with information it needed in its struggle with the auto industry.

Stork set about reforming the agency in various ways. Perhaps his most consequential decision was to shift the agency’s focus from research and development (R&D) to technology assessment. As he later explained,

“I was convinced that we would never be able to get the huge resources required to do the research and development to demonstrate the feasibility of meeting the emission standards, and that even if we were to succeed in such a task the NIH (not invented here) factor would mean that the industry would focus its efforts on proving that we were wrong rather than on adopting and improving upon our work. I had very limited technical resources and I wanted to focus those resources on analysis and rulemaking that could force the industry to develop the required technology.”60

The Clean Air Act amendments of 1970 were not based on the notion of “best available technology” as previous instantiations of the Clean Air Act were. The agency did not have to “demonstrate,” through R&D, that the emissions controls were available. The standards were set to force the industry to develop the necessary technology, and the automakers were required to show that they could not meet the standards. In November 1970, Stork wrote a memorandum outlining his belief in the priority of technology assessment over R&D and his decision to

60 Email from Eric Stork, December 31, 2009. To some degree, Stork’s point about the “not invented here” syndrome has proven true. The Office of Mobile Sources has increasingly carried out research, sometimes jointly with industry. Indeed, the so-called “Super Car Program,” fostered by Vice President Al Gore during the Clinton Administration, was largely housed at the Office of Mobile Source’s Ann Arbor office. In casual conversations I’ve had with industry representatives, all have expressed doubt not only about the quality of research coming out of the Office of Mobile Sources but also about the very fact that the agency was doing research at all. Yet, this skepticism amongst industry is equaled only by the optimism and excitement of the researchers at the agency, the press, and advocates for technical change in transportation (both environmentalists and energy independence/energy security types). The NIH factor is one explanation of industry’s doubts. Surely, however, industry representatives would argue that they are skeptical because they actually know what’s going on, unlike the press and rose-visioned futurists. The Chicago Tribune ran a series of articles on the Supercar Program in 2002.

227 move the agency in that direction.61 This decision came as a disappointment to some of the young employees, who were excited by the reformist-spirit of the emissions control efforts and came to NAPCA believing that they were basically signing up for an R&D project. Stork knew this, and he always tread a careful line between maintaining what he saw as the agency’s primary goal and alienating his staff. As he recalled,

“The most talented engineers that I had really liked wanted to do research and development, so in developing our activities it was a constant struggle to keep them focused on analysis and assessment. I occasionally let them slip a research project past me, to the disgust of my “right-hand man” for management (Janet Lane) who on my behalf closely reviewed all RFPs [request for proposals]; when Jan complained that I approved a "researchy" RFP that she had identified as being contrary to my policy, I explained to her that to keep our best engineers I had to let them win once in a while, so long as it did not take up too much staff time or money.”62

The primary product of the Office of Mobile Sources technology assessment efforts became a series of semi-annual reports.63 The reports were built through two processes: through what economists sometimes call “search” and through various in-house studies and experiments. In the first case, the EPA staff familiarized themselves with all available knowledge on emissions control, including emerging knowledge coming out in such venues as journals and professional societies. Austin and Hellman and their staff carried out many, various studies to ground the reports.

61 Email from anonymous source. 62 Email from Eric Stork, December 31, 2009. 63 Many of these reports are very hard to locate and belong to what some archivists and researchers call the “grey literature,” ephemeral publications that are rarely collected in “official” depositories, such as libraries and archives. In the case of the technology assessment reports, the very presentist attitude of many involved—including the EPA engineers and the automakers—created an atmosphere where the newest report replaced the previous one and rendered the previous one obsolete. For one such technology assessment, see Emission Control Technology Division, Mobile Source Air Pollution Control, “Automobile Emission Control—The Development, Status, Trends, and Outlook as of December 1976: A Report to the Administrator U.S. Environmental Protection Agency,” April 1977.

228

The clearest marker of Stork’s willingness to move towards the technology assessment model and away from R&D came later, when Stork voluntarily let go of the Advanced Automotive Power Systems (AAPS) division. When the Federal

Energy Agency FEA was created in 1975, it tried to centralize all federal energy programs under its banner (much like the Department of Transportation had done with transport-related programs in 1966 and the EPA had done with environmental programs in 1970). Stork recalls that the FEA was particularly interested in getting

EPA’s clean coal program, but the EPA wanted to keep it. In a meeting with Russell

Train, who was at that point the EPA’s administrator, Stork suggested that the EPA might trade away the AAPS program in order to keep the clean coal program. For

Stork, handing off AAPS solved a number of problems: first, he believed that its “pie in the sky” research conflicted with his office’s basic functions—making and enforcing regulations. Second, AAPS had a dedicated budget of $20 million, more than the other divisions within Mobile Sources combined. AAPS was the tail that wagged the dog. Third, AAPS used a great deal of that money to pay for contracts, which meant that Stork had to worry about fraud and conflict-of-interest. Finally,

John Brogan, the head of AAPS, knew about Stork’s feelings about research and wanted to go to the FEA, where he believed his work would be more appreciated.

After making the move, Brogan continued heading the FEA’s corollary of AAPS, the

Vehicle Propulsion Program, at the FEA’s successor agency, the Department of

Energy, well into the 1990s.

Stork took a strong, adversarial stance against the auto industry. While he became friendly with several auto company employees, occasionally going out to

229 dinner with them while at hearings or conferences, he could also be harsh with them when necessary. He had his “hippie” daughter cross-stitch a tapestry that hung above desk (it now hangs in his home in Arlington, Virginia).64 In Latin, it read,

“Orchides eôrum trahite cordês et mente venient” (roughly translated, “If you have them by the testicles, their hearts and minds will follow”). Hearts and flowers surrounded the words. The tapestry hung on his wall until an employee of the auto industry jotted the passage down and took it to get translated by his parish priest.

After discovering its meaning, the man called Stork’s superiors at the EPA, who politely requested that the tapestry be taken down. Eric Stork cultivated a regulatory style that mirrored his understanding of his role: if people hindered his path, they were his adversaries; and whether they sat in the auto industry or the federal bureaucracy, he would cross swords with them. He enjoyed the fight.

Stork’s style was not merely pretense; it determined his management decisions. One way to examine this is to look at how Stork responded to a classic problem in legal and political theory. In the 1930s, the German jurist and legal theorist Carl Schmitt examined the problem of legal ambiguity.65 Long before the advent of “postmodernism” and the prevalent belief in interpretive indeterminacy,

Schmitt argued that all laws have “exceptions,” difficult cases where it is not clear how to apply the law’s rule. He believed that, in such moments, it was crucial for a

“sovereign”—that is, a leader—to settle the dispute. The sovereign had to decide

64 Interview with Eric Stork, April 10, 2010. 65 Carl Schmitt, Political Theology: Four Chapters on the Concept of Sovereignty, trans. George Schwab (Cambridge, Mass.: MIT Press, 1985). In the last twenty years, Schmitt has had a marked influence on a number of leftist political theorists who were influenced by postmodernism and the “linguistic turn” in humanities. Chantal Mouffe, The Return of the Political (London: Verso, 1993); idem., ed., The Challenge of Carl Schmitt (London: Verso, 1999).

230 how the law would be read. Thus, scholars often characterize Schmidt’s position as a form of “decisionism.” But in his idealized formulation, there would only be one, great leader who would settle exceptions. (Unsurprisingly, Schmitt is widely regarded as the premiere legal theorist of National Socialism.) In examples such as the federal bureaucracy in the United States, exceptions are too bountiful to be decided by any single person, and, therefore, sovereignty and leadership are widely disseminated throughout different levels of government.66 Often decisions are only bumped up to a higher level if they could potentially threaten that superior position.

A choice that would naturally rest with a lower-level office but which could spell political trouble for the agency’s upper echelons is often made by the agency’s

Administrator. Similarly, if an agency is in charge of a given issue but the decision could cast a negative light on the President, the President often decides which way the hatchet falls. Moreover, if the issue is particularly unsavory, the superior may order their subordinate to announce the decision.

When an ambiguity in the Clean Air Act arose or, more often, if the automakers postulated such an ambiguity to increase their strategic position, Stork would issue so-called “advisory circulars,” papers that explained how the agency was interpreting the law.67 The advisory circulars did not merely announce the

66 Surely, it is difficult to imagine a case where one person could decide all of the exceptions in any legal system. 67 The advisory circulars are now available online at http://iaspub.epa.gov/otaqpub/pubadvsearch.jsp (Last accessed March 4, 2011). Business historian Joann Yates has written about how corporations managed communication to increase coordination and efficiency. While we can interpret Stork’s advisory circulars as a form of adjudication, we can also view it as a way of managing regulatory processes. In this case, the seemingly strict line between the state and private businesses breaks down, and both sides can be seen as part of a single process that requires coordination. In another example of management via issuance, Stork created papers known as “fact sheets” to manage public inquiries. Federal agencies typically receive huge

231 agency’s opinion about what the statute meant. They effectively determined the law.

In this instance, within limits, meaning followed power. Stork’s use of advisory circulars partly relied on the structure of the CAA amendments of 1970. By contrast, because of the way the Traffic Safety Act was written, the NHSB/NHTSA had created the protocol for entering rule-making procedures in order to set its own standards. Congress did not set these standards for the agency. Yet, with this rule- making habit firmly ingrained, NHSB/NHTSA also began relying on the process in order to handle any disagreements about the law, trying to work out the interpretive difference through “democratic” processes of eliciting opinions and even holding hearings. Stork and the EPA humored no such considerations. The

Clean Air Act amendments set the standards for them, and they were not in the habit of commencing rule-making proceedings, even informal ones.68 When questions arose, the agency announced how the law was going to be interpreted. Stork took it upon himself to issue the advisory circulars, though he doubtlessly discussed the issues at hand with EPA Administrator William Ruckelshaus and his (i.e.,

Ruckelshaus’s) assistants. Stork signed all of the circulars, however. This was partly because, like any good bureaucrat, Stork followed an unwritten rule: if his office had good news, it would send the information up the ladder for the EPA Administrator

numbers of inquiries either directly from the public or, more often, they are forwarded by legislators for their constituents. These inquiries can easily overwhelm the agency office place. In an effort to stem the tide, Stork and his subordinates wrote the fact sheets, which summarized the agency’s position on certain topics and the current state of knowledge on technical and scientific issues. The fact sheets have been collected in a binder that is housed at the EPA’s National Vehicle and Fuel Emissions Laboratory library in Ann Arbor, Michigan. A complete set of the documents is in the author’s possession. 68 Indeed, the agency’s lack of rule-making efforts partially explains the paucity of extant archival records from the era, since the vast majority of the NHSB/NHTSA’s archival sources take the form of “dockets,” the paper trail of rule-making procedures.

232 to announce. If bad news was the order of the day, Stork and the office would take one on the chin. In 1978, after Stork had left the EPA, the agency’s Office of General

Counsel ordered that advisory circulars should be discontinued because of their questionable legality.69

In the history and analysis of business, capacity building is associated with the question of strategy. How should a business executive develop her firm in order to have the greatest competitive advantage against other firms? Traditionally, however, studies of “state capacity” have been less oriented around studying strategic thinking. Yet, Stork based his decision to move the agency away from R&D and toward technology assessment on his understanding of the office’s strategic interest—and “competitive advantage”—in the regulatory environment. Stork’s regulatory style was highly dependent on the office’s organizational capabilities and the everyday labors of his subordinates. His strategy was to use technology assessment and analysis to ensure that he was as knowledgeable as his opponents and, thus, to frame his decisions as impeccably following the best scientific and technical knowledge available. He would not allow the automakers to outflank him in the public’s eye. Indeed, this may have been the primary role of the agency’s daily routines in Stork’s view.

Expertise Comes to the NAPCA and the EPA

As Rourke and Jones explained in their respective works on bureaucratic power, recruiting expertise is absolutely essential for any agency or organization.

69 Existing advisory circulars were updated through the early 1980s, but the agency issued no new advisory circulars after Stork left in 1978.

233

But this is one of the hardest challenges for government organizations, especially for regulatory agencies. Industry can consistently outbid the government in terms of salary and benefits for the employee. Yet, the Office of Mobile Sources did not have a great deal of trouble with recruiting. More oddly, the agency won a great deal of young, hot talent—exceptionally bright men and women who were dedicated to the cause of curbing automotive emissions. As Stork said, besides George Kittredge and

John DeKany, who were Stork’s age, “All the others were just young kids.”70 In this section, I briefly outline the biographies of six of the Office of Mobile Source’s young staff members in order to show how expertise came to the agency. While half of the people described here eventually moved to other positions either inside or outside the EPA, the others stayed, and either continue to work there or worked there until their retirement.

These young experts began arriving at the agency before Stork did. One of the things that brought young people into the agency was that, because the NAPCA was part of the PHS, they could enter the agency as Officers of the Commissioned

Corps, thereby, avoiding military service in the Vietnam War. Inevitably, some of these recruits drifted off to other opportunities over time, but a surprising number of them remained personally committed to the EPA. Indeed, many of them remained in automotive emissions control at the agency. Charles Gray was one such employee, who came to the NAPCA in 1970 after earning a Bachelor’s in mechanical engineering from the University of Mississippi. Although Gray did not have an advanced degree, like many of his peers did, he had a prodigiously talented mind for

70 Phone interview with Eric Stork, February 25, 2011.

234 technical issues. In 1976, John DeKany appointed Gray to head the regulations branch of the emissions control technology division, and Gray played a central role in passing a number of regulations during his tenure there. Gray was always interested in research, especially in alternative fuels and energy efficiency. In the early 1980s (when the Reagan administration made passing new regulations virtually impossible), Gray turned seriously to studying the technological possibilities of methanol. He eventually co-authored a book on the topic.71 In the

1990s, Gray was one of the chief researchers in the so-called Supercar program, which was a collaborative project between industry and the EPA.

Joseph Somers was another staff member who was originally hired through the PHS Commissioned Corps. He also came to the NAPCA in 1970 at the age of 25, after just having completed a Ph.D. in chemistry at Duke University.72 Somers quickly began working on emerging science issues, including characterizing exhaust emissions and unregulated (but dangerous) emissions. He was one of the central people involved in the study of sulfate emissions that I discuss later in this chapter.

Over time, Somers’ work moved him increasingly into the field of health effects research. Somers continues to work on emerging science issues in auto and fuel emissions control at the EPA’s Ann Arbor.

As Stork moved the agency towards a focus on technology assessment, two men took the helm of these efforts under John DeKany’s direction—Thomas C.

71 Charles L. Gray and Jeffrey A. Alson, Moving America to Methanol: A Plan to Replace Oil Imports, Reduce Acid Rain, and Revitalize Our Domestic Economy (Ann Arbor: University of Michigan Press, 1985). 72 Joseph Henry Somers, “Conformational Properties of Some Phosphorinane Derivatives,” PhD diss., Duke University, 1970.

235

Austin and Karl Hellman. (Neither of these men, nor the rest of the people who I discuss below, entered the agency through the Public Health Service.) In automotive emissions control circles, their annual publications became known as the “Austin-

Hellman Reports.” Stork later joked that people used to think that Austin-Hellman was one person. Austin received his B.S. in Mechanical Engineering (Automotive) from the University of Michigan in 1969.73 From 1969 to 1971, he worked as a

Mechanical Engineer for the U.S. Army Tank-Automotive Command (TACOM).

TACOM’s primary headquarters are in Warren, Michigan, and Austin was based there. At TACOM, Austin had already begun working on projects that would define the rest of his career. He focused on energy efficiency projects on military vehicles, including a jeep modified with a stratified charge engine. He also worked on a project that sought to develop emissions controls for jeeps. While developing these systems, he would come to the EPA’s labs to have them tested. Eventually the EPA staff asked him to join the agency. In 1971, Austin came to the EPA and worked as project manager in what would become John DeKany’s office, the Emissions Control

Technology Division, where Austin focused on technology assessment. In 1975,

Austin left the EPA and went to work for the California Air Resources Board, first as a Deputy Executive Officer and then, from 1978 to 1981, as the Executive Officer. In

1981, Austin co-founded a consulting firm, Sierra Research, where he continues to work. Karl Hellman grew up in Gary, Indiana. He received a Bachelor’s in

Engineering Sciences at Purdue University before entering Northwestern University,

73 Austin’s biographical information is drawn largely from his resume, which is available online: http://www.sierraresearch.com/profiletca.htm.

236 where he completed his Ph.D. in Mechanical Engineering in 1970.74 At that point, he came to the NAPCA and stayed on when it became part of the EPA. Hellman spent the majority of his time at the EPA doing various technology assessment work. But the Office of Mobile Sources was an incredibly dynamic place during the 1970s, and staff members could be called upon to work on whatever project was needed at the moment. As I will discuss in chapter six, Austin and Hellman became very involved in fuel economy measurement during the early-to-mid 1970s, and Hellman continued to develop expertise in that area after Austin’s departure. Hellman later became the branch chief in charge of characterization and technology assessment, where he continued to lead teams that produced annual technology assessments.

He remained at the EPA until the mid-2000s when he retired.

Stork also required experts and fixers in his Washington, D.C., office. During his management review of the NAPCA’s automotive programs, Stork went to each of the agency’s divisions in Ann Arbor and asked the employees what the division’s goal was. At one of these sessions, however, a young employee, Joseph “Joe”

Merenda, instantly seemed to understand what was being asked, while most others scratched their heads. Stork found Merenda to be smart, and he soon recruited

Merenda to become his technical assistant in Washington, D.C., robbing the Ann

Arbor office of one of its best and brightest. Merenda finished a Master’s in chemical engineering at Stanford before coming to the PHS in 1970.75 He’d been interested in pursuing a Ph.D. in chemical engineering focusing on air pollution issues, but when

74 Karl Harry Hellman, “A Kinetic Theory Analysis of the Evolution of Shock, Thermal, and Expansion Waves Generated by an Impulsively Moved, Fractionally Accommodating Piston,” PhD. diss., Northwestern University, 1970. 75 Interview with Joseph Merenda, Rockville, Maryland, August 16, 2008.

237 his thesis advisor tried to push him in a different direction, Merenda left and went to work for the NAPCA. Stork often relied on Merenda when he needed quick advice on technical issues. Merenda wrote many policy issue papers on a variety of topics for Stork. In 1977, Merenda moved to the EPA’s new Office of Pesticides and Toxics

Substances (which arose from the Toxic Substances Control Act of 1976).

In early 1972, Janet “Jan” Lane came to work for Stork’s office in Washington,

D.C. Stork often refers to her as his “right hand man.” She had come to the

Washington area to work on a Ph.D. in American diplomatic history at Georgetown

University.76 But when her fellowship money ran out after three years, she began working at the Agency for International Development (AID) and doing her doctoral work at nights and on weekends. A friend at AID, who knew Stork, called him trying to get a job for another person (not Lane). Stork said he couldn’t help that person out, but he did have a vacancy. When Stork described the job, the friend thought of

Lane. Lane talked to Stork, and he “worked the system” so that within a week Lane was working for him in Washington.77 Lane began to write regulations for Stork.

She was one of the key managers of the “advisory circulars” and compliance issues.

Stork told Lane that he did not want a lawyer, economist, or engineer; he wanted someone who could write.78 Lane was the first female staff member in Mobile

Sources. In 1974, she was put in charge of budget and program planning. In 1978— a month before Stork left Mobile Sources—Lane became a supervisor at the EPA’s

76 Janet A. Lane, “United States-Mexico Diplomatic Relations, 1917–1942,” PhD. diss., Georgetown University, 1971. 77 Interview with Janet Auerbach, April 18, 2010. Lane eventually married another EPA staff member named Irwin Auerbach, changing her name to Janet Auerbach. 78 Ibid.

238

Office of Solid Waste, where she worked for a year and a half before moving on to the Office of Pesticides and Toxics. (John DeKany also moved to Pesticides and Toxic

Substances during this time. Between Lane, Merenda, and DeKany, one could say that there was small exodus out of Mobile Sources to Pesticides and Toxics when the

Carter administration arrived. Part of this exodus was flight from the incoming regime, but the toxics office was also a startup operation and had an aura of excitement and “organizational verve.” ) Lane stayed with Pesticides and Toxics until 1992, when she moved to the Drinking Water office, where she stayed until she retired in 2002.

Finally, we come to the issue of “organizational vitality,” or the verve that sustains and encourages an agency’s organizational culture, and its reproduction.79

The EPA’s Office of Mobile Sources was not lacking for verve—energy was in the air, like jazz. The environmental movement was palpable; staff members rode off that with missionary zeal. Stork was, by all accounts, a demanding and exacting supervisor. His personality may have caused some to wilt, but they quickly quit or were moved to non-essential positions. Most, however, found Stork’s temperament to be invigorating. As will become clear in this and succeeding chapters, the Office of

Mobile Sources was marked by a feeling of “flying by the seat of one’s pants.” The staff members had responsibilities and daily routines, but if a problem arose, they could and would turn on a dime to address the crisis at hand. This environment

79 Sociologists and other social theorists, such as Talcott Parsons, Pierre Bourdieu, and Louis Althusser, have long examined how societies reproduce their values and norms from generation to generation. A related, if subsidiary issue, is how organizations maintain and reproduce their values and morale on a shorter time-scale, namely from day-to-day.

239 created a feeling of excitement. A corollary to this vitality was that staff members worked long hours and often worked weekends. Stork himself recalled, “I worked six days a week and half a day on Sunday. I worked from seven to seven and then came home and passed out in my armchair, much to the consternation of my wife and children.”80 One former, long-time employee suggested that the divorce of one of his peers may have arisen from these long work hours.81 The men and women of the Office of Mobile Sources worked passionately. Their passion went largely unpaid—none of them earned overtime pay—but it yielded personal rewards, a palpable organizational vitality, and the satisfaction of eventually breathing cleaner air.

Three Trials

Throughout the Office of Mobile Sources’ early years, it experienced a number of organizational trials. Here, I have chosen three such moments to illustrate the different kinds of difficulties an organization can experience. In the first case, higher powers within the EPA tempered the strict requirements of the law. Stork believed that he would have to prevent the production of Ford Motor Company’s next model year vehicles because the company had cheated the emissions control certification process. Yet, EPA administrator William Ruckelshaus maintained that the agency had to figure out some way to allow the company to begin production. He simply was not going to be responsible for thousands of people being laid off. This case

80 Phone interview with Eric Stork, February 25, 2011. 81 I heard this conjecture during an interview, but the need to protect both the anonymity of the person who made it and the person who experienced the divorce goes without saying.

240 makes an important point: superior powers—including people placed higher in the agency itself; people within external guiding agencies, such as the Office of

Management and Budget; or the White House—often keep an agency from following the strict letter of the law, whether because they are taking into consideration wider social and economic issues or for other reasons, including pressure from industry.

(This point became more important later, under the neoliberal programs of Ronald

Reagan, who attempted to stymie new regulatory rule-making and many regulatory enforcement efforts.) In the second case, the automakers were cheating—or at least

“gaming”—the regulatory system, and the Office of Mobile Sources found a way to respond, in spades. Finally, other organizational entities within the EPA often challenged both the turf and the practices of Stork’s office. Mobile Sources had to find creative ways to retain both its credibility and its hold on certain programs.

Taken together, these cases demonstrate the kinds of struggles that any regulatory agency can face and the various strategic responses available to them. Succeeding chapters, likewise, spell out representative challenges to bureaucratic practices and authority and the strategies that reside at hand.

The Ford Testing Scheme

On May 16, 1972, staff members from Ford Motor Company’s environmental office came to the Office of Mobile Source’s Ann Arbor office wanting to speak to Eric

Stork.82 He had just flown to Michigan from Washington, D.C. He told them that he

82 Interview with Eric Stork, April 10, 2010. Government Accounting Office, “Examination into the Adequacy of the Environmental Protection Agency’s Motor Vehicle Certification Activities,” June 12, 1972.

241 would be happy to meet with them but that he had a number of meetings planned for the morning. He had to catch up with his staff members. The Ford employees insisted that they had to meet with him immediately. Stork was tired and in no mood to be diverted from his plans. Yet, the Ford employees would not relent.

Though frustrated, Stork agreed. Then, the Ford staff members then dropped a bombshell: they had discovered that Ford’s testing lab had been falsifying certification data.

The Car Product Development Group (CPDG), a subdivision of Ford’s Engine and Foundry Division (E&F), was responsible for testing, and it reported its findings to Ford’s Automotive Emissions Office, which was organizationally separate from

E&F and which worked closely with the EPA on certification paperwork.83 The CPDG had effectively kept two sets of books—one for itself and one that it sent to the

Automotive Emissions Office, the data in which were, thus, reported to the EPA. The latter set of data failed to record that Ford employees had completed hundreds of acts of maintenance on test vehicles. Yet, the CPDG did record all of the unreported maintenance in its computer system. For some time, E&F had been complaining that it received “inadequate support from Ford’s Central Computer Group.”84 The computer group or its superiors finally capitulated and decided to make a report of

CPDG’s computer records. When it printed the data, however, the group realized that there were two sets of records and that they diverged wildly. It reported the discrepancy to the vice president charged with emissions control on May 11. For

83 GAO, “Examination,” 4. 84 Ibid.

242 five days, Ford officials debated how to handle the situation but finally decided to report it to the EPA and to withdraw its certification applications.

The law stipulated that the car could receive a significant tune up only at

25,000 miles. If a vehicle required unscheduled maintenance—say, a part on the car broke—the agency had to request permission to carry out “unscheduled maintenance” from the EPA. The agency usually granted permission because, after all, the CAA did not require all of the automobile’s systems to last for 25,000. If something “normal” had malfunctioned, it should be fixed.85 Yet, the Ford employees had been jiggering with the engines during the entire testing process, changing spark plugs, filters, and other parts. While this maintenance may seem like minor details, they made the cars run much more smoothly and, thus, produce fewer emissions than they normally would have. Over the course of the succeeding weeks, the number of instances of maintenance to which Ford admitted increased. Later, as the Government Accountability Office Reported, “Ford identified 442 instances of unauthorized maintenance on 26 test vehicles.”86

Stork was dumbfounded. He believed that he would have to hold up the company’s production of 1973-model year cars. The certification process took nearly a year, but only about four months remained until the company would be making the switch to the new model year vehicles. Stork did not see any way around holding up production, but first he had to talk to his superiors in Washington, D.C.

85 The distinction between what I am calling a “normal” malfunction and an abnormal malfunction is an important one. For example, it could be fairly typical for a given engine to foul spark plugs and render them ineffective, which would greatly increase emissions. The EPA would take such a problem very seriously, but if a broken system was not directly related to emissions control but was keeping a test from going on, the agency would allow a repair. 86 GAO, “Examination,” 4.

243

Ford’s impropriety was the latest in a line of company missteps regarding federal air pollution regulation. In 1970, Ford ran behind on its certification efforts on 1971-model heavy-duty vehicles, mostly large trucks, and asked the EPA for conditional certification with the understanding that Ford would modify any engines that were found to fail the standards.87 The EPA granted conditional certification in early January 1971 but revoked certification for two engine models eleven days later. The next year, Ford shipped 200,000 cars before they were certified. The company later signed a consent decree with the Department of Justice and paid a $10,000 fine, or, as one government report cynically put it, “about 5 cents per vehicle.”88

Ford executives had discovered the internal deception around May 11.89

They characterized those responsible as “supervisory technical employees,” which, for the next nine months, became the standard term for those at fault.90 Ford went to the press about the problem on May 23. A day later, Ford reassigned the four supervisory technical employees, though spokespersons refused to say where. To the degree that executives were originally unaware of the lower-level employees’ subterfuge, it reveals, once again, that corporations are not monolithic entities. The company had trouble getting its cars certified a year earlier, and the employees were under considerable pressure to get the cars passed. Under such circumstances, the incentives to deceive were high even within the company. As

87 Ibid., 3. 88 Ibid. 89 Jerry M. Flint, “Test Fault Perils Sale of ’73 Fords,” New York Times, May 23, 1972. 90 “Ford Reassigns 4 in Faulty Tests,” New York Times, May 24, 1972. “Ford ‘Reassigns’ Employees in Test Snarl, Hopes to Market ’73 Models on Schedule,” Wall Street Journal, May 24, 1972.

244 one government report on the scandal stated, “Ford officials told us that [the division in charge of testing] had a vested interest in ensuring that the vehicles met

Federal emission standards.”91 The division would be on the hook if the cars did not pass. Ford executives did not have adequate control over the certification division.

In the midst of the controversy, Ford changed its organizational structure such that the certification testing labs reported directly to the company president,

Lee Iacocca.92 It also appointed a new director of certification, William A.

McConnell.93 Certification testing had been a part of Ford’s Engine and Foundry

Division, but, on May 23, the company reorganized, and certification became a part of the Environment and Safety Engineering Staff.94 Certification’s original placement in the Engine and Foundry Division further explains the deception. The

Division was one of Ford’s oldest. It was a product of pre-regulation Detroit. Many in the Division had been Ford employees for decades. They had grown used to a certain culture and institutional arrangement, and now the federal government was coming in and telling them what to do. If there was low-level resentment against federal intervention, it would have nested here. The Environment and Safety

Engineering Staff, however, were mostly fresh faces in the company. Most of them believed in the cause of making the car safer and cleaner, regardless of their political views of regulation. After the testing scandal, Ford created a strict line between these two entities. As one company memo read,

91 GAO, “Examination,” 4. 92 Agis Salpukas, “Court Fines Ford $7-Million in Suit on Pollution Test,” New York Times, February 14, 1973. 93 Charles Camp and Walter Mossberg, “Sniffing at the Tailpipe: Fuss at Ford Shows How Weak Enforcers Are,” Wall Street Journal, June 6, 1972. 94 GAO, “Examination,” 4–5.

245

“Once a vehicle is transferred to the Environmental and Safety Engineering Staff for certification mileage accumulations and testing, neither the Engine and Foundry Division nor any other organization will touch the car or have access to it except as directed by the Environmental and Safety Engineering Staff."95

Some of the steps that Ford took during the controversy—including locking the test vehicles’ hoods to “prevent tampering” and storing vehicles in guarded areas during cooling periods—suggest both the pressure that certain divisions felt and the lengths to which the company was willing to go to build trust with the EPA.

When Stork went to Ruckelshaus, he was still dumbfounded. He told

Ruckelshaus that he did not see anyway around holding up the production of 1973- model year cars until the certification was done. Ruckelshaus was more circumspect, however. If the auto plants closed down, nearly 170,000 people would be temporarily laid off, and, as the Wall Street Journal pointed out, these lay offs would come during a (presidential) election year.96 For Ruckelshaus, pressure from the White House was a mixed bag:

There was pressure from the White House, particularly towards the end of the Nixon first term, to back off on some of these environmental issues. I didn’t feel pressure from the President. I knew that it was a political issue obviously, and, therefore, we had better have a good record. And as far as I was concerned, having a good record meant that we understood and enforced the laws as they were written and that we were striking the proper balance between the environment and other social concerns.97

Ruckelshaus tried to handle the election year problem by distancing his decisions from the Nixon Administration, framing his choices as if they were his alone, and

95 Ibid., 5. 96 Charles Camp and Walter Mossberg, “Sniffing at the Tailpipe: Fuss at Ford Shows How Weak Enforcers Are,” Wall Street Journal, June 6, 1972. 97 Phone interview with William Ruckelshaus, February 20, 2009.

246 putting in the work to ensure that his decisions were based on the best available reasons, science, and information. Yet, it is hard to imagine that the election year did not influence Ruckelshaus’ thinking.

Stork recalls that Ruckelshaus told him to find any solution that did not involve shutting down Ford.98 Ruckelshaus does not remember saying this, but he believes it is consistent with his regulatory philosophy. First of all, Ford had come forward and reported itself, and there was precedent for reducing a penalty if a firm turned itself in. Second, Ruckelshaus believed (and continues to believe) that agencies had to take account of the economic impacts of their regulations:

It seems to me you have to do what is right for the whole, for the economy, recognizing that somebody may get off in the process. And I think the obligation of federal officials is really to the whole. You’ve got to develop the capacity to step back and say, “Wait a minute, what’s the effect of what I’m doing here?” And the EPA has the power—particularly it had the power in those days—of enormous significance to the economy of the country. And we just had to be careful about what we did.99

Moreover, he claimed,

The legitimate thing for me to have been concerned about was the discrediting of the whole regulatory effort to improve the environment or reduce pollution because we had these drastic economic effects that could not be justified on the basis of improved health, or at least in the public’s minds. So we had to look at what we were doing and its impact not just on the individual firm or on the EPA’s image as much as it might undercut public confidence that the government was acting wisely and in their interest, and, therefore, discredit their support for other environmental requirements necessary to make sure that public health and the environment were protected. I think that’s a legitimate concern.100

98 Interview with Eric Stork, Arlington, Virginia, June 11, 2008. 99 Phone interview with William Ruckelshaus, February 20, 2009. 100 Ibid.

247

EPA staff members were also aware that they were pulling off a balancing act: they could alienate the public by enraging environmentalists or by generating huge, temporary unemployment. One EPA official asked the Wall Street Journal, “How would it look if we whitewashed this and then some Congressman uncovered a grand conspiracy?”101 In this context, the EPA had to take into account both the environment and the health of the firms it regulated.

On May 24, both Ford and the EPA told the press that it would be difficult for the company to release its 1973 line of vehicles as planned. In response to Ford’s indiscretion, the EPA had changed its procedures but only for dealing with Ford’s retest. The new procedures, including examining test vehicles prior to testing and carrying out “spot-checks,” simply could not be applied universally for want of resources. Ford’s retests alone would entail over 1,300 hours of overtime at EPA, a huge cost for a cash-strapped organization.102 The EPA told the GAO that it was considering a number of new testing systems: it might “(1) make unannounced spot inspections of the auto companies' records and test facilities, (2) station inspectors at the auto companies' test facilities to provide continuous monitoring, or (3) assume responsibility for some or all testing and mileage accumulation of the companies' prototypes.”103 But it was nearly impossible to see where resources for any of these developments would come from.

The EPA and Ford worked out a complex arrangement. Ford would continue producing ’72-models beyond the planned end date to keep its people working and

101 Charles Camp and Walter Mossberg, “Sniffing at the Tailpipe: Fuss at Ford Shows How Weak Enforcers Are,” Wall Street Journal, June 6, 1972. 102 GAO, “Examination,” 14. 103 Ibid.

248 products moving. Then, it would begin producing ’73-models before the certification was complete, but these cars would only be shipped to dealers lots and stored, not sold. Furthermore, Ford agreed to more stringent certification procedures for the next decade. Finally, the Justice Department would fine Ford seven million dollars. A United States Attorney claimed that the size of the fine helped “demonstrate that the Government means business in enforcing the provisions for the Clean Air Act.”104

Inevitably, however, environmentalists and the popular press interpreted the settlement as a sign of the EPA’s feebleness. As a subtitle of a Wall Street Journal article read, “Ford Imbroglio Shows Clean Air Enforcers How Weak They Are.”105 It went on to say that, though the automakers cast Washington as a “growing regulatory monster,” the testing scandal demonstrated that “the monster is in truth a toothless tiger.” The settlement can be interpreted in another way, however. Stork now believes that Ruckelshaus made the right decision in allowing Ford to continue producing cars, but Ruckelshaus did not follow the letter of the law in making his decision. . In either case, Ruckelshaus effectively constricted the Office of Mobile

Sources’ practices and evaded a strict interpretation of the law. To the degree that

Ruckelshaus may have felt pressure from above to make this decision, the limitation of practice extended to the White House.

104 Agis Salpukas, “Court Fines Ford $7-Million in Suit on Pollution Test,” New York Times, February 14, 1973. 105 Charles Camp and Walter Mossberg, “Sniffing at the Tailpipe: Fuss at Ford Shows How Weak Enforcers Are,” Wall Street Journal, June 6, 1972.

249

Defeat Devices

Sometime in early June—or possibly May—1972, technicians from the

American Motors Corporation brought two cars to the EPA’s Ann Arbor labs to be certified. The cars were pulled into the testing room, put on dynamometers, and had hoses that collected the vehicles’ emissions connected to the tailpipes. The AMC cars failed the test. They produced NOx levels above the federal limit. The AMC technicians were displeased and stood scratching their heads. “It must have been a bad sensor,” one of the AMC techs suggested.106 “What bad sensor?” asked an EPA staffer. The answer to that question created a controversy that hung over the agency for half a year.

On June 15, R. J. Belanger, a Supervisor at AMC, sent a letter to Ed Brune,

Director of the Office of Mobile Sources’ Division of Certification, explaining what had gone wrong.107 The AMC vehicles contained compensating valves that turned the cars’ exhaust gas recirculation (EGR) systems off under extreme heat circumstances. After these cars went through a 4,000-mile test at the company, as

Belanger wrote, “their respective compensator valves [were] removed for calibration documentation.”108 AMC technicians then put the wrong compensators back on the vehicles. As Belanger explained, “The released hardware for this application has a low end temperature tolerance of 115° F, as compared to the 100°

F actuating temperature of the units which actually ended up as being tested at the

106 Interview with Eric Stork, April 10, 2010. 107 R. J. Belanger to E. J. Brune, June 15, 1972. The Belanger letter is included in a binder in which Eric Stork collected documents related to the issue of “defeat devices.” The author possesses a complete copy of the documents in this binder. 108 Ibid.

250

EPA.”109 Belanger noted that the 100° F units were “the original hardware considered for this task,” but the company realized that these lower temperature units turned off the Exhaust Gas Recirculation too often “for the specific reason you have just experienced.”

For the EPA, however, this event opened up a whole can of worms. In the instance of the AMC case, 15° F was not a wide margin in a mechanical environment that often experienced hundreds and thousands of degrees. But the wider problem was one of “bounded rationality,” or limits to current knowledge.110 What if other automakers were putting similar systems on their cars? The Office of Mobile

Sources simply did not know the extent of this technology.

On July 7, 1972, Brune wrote a memorandum summarizing his division’s experience with automotive technologies that the EPA called “defeat devices.”

These devices governed emissions control systems on automobiles, and, typically, they caused the systems to turn off under certain operating conditions. Brune listed three reasons that automakers put defeat devices on the vehicles, but he hedged that they did so “basically for two reasons and possibly a third”:

1. To prevent engine or vehicle damage during extreme operating conditions [when components become very hot]. 2. To reduce control system effectiveness under ambient or operational conditions which are characteristically low emission modes in order to improve engine economy and/or performance.

109 Ibid. 110 Herbert A. Simon, Administrative Behavior: A Study of Decision-Making Processes in Administrative Organization (New York: Macmillan Co., 1947).

251

3. To reduce control system effectiveness under ambient or operational conditions which are not represented on the Federal test in order to improve engine economy and/or performance.111

Brune hedged because this third reason for using defeat devices amounted to a de facto violation of the Clean Air Act. It meant that automakers were creating cars that would meet the narrow demands of the current automotive emission Federal

Test Procedure but that would shut down emission controls under conditions that fell outside the procedure. This was a serious accusation, which partly explains why

Brune was being so careful. The Federal Test Procedure in 1972 required tests to be carried out in conditions with air temperatures between 68º–86°, a very narrow temperature band.112 EPA officials had mandated this requirement so that tests were repeatable under typical laboratory conditions and, thus, the agency’s decisions based on the tests would be unassailable. No one could point to errant lab conditions and say, “This ambient air temperature explains and dismisses your finding.” Yet, if automotive engineers were designing their emission controls to work only within this narrow band, they could be designing their devices to beat the test while not fulfilling the spirit of the CAA. As Brune wrote, “The Division of

Certification and Surveillance is concerned that some of these ‘part-time’ control systems may technically satisfy the regulations, but clearly violate the intent of the law.”113

111 Memorandum, E. J. Brune to Director, MSPCP, EPA, OAP, “Characterization and Use of Emission Control Systems That Operatoe Only Under Specific Ambient Conditions or Vehicle Operational Modes,” July 7, 1972, 1. 112 Ibid., 2. 113 Ibid., 1.

252

As Brune noted, “The previously used seven-mode test cycle practically encouraged the use of these ‘part-time’ control devices. The seven-mode cycle and accompanying procedures measured vehicle emissions only during select portions of the test.”114 This test undoubtedly increased incentives for the automakers to game the system; they could simply design the automobiles to function in low- emission modes during those particular portions of the test. The new federal test procedure that the EPA adopted in 1972, however, “sampled [emissions] throughout the entire test.”115 While this improved test closed some loopholes, others may have remained.

In their now classic essay “The Social Construction of Facts and Artifacts: Or

How the Sociology of Science and the Sociology of Technology Might Benefit Each

Other,” Trever Pinch and Wiebe Bijker discuss the notion of “interpretive flexibility,” wherein understandings of technological artifacts and their design can differ between individuals and interpretive communities.116 The interpretive flexibility surrounding defeat devices was this: automakers cast the systems as necessary components that guaranteed the safety and functionality of the car, while members of the Office of Mobile Sources claimed that they broke the law.

Eric Stork reports that one auto executive said that he knew the automakers had been beaten when he read the words “defeat device” in the newspaper.117

114 Ibid., 3. 115 Ibid., 3. 116 Pinch and Bijker, “The Social Construction of Facts and Artifacts: Or How the Sociology of Science and the Sociology of Technology Might Benefit Each Other” in The Social Construction of Technological Systems: New Directions in the Sociology and History of Technology, eds. Wiebe E. Bijker, Thomas P. Hughes, and Trevor Pinch (Cambridge, Mass.: The MIT Press, 1987), esp. 40–1. 117 Interview with Eric Stork, Arlington, Virginia, April 10, 2010.

253

Interpretive flexibility had been closed through rhetorical framing; Stork had outflanked the automakers in the national press.118 The EPA had framed the issue such that automaker resistance looked rapacious. If diminishing air pollution was more sacred than “motherhood,” as Senator Eugene McCarthy claimed, then introducing devices that “defeated” pollution control efforts was tantamount to sin.119 But it wasn’t simply a case of the Office of Mobile Sources being the first group to get the message out. The agency excelled at breaking down complex technical topics into stories that could be grasped easily by the average news consumer. Whether called “risk communication” or “technical education,” the end result was the same: the EPA framed its new releases and follow-up interviews in such a way that the recently revealed “regulators” were unequivocally “defeat devices” that broke the letter and spirit of the CAA.

The defeat devices still presented problems for the EPA, however. As Brune stated in his memo, “There is nothing intrinsically wrong or illegal about devices which reduce control system effectiveness under certain conditions. The issues are the integrity of manufacturers in utilizing the defeat devices and the motive behind their use. In short, are the devices intended to ‘beat the test’?”120 It was unclear how

EPA administrators should control defeat devices since they were not explicitly banned in the enabling legislation. Brune outlined four policy alternatives for EPA

118 Douglas Williams, “Ford Asks EPA Approval to Use Disputed Device,” Detroit Free Press, December 22, 1972; Jerry M. Flint, “Auto Makers Told to Remove Device,” New York Times, December 16, 1972. 119 Donald A. Colburn, “Washinton’s RX: Will the Cure Be Worse than the Disease?”, Automotive Industries, February 15, 1974, 72. 120 Memorandum, E. J. Brune to Director, MSPCP, EPA, OAP, “Characterization and Use of Emission Control Systems That Operate Only Under Specific Ambient Conditions or Vehicle Operational Modes,” July 7, 1972, 8. Underlining in the original.

254 decision makers: ban defeat devices outright, allow the devices only for “safety reasons,” allow them only if they function during the test (this way their influence would be captured by the test), or allow the defeat devices outright as long as the vehicle passes the test (and regardless of whether the devices activated during the tests).121 He listed pros and cons for each of these options. Neither of the extreme options, banning or permitting the defeat technologies categorically, were ideal. On the one hand, if the devices were banned, automakers would have no way to protect the engine or vehicle body from potentially dangerous overheating and fires.

Moreover, defeat devices made sense under certain conditions where fuel efficiency and vehicle performance would be reduced with such technology but where emissions would not be affected by it. Finally, as with any policy that limited these technologies, an outright ban would require the EPA to define what exactly a “defeat device” was. This task would not be simple. Any definition would probably create further loopholes. Furthermore, it was not clear that the EPA had the legal authority to create such a definition. On the other hand, if the agency gave such technologies a pass, “vehicles [will be] allowed to be ‘test beaters’, i.e., the emission levels measured on the test are not representative of what the vehicle will do on the road under real, normal conditions.”122 As Brune asked, “Why should a vehicle be allowed to pollute under some condition merely because that condition is not represented on the Federal test?”123

121 Ibid., 9-10. 122 Ibid., 10. 123 Ibid.

255

Brune made several suggestions as to how the EPA should resolve the defeat device controversy. The first suggestion was fairly obvious: “Obtain additional information.”124 Yet, this proposal went to the very heart of the EPA’s troubles—a lack of knowledge. Second, Brune argued that a policy should be formulated such that, if an emissions control device worked during testing, it also had to work on the road. Third, Brune pointed out that “no justification for defeat devices which sense ambient conditions has been presented.” The dangers that could be averted through defeat devices were all internal to the engine: if it overheated, multiple systems in the car could be damaged. The only reason that EPA staff members could come up with for why cars had ambient air sensors was to beat the test. Finally, Brune suggested pithily, “Seek legal authority to enforce the formulated policy on defeat devices.”125

At first, Stork handled the matter through an Advisory Circular that he issued on December 11, 1972.126 Stork’s original inclination was to ban defeat devices for the next year, the 1974 model year, which the Advisory Circular spelled out. Additionally, since the automakers could use the same certification information to re-pass a car if the car’s systems were unchanged the next year, any car that had a defeat device on it in the 1973 model year would need to be recertified from scratch.

Stork believed it was best to wait until the model year 1974 because the 1973 model year cars were already in production; banning the devices at that point would have

124 Ibid., 10. 125 Ibid., 11. 126 MSPC [Mobile Source Pollution Control] Advisory Circular, No. 24, “Prohibition of Use of Emission Control Defeat Devices,” December 11, 1972. The Advisory Circular is in the Defeat Control binder and, thus, in the author’s possession. But the complete collection of Advisory Circulars is also available at http://iaspub.epa.gov/otaqpub/pubadvsearch.jsp.

256 been too disruptive. But EPA Administrator Ruckelshaus would not tolerate such systems. In a meeting between Mobile Sources staff and Ruckelshaus, as Stork later recalled, the administrator said something to the effect of, “There is no way I can justify the use of this scam for another half a year—give them the absolute minimum time to produce cars with those devices before they have to come off.”127 Four days later, on December 15, the EPA ordered the automakers to remove the devices, depending upon the system, either by February 15 or March 15, 1973, effectively giving the companies between 60 and 90 days to get the fraudulent systems off of their cars.128

The Sulfate Flap

In late 1972, a contractor working for the EPA, the Dow Chemical Company, found a surprising fact: particulate emissions were higher on vehicles equipped with catalytic converters than on ones that weren’t.129 The Ford Motor Company had provided the catalysts that the contractor had used, and after Ford heard about the particulates issue, it asked to see the filters used in the tests. After doing some testing of its own, Ford found sulfuric acid on the filters. Ford sent a letter to the

127 Email from Eric Stork, July 15, 2009. 128 Press release, Environmental Protection Agency, January 3, 1973, Defeat Device binder. 129 Emission Control Technology Division, Office of Mobile Sources, “Automobile Sulfuric Acid Emission Control—The Development Status as of December 1975: A Report to the Administrator, EPA,” pg. 2-1. The author was given a copy of the report by Eric Stork, but it is also available online at: http://nepis.epa.gov/EPA/html/newtitles.html. (Last accessed February 24, 2011). See, also, Eric Stork, “The United States Experience with Imposing Automobile Emission Standards,” A Paper for the Australian Society of Automotive Engineers, Perth, Australia, September 22, 1976. Given to the author by Eric Stork. In this paper, Stork described the controversy regarding catalytic converters and sulfate particulates as the “sulfate flap,” an unnecessary state of worry created by his adversaries’ imprudent claims. I have borrowed this phrase for the title of this section.

257

EPA on February 5, 1973, warning the agency about the presence of the toxin.130

The greatest concern was that the catalytic converter, which looked like the most likely technology to “solve” the automotive air pollution problem, was going to create an even worse hazard.

Throughout its history, the Office of Mobile Sources did scientific work on furthering knowledge about auto emissions. The core of this work involved characterizing the chemical makeup of emissions, including unregulated but potentially hazardous compounds. In the Office of Mobile Sources, the sulfuric acid issue fit squarely within this mission, and as an assignment, this study fell on the shoulders of Joseph Somers, a chemist who worked in the Office of Mobile Sources’

Emission Control Technology Division. Soon after the EPA received the letter,

Somers began work on characterizing the sulfuric acid problem; this project would take up the majority of his time for the next two to three years.

Unfortunately for the Office of Mobile Sources, another part of the EPA, the

Office of Research and Monitoring (ORM) [which later became the Office of Research and Development (ORD)], also took up the sulfate issue.131 Like the Office of Mobile

Sources, most of the work of ORM’s predecessors in the Public Health Service was

130 Ibid., 2-2. Sulfate was not the catalytic converter’s only chemical by-product that worried regulators. At one point, someone suggested that catalytic converters would also produce hydrogen cyanide (HCN). But Joe Merenda did what Stork called a “back-of-the-envelope equation” and discovered that carbon monoxide would kill a person many times over before HCN got to them. Merenda explained that the “back-of-the-envelope” study amounted to a fast, informal look at the problem but that the solution was quickly apparent. Interview with Eric Stork. Interview with Joe Merenda, July 16, 2008, Rockville, Maryland. 131 The historian Jongmin Lee has begun some fascinating research that views the sulfate issue from the viewpoint of the EPA’s Office of Research and Monitoring. See, for instance, “Regulatory Engineering in Pollution and Health Research: U.S. Environmental Protection Agency’s CHESS, 1969– 1977,” presented at the annual meeting of the Society for the History of Technology, Tacoma, Washington, October 2, 2010.

258 originally performed at the Robert A. Taft Sanitary Engineering Center in Cincinnati.

And also like the Office of Mobile Sources, the ORM transferred the bulk of its work out of Cincinnati to a new primary laboratory in Research Triangle Park in North

Carolina, though the ORM retained some functions in Cincinnati. Some of the ORM’s most important work lay in the Community Health and Environmental Surveillance

System (CHESS), an epidemiological study of air pollution in eight cities. Sulfur oxides were among the air pollutants that the ORM studied, and so the office became very concerned when it heard that automobiles might be increasing the amount of sulfur oxides in the ambient air.132 In September 1973, the ORM announced the results of its research, claiming that, while automobiles had typically emitted the

133 sulfur in gasoline as SO2, a basic precursor of sulfuric acid. This chemical compound would, thereby, combine with water, creating a mist of particulate matter that had deleterious effects on human lungs (and other organic matter as well, especially considering that sulfuric acid was a precursor to what would later become known as “acid rain”). Therefore, catalytic converters were potentially dangerous to the population’s health.

This announcement started a bureaucratic face off within the EPA. The Office of Mobile Sources was politically deeply invested in the catalytic converter. The office’s job was to make sure that the automakers met the emission standards, and its staff members believed that, in the short run, the catalytic converter was the only way to achieve those standards. The office’s credibility was on the line, as was EPA

132 For the Office of Research and Monitoring’s work on sulfur oxides, see Environmental Protection Agency, “Health Consequences of Sulfur Oxides: A Report from Chess, 1970–1971,” May 1974. 133 Eric Stork, “Imposing Automobile Emission Standards.” See also, “Do Auto Catalysts Meet the Safety Test?” Business Week, October 27. 1973, 116

259 administrator Russ Train’s. Meanwhile, ORM believed it was best for the EPA to delay any standard that would force automakers to deploy catalytic converters because of their potential threat to public health. “It’s like watching two armies form on a battlefield,” one EPA staff member said.134

The battle soon went public, with both sides speaking to the press. On one side stood, Stanley Greenfield, the head of ORM and two of his subordinates, John

Finklea, the head of research, and John Moran, a program director at Research

Triangle Park135 Greenfield had worked for the RAND Corporation from 1950 to

1970, acting as a department head for the second decade. He did a two-year stint as a science advisor for the Air Force before coming to the EPA as the Assistant

Administrator of ORM. Finklea was a controversial figure. Possessing a Doctorate in

Public Health from the University of Michigan, he was later accused of overplaying the negative health effects of sulfates in the CHESS studies.136 Finklea left the EPA to become the director of the National Institute for Occupational Safety and Health, a considerable step up the ladder, but he resigned under a pall of organizational discontent. John (Bernard) Moran worked as an automotive specialist at Dow

Chemical Company from 1962 until he joined EPA’s Research Triangle Park lab in

1971. Moran was, by far, the most vocal critic on the ORM side, going so far as to say that he would not buy a catalytic converter-equipped car because of the technology’s potential health risks. Business Week claimed that Stork’s direct supervisor, Robert L. Sansom, the Assistant Administrator for Air and Water

134 “Do Auto Catalysts Meet the Safety Test?” Business Week, October 27. 1973, 116. 135 Ibid. On Greenfield, Who’s Who in America—1994. On Moran, Who’s Who in America—2008. 136 The best current treatment of this story is Jongmin Lee, “Regulatory Engineering in Pollution and Health Research: U.S. Environmental Protection Agency’s CHESS, 1969–1977.”

260

Programs, was “soft-pedaling the dispute, but is expected to back the catalysts.”137

Therefore, on the Office of Mobile Sources’ side, Stork and his staff members stood almost alone.

Stork cast the dispute as a merely bureaucratic issue. He believed that the

ORM was trying to get a piece of mobile sources, which at the time was the most prominent issue on which the EPA worked.138 The ORM had seen its budget slashed for the 1974 fiscal year from $881,000 to $150,000.139 Moran told some congressional staffers about the sulfuric oxide issue while they were visiting

Research Triangle Park. Congressman Paul G. Rogers (D—Florida) then pressured the interim administrator between Ruckelshaus and Train, and Moran was given $2 million to study the relationship between catalytic converters and sulfuric oxides.

Stork described Moran as “a researcher who is using every possible means to get research money for his own interests.”140 Moran responded, “I hope it works out that we don’t have a problem. But I’ll be damned if I’ll apologize for sticking my neck out and insisting this program be funded the way it is now being done.”141

Meanwhile, Robert S. Leventhal, the executive vice president of Englehard Minerals

& Chemicals Corporation, the company that was supplying catalysts to both General

Motors and Ford, accused Moran of fabricating a “controversy” in order to win

137 “Do Auto Catalysts Meet the Safety Test?” Business Week, October 27. 1973, 116. 138 Phone interview with Eric Stork, February 25, 2011. 139 Lawrence Mosher, “Catalytic Converter. Next Street Hazard?” The National Observer, undated, from Stork’s clipping files. 140 Ibid. 141 Ibid.

261 funds. As Leventhal claimed, “I have no knowledge of what Mr. Moran’s study is. I don’t think he even has a study.”142

Over time, however, Russell Train, who assumed the position as administrator during this sulfate flap, moved to the side of Mobile Sources. He testified in front of Congress, defending the technology. The agency as a whole had staked much of its credibility in its ability to regulate the automobile, and the catalytic converter appeared to be the answer. If the EPA had to set aside this technology, it would have been at a loss for how to proceed. Thus, Train’s early instincts on the matter may have been more a matter of personal and organizational stakes than one of good science.

The controversy surrounding catalytic converters, more or less, forced Train to suspend the 1977 emission control standards on March 5, 1975.143 In mid-1975,

Moran was scheduled to appear at hearings of the House Subcommittee on

Environment and the Atmosphere. Indeed, the Subcommittee had specifically requested Moran to appear. But the EPA substituted Stork for Moran at the last second.144 Moran appeared as a “backup witness,” but it was Stork who spoke, and

Stork spoke in defense of catalytic converters. Stork, not Moran, represented the

EPA’s policy stance. Train, thus, muzzled Moran.

Luckily for the Office of Mobile Sources and the Office of the Administrator, another powerful group was also deeply invested in catalytic converters—General

142 Ibid. 143 Letter, Eric Stork to Representative James F. Hastings, November 14, 1975, in Eric Stork’s personal collection. Ned MacKay, “The EPA Sniffs Sulfuric Acid,” Palo Alto Times, March 12, 1975. 144 “EPA Would Rather Switch than Fight, Substitutes Stork for Moran,” Environmental Health Letter, July 15, 1975.

262

Motors. The controversy around catalytic converters boiled down to a single issue: automobiles contributed less than one percent to the overall levels of sulfur oxides in the ambient air.145 The real question was what the effect of auto-generated sulfur oxides would be in places with dense traffic. From the beginning, General Motors challenged the ORM’s findings on catalytic converters. The ORM claimed that sulfate emissions stood at an average of 0.05 grams per mile and that the level could spike to 46 micrograms per cubic foot of ambient air.146 The EPA had earlier argued that

10 micrograms per cubic meter should be the threshold for safety.

General Motors definitively settled the sulfuric acid issue through a road test at its proving grounds. On June 11, 1975, the Office of Mobile Sources met with representatives from GM in Ann Arbor to discuss the possibility of carrying out a real world study of sulfate emissions.147 The ORM studies had all been based on dynamometer tests that were then extrapolated using models. In October 1975,

General Motors would settle the issue by recreating the real deal—it drove a large fleet of cars around its Proving Ground in Milford, Michigan, simulating rush hour traffic. As one publication later put it, “In this study, GM operated a fleet of 352 catalyst equipped cars on their 10 km north-south straightaway. The cars ran for 2 hours at a speed of 80 km/h (50 mph), which is equivalent to a traffic density of

145 Eric Stork, “Imposing Automobile Emission Standards.” 146 “Do Auto Catalysts Meet the Safety Test?” Business Week, October 27. 1973, 116. 147 Roger Strelow, Assistant Administrator of Air and Waste Management, to Ernest R. Starkman, Vice President, Environmental Activities Staff, General Motors Corporation, June 197, 1975. Letter, Strelow to Starkman, July 7, 1975. Both in Eric Stork’s personal collection. Strelow replaced Robert Sansom as the Assistant Administrator of Air and Waste Management and as Stork’s direct supervisor after Sansom left the position.

263 approximately 5500 vehicles/hour on a four lane highway.”148 With GM, the EPA, and other groups and automakers collecting data on the ambient air, the study found that sulfates from catalysts were insignificant. The sulfate flap was over.149

Greenfield, Finklea, and Moran all left EPA between 1974 and 1976. The controversy over the catalytic converter may not have been the central reason for their departures. But Russell Train’s decision in favor of the catalytic converter left them in an essentially untenable position at EPA. Stork recalls that, in a meeting,

OMR staff once verged on threatening Train, claiming that, if the administrator continued down the road with catalysts, OMR would not be able to support him.150

After the GM experiment, OMR lost all credibility with both the EPA’s administrator and with the public.

The EPA eventually regulated sulfur oxides, using section 211 of the Clean

Air Act to force their removal from gasoline. But the agency promulgated the sulfur oxides regulations not because of their health effects but because, like lead, they

“poison” catalytic converters. This decision had real health benefits, as sulfur oxides made a significant contribution to the automobile’s overall output of particulate matter.

Within a few years of forming, the EPA’s Office of Mobile Sources—under

Eric Stork’s direction—had become a formidable regulatory agency. Whereas,

148 William E. Wilson et al., “General Motors Sulfate Dispersion Experiment: Summary of EPA Measurements,” Journal of the Air Pollution Control Association, Vol. 27, No. 1 (January 1977), 46–51. 149 Papers from this effort were eventually collected into the volume R. K. Stevens, P. J. Lamothe, W. E. Wilson, J. L. Durham, and T. G. Dzubay, eds., The General Motors/Environmental Protection Agency Sulfate Dispersion Experiment (Research Triangle Park, NC: Environmental Protection Agency, 1976). 150 Email from Eric Stork, February 26, 2011.

264 earlier federal automotive pollution control efforts had focused on R&D, the Office of

Mobile Sources moved decisively towards the production of knowledge through technology assessment. With this information in hand, the agency and its administrators could “compete” with the automakers in the regulatory environment.

In the two subsequent chapters, I show how the agency used these capabilities to gain “competitive advantage” over both the automakers and other federal agencies.

Chapter 5—Establishing the State of the Art: Administering the Clean Air Act of 1970, 1972-1973

In many ways, the work of Eric Stork and his staff at the Environmental

Protection Agency’s (EPA) Office of Mobile Source Air Pollution Control (hereafter,

Office of Mobile Sources), which included refining federal emissions tests and building the capacity of the agency, was all prelude to the coming of the standards set in the Clean Air Act Amendments (CAAA) of 1970. Among other provisions, the law required automakers to meet stringent emissions levels by 1975, but it contained a safety valve in case the automakers could not attain the standards. If the automakers believed they were in such a position, they could petition the EPA to suspend the standards for one year. The automakers then had to demonstrate that they had made “good faith” efforts but that the technology was not yet “feasible.”

Both of these terms were hard to define, and they set up a battle over what was technologically possible in the world of emissions control. Ultimately, the EPA

Administrator had to make the final decision whether to delay implementation of the regulations.

Everyone at the EPA knew that the petitions were coming. The question was not if, but when the automakers would submit them. As the deadline for suspension applications approached, the EPA began planning how it would run the hearings that would determine the Administrator’s decision. It held the first hearings in April

1972, and, on May 12, 1972, William Ruckelshaus, the first EPA Administrator, announced that he would not grant a one-year suspension of the standards. Yet, several of the automakers sued the EPA, claiming that Ruckelshaus had not given

265 266 enough reasons to deny the applications. On February 10, 1973, Judge Harold

Leventhal ruled, in International Harvester v. Ruckelshaus, that the auto companies had a point: Ruckelshaus and his assistants at the EPA had not given the automakers, the courts, or the public enough justification for why the EPA had rejected the applications. The EPA then held a second round of hearings, and, finally,

Ruckelshaus suspended regulations until the following year. Although the story of automotive emissions control was not yet over at this point, I will argue that it was largely “settled” in several key senses.1

This chapter recounts and examines this process: how the EPA held hearings over whether to suspend the automotive emission standards and how it finally made the decision to do so. In many ways, this chapter is a meditation on the role science and technical expertise play in governance, which scholars have studied for several generations.2 I have chosen, however, to focus on a small aspect of this

1 The EPA held a third set of hearings in early 1975, but by this time the debate had changed considerably, as I will discuss briefly at the end of this chapter. The issue of whether the standards should be suspended eventually returned to Congress when the EPA no longer had legislative authority to give extensions to the automakers. Congress later passed the Clean Air Act amendments of 1977, which froze emissions standards at their interim levels for five years. 2 Scholars in the history of science, the history of technology, and Science and Technology Studies (STS) have studied the relationship between science and engineering, on the one hand, and government, on the other, for several decades. Important early works include A. Hunter Dupree, Science in the Federal Government: A History of Policies and Activities to 1940 (Cambridge, Mass.: The Belknap Press of Harvard University Press, 1957); Don K. Price, The Scientific Estate (Cambridge, Mass.: Harvard University Press, 1965); and Daniel J. Kevles, The Physicists: The History of a Scientific Community in Modern America (New York: Alfred A. Knopf, 1978). Some more recent works are Yaron Ezrahi, The Descent of Icarus: Science and the Transformation of Contemporary Democracy (Cambridge, Mass.: Harvard University Press, 1990); Sheila Jasanoff, The Fifth Branch: Science Advisors as Policymakers (Cambrdige, Mass.: Harvard University Press); idem., Designs on Nature: Science and Democracy in Europe and the United States (Princeton: Princeton University Press, 2005); Brian Balogh, Chain Reaction: Expert Debate & Public Participation in American Commercial Nuclear Power, 1945–1975 (Cambridge: Cambridge University Press, 1991).

267 larger topic, namely how regulation produces knowledge.3 I will argue that, contrary to views that see regulation as an impedance to “innovation,” regulation is an engine of knowledge-production of all sorts.4 In some instances, that regulation produces knowledge is a priori, obvious, and even trivial; yet, at each turn, I emphasize the many varieties of knowledge that regulation fostered. First, of course, automotive emission regulations generated new technical knowledge. The CAAA forced automakers to spend hundreds of millions—eventually billions of dollars— on studying emission control systems and the science of air pollution, and many parties—automakers, automotive parts suppliers, academics, environmentalists, and government employees—learned a great deal about emissions science and technology in the process. New regulations also produce new situations, and the actors in those situations must come to grips with their novel environment, including coming to know each other. In the beginning, the automakers and regulators were all non-cooperative players who shared little information, but they eventually built interpersonal knowledge and, in time, even trust. Federal regulators chiefly learned about the automakers through legally-sanctioned surveillance—by making tests of the automakers’ vehicles and requiring automakers to produce and

3 In any act of regulation, there are multiple kinds of uncertainty. Technology-forcing regulation contains an intractable uncertainty, namely what will happen in the future. But in creating and enforcing standards that are technology-forcing, regulators face another kind of uncertainty— knowing fully the present state of a technology. To predict where a technology can be in the future, one has to have some sense of where a technology is currently. The consent decree, however, had effectively blocked the formation of this knowledge. No one had a good idea of where emissions control technologies were at across the board. 4 Margaret Taylor (and her co-authors) have written about how regulation fosters knowledge- production and information-sharing in Margaret R. Taylor, “The Influence of Government Actions on Innovative Activities in the Development of Environmental Technologies to Control Sulfur Dioxide Emissions from Stationary Sources” (PhD diss., Carnegie Mellon University, 2001); Margaret R. Taylor, Edward S. Rubin, and David A. Hounshell, “Regulation as the Mother of Innovation: The Case of SO2 Control,” Law & Policy, Vol. 27, No. 2 (April 2005), 348-78.

268 provide information—in a dynamic akin to what Michel Foucault called

“discipline.”5 Additionally, both regulators and the regulated learned the full meaning of the law through court decisions on the topic. In the courts, the EPA learned the limits of its power, while the automakers reduced their uncertainty about the regulatory environment.

Moreover, the regulations created another, perhaps unintentional, and certainly ironic, result. As regulators struggled to understand the state of automotive technology and where it could be in the future, they discovered and established the state of the art in automotive emissions control. Anti-trust rules had prevented automakers from discussing their technologies. The auto companies had signed a consent decree limiting such talk in 1969. For this reason, no one truly knew where emissions control technology sat. The hearings created an environment in which automakers reported openly on what they knew and how they knew it.

Thus, regulation produced a great deal of information useful to government, antipollution advocates, and industry. When processed through dialogue, inquiry, comparison, this information yielded scientific, technical, and industrial knowledge.

This chapter also examines two secondary but complementary issues: First, I will suggest that debates during the two suspension hearings and International

Harvester vs. Ruckelshaus were primarily about what form of rationality should comprise government regulation; that is, what the proper form of governance was.

The EPA attempted to create a system for handling objections to its decisions that would both tame conflict and channel remaining adversarial energies into a

5 Michel Foucault, Discipline and Punish: The Birth of the Prison, trans. Alan Sheridan (New York: Vintage Books, 1977).

269 controllable and predictable procedure. Firms contested the form of these procedures. Moreover, the EPA and automakers disagreed about both how to produce truth and what kinds of truths needed to be produced in order to carry out the regulations. On the one hand, the automakers often focused on a juridical notion, asking whether the regulations were just. On the other, the EPA wished to restrict the regulatory process to a veridical, or truth-making, project, and to asking strictly technical questions about whether a given emissions control technology was

“feasible.”6 Some of these conflicts over definitions arose from ambiguities in the legislation. In the absence of sufficient technical progress, the automakers were asked to show that they had made “good faith efforts” to developing emissions controls. How one was to judge “good faith” was a hairy question filled with social and political implications. Ultimately, however, notions, like “feasibility” and “good faith,” showed how ideas of truth-making and ideas of justice were inextricably intertwined. In this case, the political was personal, and the personal was technological.

Second, the CAAA created a politics of possibility. Automakers argued that the standards could not be met; catalyst makers and environmentalists said they could. The agency and automakers disagreed fundamentally about the nature of technological change and what model of innovation should inform regulatory acts.

As one EPA employee later recalled, the agency believed that something like the

6 The distinction between juridical and veridical notions of activity comes from Michel Foucault’s The Birth of Biopolitics: Lectures at the College de France, 1978-1979, trans. Graham Burchell (New York: Palgrave Macmillan, 2008), especially lecture 2. Foucault applies this distinction to two different conceptions of the way “markets” work—as sites of justice or as sites to truth-making—but the distinction also holds well for the activities of many federal bureaucracies.

270 invention of the light bulb should serve as the ideal model of introducing new technologies. The producer would choose the most successful example out of a generation of attempts and further develop it, as Edison had done with the light bulb. Propositional knowledge about why the control technology worked was not essential to moving forward. The automakers focused on what they could produce on a mass scale in short order and asserted that regulations should be based on the average technology available, not the best iteration in a generation. The debate around this issue often focused on whether most or all of the manufacturers’ vehicles should meet the emissions standards or whether test results for a given vehicle or engine class should be averaged. In this way, this debate also touched on the issue of how alike any two mass-produced objects are.

This chapter proceeds through four chronological sections. It begins with the members of the EPA deciding how they would organize hearings arising from the automakers’ petitions for suspension. Next, it turns to the first set of hearings and

Ruckelshaus’s decision coming out of them. Yet, several automakers challenged this decision in court, so the third section examines the DC Circuit Court’s decision to remand the suspension to the EPA and Ruckelshaus for further consideration.

Finally, the chapter turns to the second round of hearings and the EPA’s creation of a statistical methodology that would settle the issue of what “feasibility” meant.

Before considering these issues, however, we must examine the technological systems that were under debate during the period, particularly the catalytic converter and, though less important, the thermal reactor. In many ways, the limits of these systems constituted the limits of regulation.

271

Technical Considerations

The Clean Air Act Amendments of 1970 set a high bar for automotive emissions control. It required automakers to reduce carbon monoxide (CO) and unburned hydrocarbons (HC) by ninety percent of 1970 levels by 1975. Likewise, nitrogen oxide (NOx) emissions had to be cut the same amount by 1976. This translated into 0.41 grams per vehicle mile HC, 3.4 grams CO, and 3.1 grams NOx in

1975. In 1976, NOx would be reduced to 0.4 grams. To understand how intense this change was, we should consider that, according to one estimate, uncontrolled 1968 vehicle emissions were around 17 grams per mile HC, 125 grams per mile CO, and 6 grams per mile NOx.7

The automakers had met the standards set by the Motor Vehicle Pollution

Control Act of 1965 through a number of engine modifications, including

“adjustments to the carburetor to provide leaner fuel-air mixtures and improved mixture uniformity, controlled heating of the intake air, increased idle speed, retarded spark timing, reduced engine compression ratio, and improved cylinder head design.”8 Different manufacturers used a variety of other systems as well, including using air pumps to burn up more HC and CO in the exhaust manifold.9 HC emissions were also heavily curtailed by eliminating emissions lost through

7 Semiannual Report by the Committee on Motor Vehicle Emissions of the National Academy of Sciences to the Environmental Protection Agency, January 1, 1972, 12. 8 Ibid. 9 As described in the previous chapter, the manufacturers use of air pumps was a major factor that pushed the EPA to change their testing procedures, since the additional air effectively “watered down” the controlled emissions’ concentration.

272 evaporation.10 Yet, major technological breakthroughs were not forthcoming.

Indeed, in 1969, the Justice Department brought suit against the four largest domestic automakers accusing them of conspiring to hold up the development of emissions control technologies.11 The manufacturers, it seemed, were were unwilling to deploy any emissions control technologies that they did not all have.

The suit was eventually settled by a consent decree in which the companies agreed to stop sharing research about emissions control technologies.

New technologies would be necessary to meet the standards set by the Clean

Air Act Amendments of 1970. Engine modifications alone would not be sufficient.

In passing the amendments, Congress knew that the automakers did not have the technologies available; the companies’ research branches had to develop them. In this sense, the law was “technology-forcing”: it would push the auto firms to develop new technologies.

During the early 1970s, automakers experimented with catalytic converters that had two different physical configurations, beads and monoliths (which resemble honeycombs), and two basic kinds of materials, either noble or base metals. In each case, the catalyst was applied to the material, either beads or a monolith, and this material was, in turn, placed in a “can” that would be installed in an automobile’s tailpipe. Automakers preferred base metals simply because of cost.

Moreover, the most commonly used noble metal in catalytic converters was platinum, one of the most expensive precious metals of all. The bulk of platinum— nearly 1.5 million troy ounces yearly—came out of Russia and South Africa, two

10 Blow-by HC emissions had been controlled since 1963. 11 International Harvester v. Ruckelshaus, 623

273 places considered politically unstable by the day’s standards. One GM employee estimated that, if the entire US auto fleet moved to platinum catalysts, it would require between 1.3 and 1.4 million troy ounces a year.12 This would nearly double global production, requiring major changes in the platinum industry, greatly increasing the number of miners needed, and, probably, vastly increasing the price of platinum. Others, however, including employees at the catalyst-producing companies Matthey-Bishop and Englehard, believed that producing enough platinum would not be a problem.13 As we will see, this dynamic of automakers making pessimistic predictions and suppliers making optimistic ones ran throughout the era.

During this period, catalytic converters also had two basic flaws, or “reverse salients,” constraining their successful development—physical durability and chemical “poisoning.”14 The catalytic materials could break down physically for a number reasons, but the most common were from intense heat and from simple wear in being shaken through vibrations during operation. The converters had to work under extremely hot conditions, typically at least 1,200 degrees Fahrenheit.

This was especially true in “overtemperature” conditions, such as high elevations or engine misfires, when temperatures could reach nearly 2,100 degrees. Moreover, a malfunction, like a misfire, could cause the temperature of catalytic materials to spike within fractions of a second. Automakers envisioned making “blow-by

12 Hustead, “Auto Emissions Extension,” 1469. 13 Sansom, “Auto Emissions Extension,” 1471. 14 On how engineers frame problems around “reverse salients,” see Thomas Hughes, Networks of Power: Electrification in Western Society, 1880–1930 (Baltimore: Johns Hopkins University Press, 1983), 14

274 systems” that would vent the gases if they became too hot, but no one knew how to build a system that would react that quickly. At GM, overheating sometimes led to beaded catalysts becoming so hot that they would burn through the “grid plate” that held them, though they were retained by the external casing.15 In at least one case, after testing a converter at high speeds, the floor mats began to “smolder.”16

Additionally, the threat of fire on asphalt or, especially, leaves and grass was a constant concern.17 In GM tests, converters did occasionally “char” grass that they were parked over. For these reasons, converters had to be insulated both above and below, which increased unit costs.

The beaded catalysts were especially susceptible to breaking down from vibrations. Often, beaded catalysts would slowly be turned into dust and would blow away through the exhaust system. The automakers were concerned about the possibility of extremely hot catalysts breaking down physically, “blowing out” the exhaust system, and injuring someone standing behind the vehicle. As one GM employee put it, “[We] blew the beads out to see, suppose you were walking behind a car, a tailpipe, at the exact instant they let go, would the beads hurt. They are warm. They sting. But there was no breakage of the skin, no severe burn or anything.”18

Poisoning of the catalysts was a more complex issue. Under different conditions, chemicals would bond to the catalytic materials, reducing the effective,

15 Stempel, “Auto Emissions Extension,” 1445. 16 Stempel, “Auto Emissions Extension,” 1446. 17 Eric Stork eventually took part in litigation over a case of fire supposedly caused by a catalytic converter. He testified for the auto companies and maintains that catalytic converters were not really a fire hazard. 18 Stempel, “Auto Emissions Extension,” 1445.

275 catalyzing surface area and stopping the catalysis. The chemical industry had used catalysts for a long time but under very controlled conditions. Conditions in cars varied wildly, chaotically, depending on such mundane factors as weather and the user’s driving style. Lead was the primary culprit. But sulfur and phosphorous also damaged the catalysts.19 The discovery that sulfur would poison catalysts was rather new at that time. It took catalyst suppliers by surprise because they were used to industrial processes that used extremely high heat and that had a “reducing atmosphere.”20 Beyond creating a “rotten egg” smell in the exhaust, sulfur bonded to catalysts at lower temperatures. If the converter was brought up to high temperatures, the sulfur would oxidize off, but this increased the probability of other heat-related problems emerging. The borderline for sulfur buildup lay between 1200 and 1300 degrees.21 Any higher and the sulfur would begin to reduce. But if someone only used his or her car to run errands, it might never reach these temperatures for sustained periods of time, and its catalyst would be poisoned. To deal with the lead problem, the automakers pushed for creating unleaded gasoline in order preserve the catalysts. General Motors had once owned a hefty share of Ethyl Corp, the primary manufacturer of the tetraethyl lead gasoline additive, so for General Motors to be advocating for the removal of lead from gasoline was a significant statement. Unsurprisingly, petroleum companies pushed back. GM president Ed Cole famously announced at a California Air Resources

19 Hustead, “Auto Emissions Extension,” 1403-4. 20 Starkman, “Auto Emissions Extension,” 1458. As I will discuss at the end of the chapter, sulfur became a major point of controversy when some asserted that catalytic converters would produce sulfuric acid, which was hypothetically more hazardous than smog. In other words, catalysts would be creating a greater health hazard than the one they were mitigating. 21 Hustead, ibid., 1459.

276

Board meeting that, if the government would ban lead in fuel, he would deliver a functional catalytic converter.22

Unleaded gasoline, however, provided a rich example of managing complexity in systems: change one thing in order to solve a problem and you are likely to create new problems. In the 1970s, for reasons related to both emissions control and fuel efficiency, manufacturers began adopting electronic fuel injection.

Auto engineers had used mechanical fuel injection in diesel engineers for many years, but the oily nature of diesel fuel gave it lubricating properties that unleaded gasoline lacked. Engineers found that unleaded fuel more quickly deteriorated fuel injection systems. For instance, Volvo learned that unleaded gasoline created deposits on the injection needle, the part of the fuel injection system that projected into the cylinder.23 In one of its test vehicles, the fuel injection system failed from using unleaded gasoline before the catalytic converter did. Unleaded gasoline was necessary for catalytic converters, but it was by no means a trouble-free proposition.

Several automakers also considered, and, indeed, some ended up deploying, thermal reactors to control emissions. Thermal reactors were enlarged exhaust manifolds. They had a thin steel core inside a cast-iron shell. The thin steel heated very quickly once the car was started, while the shell prevented heat loss. Exhaust would come from the cylinders and mix with secondary air drawn in from an air

22 On Cole, Tom McCarthy, Auto Mania: Cars, Consumers, and the Environment (New Haven: Yale University Press, 2007),179. 23 Elljbaer, “Auto Emissions Extension,” 91. Clogged fuel injectors continued to be a problem until 1994, when the Federal government required petroleum companies to put detergent in gasoline to stave off the problem.

277 pump, which added oxygen to the air mix. The reactor’s enlarged size allowed air to remain longer than it would in a normal exhaust manifold. The hot core then would burn off some of the HC and CO in the exhaust. A catalytic converter would then be used to clean up any remaining HC and CO (and eventually NOx with the three-way converter).24 Thermal reactors had several problems, however. First, they significantly increased fuel use, which was not a concern in the emissions control hearings until 1973 when the OPEC oil embargo made it clear that any effective emissions control system could not also greatly decrease fuel efficiency. Second, thermal reactors had to operate at severely high temperatures. Normal operating temperatures were between 1600-1800 degrees Fahrenheit, but reactors had to be able to tolerate temperatures as high as 2200-2300 degrees Fahrenheit, when a car overheated or if a cylinder was misfiring, thus, putting fuel directly into the exhaust manifold, thereby, increasing the reaction temperature. This high heat often caused the core to sag, even crack. Moreover, the thermal reactor had to be mounted directly on the cylinder head. This created ideal conditions for a fire in the engine, as this extremely hot piece of steel came in close contact with fuel lines and other flammable materials. In some of its tests, Volvo had attached a catalytic converter directly to the exhaust manifold, in much the same location as a thermal reactor.

When a fuel injector sprang a leak, the leaking fuel and then the entire engine caught fire. Typically, the limits of the thermal reactor were less dramatic. They simply failed. By the time of the first hearings, none of the industry’s thermal reactors had made it through the 50,000 mile durability test.

24 Automakers adopted the three-way catalytic converter—so named because it controlled CO, NOx, and HC—after more stringent NOx standards came into effect in the late-1970s.

278

Organizing the Hearings: Rationalities of Regulation

As noted earlier, the CAAA contained a “safety valve” for the auto manufactures, which was largely a response to a central problem: legislators wanted to force the manufacturers to reduce air pollution, but no one could predict the future. No one could say where emissions control technology would be in five years, even if the automakers and their suppliers gave their all to its development during this span of time. If automakers made a “good faith effort” to develop emission controls but had not found a “feasible” control system, they could petition the EPA for a one-year suspension of the standards. If the EPA decided to grant a suspension, the Administrator was to set interim standards that reflected the greatest degree of control possible at that time. Yet, as we will see, what exactly was required to prove that the standards could not be met became a controversy between the manufacturers, the EPA, and the courts. The initial decision fell to the first EPA Administrator, William Ruckelshaus.

The applications came as expected. Volvo sent the first one, which arrived on

March 13, 1972. Chrysler, Ford, General Motors, and International Harvester sent in their respective applications soon after that.25 If manufactures submitted a petition, the CAAA required the EPA to hold hearings in which the agency was to discern the verity of the manufactures’ claims. The Act presented the EPA with an uneven field of dictates. Some requirements, like the mandated emissions levels, were spelled out in great detail in the law; others, the administrative regulatory procedures,

25 Letter from William Ruckelshaus describing the Second Decision, “Decision of the Administrator of the Environmental Protection Agency Regarding Suspension of the 1975 Auto Emission Standards,” Subcommittee on Air and Water Pollution, Committee on Public Works, 93rd Congress, First Session, April 16, 17, and 18, 1973, 3. From here on, noted as “Second Decision.”

279 barely contained any direction at all. The suspension hearings belonged to the latter category. Laws, such as the Administrative Procedures Act, set protocols for rule- making and other government actions but were silent on the issue of these hearings.

In such times, bureaucrats often turn to examples from other agencies, but

Ruckelshaus and his advisors believed that they had no models to run on. As George

Allen, a lawyer who functioned as Ruckelshaus’ chief legal advisor on the suspension hearings, later claimed, “The decision involved was not really like any regular, recurring regulatory function of an existing agency.”26 Without such legal direction or a model to work from, employees of the EPA became responsible for structuring these sui generis hearings.

In creating the procedures for the suspension hearings and decision, the EPA had two primary goals. First, it had to gather the information necessary for making the decision. Second, the EPA believed it critically important that its public actions be seen as rational; thus, the agency had to devise procedures to make them seem so. Historians and sociologists have long noted that states formed bureaucracies and adopted scientific and statistical methods to downplay the political, contested nature of their decisions.27 The CAAA had taken something that once fell under

26 George Allen, National Research Council, “Transcript from Meeting on Resolution of Scientific Issues and the Judicial Process: International Harvester v. Ruckelshaus,” May 14, 1976, 7. From here on, “NRC Transcript.” Since this document is the transcript of a discussion, I will note who made these claims in the first entry of each footnote. Ruckelshaus knew George Allen from his undergraduate days at Princeton and hired him onto the agency. All attempts to learn about George Allen’s biography and whereabouts through former associates, including Ruckelshaus and Stork, and biographical reference tools have failed. 27 See, for instance, Stephen Skowronek, Building a New American State: The Expansion of National Administrative Capacities, 1877–1920 (Cambridge: Cambridge University Press, 1982); Theodore M. Porter, Trust in Numbers: The Pursuit of Objectivity in Science and Public Life (Princeton: Princeton University Press, 1996); and Jonathan Kahn, Budgeting Democracy: State Building and Citizenship in America, 1890–1928 (Ithaca: Cornell University Press, 1997).

280 corporate control and placed it in the hands of the state. Between the concerns of the automakers and the concerns of environmentalists and consumer advocates, the

EPA’s activities would be closely watched. Controversy could not be avoided. Thus, the EPA had to build procedures that would make the decision seem rational and, therefore, unassailable when it was handed down.

Eric Stork and his staff members at the EPA’s Office of Mobile Sources discussed publishing requirements for the petitions, but they did not want to seem to be encouraging suspensions requests. So, the EPA staff members waited for the first petitions to arrive and promptly wrote back that the petitions provided inadequate information and gave the firms a template for the revised petitions.28

But the EPA was still unclear about what information it should ask for. What needed to be known? As Stork subsequently recalled, “ In addition to not having the answers, we didn’t even understand all of the questions yet.”29 Figuring out what the questions were but one form of the EPA’s organizational learning. In many ways, members of the agency were naïve.

At that moment, however, the EPA staff members could not wait for learning.

They needed information now. Stork and Allen discussed the best way to request information and what information to request.30 Allen drew attention to section 307

(a) of the CAAA amendments, which required firms to furnish any data requested.

When Stork pointed out that they still did not know what to ask for, Allen said,

28 Eric Stork, NRC Transcript, 9. 29 Ibid., 5. 30 Ibid.

281

“Well, just require them to give you all their data.”31 This tack was a blunt tool and created great inefficiencies, but it was their only way to ensure that any question asked during the hearings was fair game and that the agency knew as much as possible before hand. Stork had blackly humorous fantasies of “GM dump trucks moving up to our lab, the tail gate lifting and tons of paper dropping down on Karl

Hellman and his three associates.”32 The automakers turned over thousands of pages of engineering data and scientific studies. A Chrysler employee characterized the process as the “complete purgative . . . of everything that was going on in the corporation.”33 Hellman, Austin, and other Office of Mobile Sources’ staff members spent many late, frantic nights poring over paper. The industry submissions were a mixture of raw data and rhetoric, pleas that the standards could not be met. But, as later Stork pointed out, the industry information served two functions: it made the companies’ cases for suspension, and it fulfilled the EPA requirements for information.34

The information request for the suspension hearings built on the back of the

Office of Mobile Sources’ technology assessment program, which was described in

Chapter Four. In August or September of 1971, the Office of Mobile Sources sent out a technical questionnaire to the automakers, and by October or November, the agency started receiving replies.35 All of this information was gathered under promise of confidentiality. But in early 1972, Ruckelshaus began questioning the

31 Ibid. 32 Ibid. 33 Charlie Heinen, NRC Transcript, 4. 34 Eric Stork, NRC Transcript, 4-5. 35 Ibid., 6.

282 wisdom of this promise. He believed that, if the hearing process was to be credible, it would have to be open to public scrutiny. This meant that the industry data would have to be public as well.36 In early 1972, Ruckelshaus held a meeting, which both

Allen and Stork attended, about the matter. The point was hotly contested, but

Ruckelshaus decided that transparency was the only way to proceed. Stork strongly disagreed.37 He had signed the letters that accompanied the early information requests, and he believed reversing course on confidentiality would undermine his own trustworthiness. He lost in the end, however.

Though the EPA had secured a great deal of information through the suspension applications, it still had to decide how to structure the hearings. But what form should they take? Ruckelshaus and Allen were both lawyers. Stork was not, but his father had been one. All were suspicious of modeling the suspension hearings on court proceedings. They believed that industry lawyers, if given the chance, would bog down the hearings through administrative requests and cross- examination. Since the EPA was under a strict timeline, Ruckelshaus and company felt that they could not lose time to following such procedures.38 Moreover,

Ruckelshaus did not want to put his scientists, engineers, and technicians on the stand. If industry lawyers cross-examined EPA employees, they might undermine the EPA’s already delicate credibility, and Ruckelshaus always paid attention to maintaining the agency’s trustworthiness. As Allen put it, “We didn’t want the

36 There were some exceptions to this rule. Production numbers and other information that firms considered extremely sensitive and that would conceivably aid their competitors were redacted. 37 Eric Stork, NRC Transcript, 6. 38 George Allen, NRC Transcript, 8.

283 hearing to turn into a process in which counsel would interrogate witnesses in the usual way that counsel do, trying to elicit peripheral matters that seem highly prejudicial or dramatic and yet never groping with the entire mass of data that we thought had to be analyzed.”

This assertion went to the heart of the disagreement between the EPA and industry, namely the question of what truth the hearings would produce. As evidenced by the early industry testimonies at the hearings, the firms sought to put the Clean Air Act and the EPA’s procedures under question. For industry, the question was whether the law made sense and whether the government was acting rationally and justly. Ruckelshaus did not want to place the law or his employees in doubt. For the EPA, the question was whether the industry had proven that it could not meet the emissions standards.

The disagreement partly rested on the issue of burden of proof. The agency did not adopt the traditional concept of burden of proof that stands in “adjudicatory procedures.” Ruckelshaus said, “ We are not taking the position that an applicant would suffer unfavorably simply because the affirmative evidence he submits may be inadequate to support his contentions in his application.”39 In other words, the

EPA would not penalize an automaker simply because it did a bad job of making its case. Instead, Ruckelshaus would “consider all information available to him from any source.” The issue at hand in the decision was the fact of current technological development, not how automakers had presented this development. Yet,

Ruckelshaus argued, “In this practical sense, therefore, it is incumbent on the

39 Ruckelshaus, Emissions Extension, 22.

284 applicants to make certain that the state of the record at the end of the proceeding permits the administrator to make the determinations which the applicant believes should be made. In that sense, and only in that sense, do we assert that the applicant has a burden of proof.”40

The companies were not allowed to cross-examine each other during the hearings. Near the beginning of the hearings, Ford Motor Company had requested that the auto companies be allowed to cross-examine each other because the companies shared economic straits, in effect “the applicants’ livelihood and business interests are in jeopardy.”41 These shared interests were supposed somehow to make it appealing for them to be able to ask each other questions, perhaps as a means for bypassing the consent decree. Ruckelshaus didn’t agree, however. Cross- examination would jam up the proceedings that were already under extreme time constraints. As important, Ruckelshaus did not believe that the automakers’ shared economic interests entitled them to any special treatment in the hearings. As he said, “The Agency does not agree that the interest of an applicant in these proceedings rises above the interests of the general public.”42 Furthermore,

“Even if we consider only the kinds of economic and property

interests which the [Ford petition] dwells on, we have witnesses here

who represent equally strong economic interests; the interest of

suppliers in the outcome of this proceeding is as strong as the

interests of the manufacturers of the basic vehicle. The interests of

40 Ruckelshaus, Emissions Extension, 22-3. 41 Ruckelshaus, Emissions Extension, 20. 42 Ruckelshaus, Emissions Extension, 21.

285

business men who do business in urban areas that are threatened by

unhealthful air pollution caused by automobile emissions is an

economic interest which is in many cases as strong as the economic

interests of the manufacturers in continuing to produce vehicles

which do not meet a stringent emissions standard.”

The hearings went like this then: the witnesses would show up in the morning, and EPA employees would begin grilling them about their emissions control programs. In at least one way, the EPA took a cavalier attitude towards forming the hearings. The agency did not think much about judicial review when creating the hearings, though it did when writing the decision.43 This attitude and the agency’s whole approach probably helped foster the set of circumstances that led the automakers to sue the agency and then persuaded Judge Harold Leventhal to rule against the EPA in International Harvester v. Ruckelshaus, as I will discuss later.

The First Hearings: Questioning the Companies, Trading Analogies

The hearings began on Monday, April 10, 1972 in an auditorium at the

Department of Commerce in Washington, D.C. Ruckelshaus started the hearings off with strong words, saying, “I fully share Congress’ conviction that the automobile cannot be permitted to continue to menace the health and welfare of large numbers of the people of this nation.”44 Sticking to his desire to limit legal jousting and to focus primarily on technical issues in the hearings, he announced, “For the purpose of making the decision, the wisdom of that standard is not debatable, nor is the need

43 George Allen, NRC Transcript, 4. 44 William Ruckelshaus, Extension Hearings, vol. 1, 4.

286 for prompt development of more effective automobile control systems debatable.”45

He believed that the decision about the suspension broke down into four factors: first, it had to be the case “that suspension is essential to the public interest or the public health and welfare of the United States.”46 In his definition of public interest,

Ruckelshaus included both that the automakers would be able to produce enough cars to meet demand and that the standards would not create massive layoffs or economic damage.47 Second, the automobile manufacturers had to show that they had made “good faith efforts,” though there was no agreement on how such a thing would be shown. Third, the companies had to demonstrate that there were no feasible emissions control technologies available, which required them more or less to make a negative proof. They could also show, if an effective technology was available, that they did not have sufficient lead time to meet the standards. Finally,

Ruckelshaus would take into account information from other sources, including the

National Academy of Sciences.

One cannot help being struck by the number of analogies traded between participants at the hearings, especially on the part of automakers. At times, the regulation table seems like a trading-zone for analogies, a place to share stories.48

These stories were effectively a way for regulators and the regulated to teach each

45 Ruckelshaus, Extension Hearings, vol. 1, 4. 46 Ruckelshaus, Extension Hearings, vol. 1, 6. 47 Interview with William Ruckelshaus. 48 The hearings (again, unintentionally) became what Peter Galison calls a “trading zone,” a space where varying parties can exchange information, though they may disagree about world-views and philosophical assumptions. Galison focuses on different groups within theoretical physics, and so the players in his trading zone amount to “subcultures of instrumentation, experiment, and theory.” (Galison, 803) The actors in the suspension hearings could be viewed as different organizational subcultures of government (regulators), industry (automakers and their suppliers), and activists (environmentalists and consumer advocates). Galison, Image and Logic: A Material Culture of Microphysics (Chicago: The University of Chicago Press, 1997).

287 other. In his work, Doing without Concepts, the philosopher of science Edouard

Machery claims that analogies, “exemplars” in his terminology, are one form of reasoning along with two others, prototypes and theories.49 He argues that, for too long, philosophers and psychologists have treated concepts as natural kinds and have needlessly and groundlessly folded distinct cognitive processes under this umbrella term. We treat them as if they were one thing, but they are different things.

In this context, we should ask why analogies—as opposed to other cognitive processes—play such an important role in science and technology policy (STP).50

For instance, the title of Donald Stoke’s Pasteur’s Quadrant, in many ways the best book on the interplay of science, technology, and the policies that foster them, is itself an analogy: Stokes argues that science and technology policy would excel if it focused on “basic research that seeks to extend the frontiers of understanding but is also inspired by considerations of use.” This type of research, which Stokes calls

“Pasteur’s Quadrant,” is so named because it resembles Louis Pasteur’s methods. To work in such a way is to work as Pasteur worked. Analogies bloom like dandelions

49 Edouard Machery, Doing without Concepts (Oxford: Oxford University Press, 2009). The difference between an exemplar and a prototype can be confusing. The basic difference is this: prototypes are ideal models drawn from a host of experiences. They are statistical averages. We have a prototype “dog” from all of our past encounters with dogs. Exemplars, however, refer to specific, representative cases, or, in Machery’s heavy prose, they are “sets of representations of specific category members.” Machery, 173. When reasoning with exemplars, the object of reason is compared to specific analogues, and if it is found to be like them, it becomes part of their set or class or category. The reasoner argues, “It is like this.” Now, it seems intuitive, and indeed some philosophers and psychologists have argued, that prototypes can subsume exemplars, but experimental studies support the view that people use both prototypes and exemplars under different conditions. Machery, 179-183. I would suggest that people rely on exemplars especially when they cannot form a representative (that is, statistically average) model of a given class of objects. We could call such classes “non-prototypical (or non-prototypeable) classes.” I would suggest further that “scientific discovery” and “technological invention” are two such non- prototypical classes, and, thus, decision-makers draw heavily on exemplars when reasoning about science and technology policy. 50 Donald E. Stokes, Pasteur’s Quadrant: Basic Science and Technological Innovation (Washington, D.C.: Brookings Institute, 1997), 74.

288 throughout discussions of STP. Often when people discuss global climate change and the technological change necessary for mitigating it, they liken our efforts to a new Manhattan Project.51 In pointing out the dominance of analogies in STP, I am not decrying their existence. If Richard Neustadt’s and Ernest May’s Thinking in

Time, which examines the role of analogies in decision-making, taught us anything, it is that there are good analogies and there are bad analogies and there are ways of telling between them, though sometimes only retroactively.52 Instead, I wish to point to analogy’s role in STP as an unexplained puzzle. For all the work scientists, philosophers, historians, and social scientists put into capturing the essence of scientific discovery and technological creativity, the essence continues to escape us.

Left without repeatable data for induction, we grasp at likeness.

In the EPA’s suspension, analogies played a crucial role in learning and teaching. On the one hand, the auto industry used them to teach regulators about industry structure and the traditional division of intellectual labor between the automakers and their suppliers.53 On the other hand, the regulators put forward analogies to question the automakers’ practices and to encourage increased research and development.

51 This was also true of energy independence in the 1970s and afterwards. President Barack Obama recently made such an analogy in his 2011 State of the Union Address. 52 Richard E. Neustadt and Ernest R. May, Thinking in Time: The Uses of History for Decision-Makers (New York: The Free Press, 1986). 53 On the “division of intellectual (or ‘innovative’) labor,” see Ashish Arora, Andrea Fosfuri, and Alfonso Gambardella, Markets for Technology: The Economics of Innovation and Corporate Strategy (Cambridge, Mass.: The MIT Press, 2001). Also, Jaegul Lee and Francisco Veloso have examined how central firms and suppliers partition knowledge to foster innovation. Lee and Veloso, “Inter-Firm Innovation under Uncertainty: Empirical Evidence for Strategic Knowledge-Partitioning,” Journal of Product Innovation Management, Vol. 25., Issue 5 (2006), 418–435.

289

Though only five automobile companies applied for suspension, the EPA subpoenaed the other major vehicle manufacturers, both foreign and domestic, who sold cars in the United States. Though these companies could not meet the standards either, they had not yet filed petitions; the EPA subpoenaed them so that everyone would be at the table at one time. The agency also subpoenaed suppliers of catalysts and other devices and materials that would potentially play a major role in emissions control. Ruckelshaus announced in his opening statements that the EPA had invited these suppliers because they would likely contradict the auto companies that had applied for suspension.54 The agency had invited public interest groups for the same reason. If the manufacturers were going to get a suspension, they were going to have to fight for it.

The people who testified at the first hearings can be roughly broken down into four groups: the Big Three—General Motors, Chrysler, and Ford—and other domestic manufacturers, foreign manufacturers, catalyst companies and other suppliers, and public interest and governmental groups. Each of these groups generally towed a party line that corresponded to their respective “interests” in emissions control. The domestic auto companies all sounded the same note: the standards could not be met. The Big Three were the major voices here. American

Motors Company, for instance, depended almost wholly on General Motors for research because it did not have the scale to carry out major research projects and because it had been cash strapped and barely profitable for years.55 Catalyst makers

54 Ruckelshaus, Emissions Extensions, 8. 55 For a popular recounting of AMC’s financial woes, see Keith Bradsher, High and Mighty: SUVs—The World’s Most Dangerous Vehicles and How They Got that Way (New York: Public Affairs, 2002).

290 all claimed that the standards could be attained, and, of course, if the automakers adopted catalytic converters, they would require hundreds of tons of catalyst daily, opening a huge new market. The catalyst makers’ potential profits could not have been tidier. Environmentalists and consumer advocates invariably believed that the standards fell within automakers’ grasp and argued that the companies were simply dragging their feet by applying for suspension.56 Foreign manufacturers were, by far, the most divided group. Interestingly, this divide fell along East-West lines.

European automakers were pessimistic about meeting the regulations’ requirements. (They had also put far fewer resources into research and development on emissions controls.) Asian companies put a much brighter spin on the prospects of emissions reductions. To some degree, they had a jump on domestic automakers on many of the factors that influenced emission controls.

Their cars were smaller, and they had put more into alternative power sources, including the Wankel engine and Honda’s CVCC engine. But, at the same time, they were probably overly optimistic in their estimates of what they could attain in the short run. Part of their positive spin at the hearings may have arisen out of a need to assuage bureaucracies in their home countries, especially the Ministry of

International Trade and Industry in Japan.

56 The consumer advocates’ and environmentalists’ unwillingness to violate liberal norms is striking. Though one community organizer did suggest that the ideal was that people would stop driving, most other progressive witnesses steered well clear of such talk. In his testimony, David Hawkins, who was, at that time, a lawyer at the National Resources Defense Council and who later, under President Carter, became the EPA’s Assistant Administrator of Air, Noise, and Radiation, felt compelled to say, “A halt to manufacturing operations in Detroit in 1975 is neither a real nor desirable possibility. Accordingly, such a result should not be hypothesized in calculating whether a suspension is essential to the public interest..” Hawkins, 2572. If nothing else, Hawkins’ statement demonstrates how much debates over emissions control in the United States took place within the framework of a liberal consensus.

291

Automakers were centrally concerned about three issues, uncertainties that they believed the EPA could alleviate. First, no one had a clear picture about what the EPA’s eventual policy on vehicle maintenance would be. None of the automakers were making it anywhere close to the 50,000 mile durability requirements of the Clean Air Act Amendments. Catalytic converters fouled by that point, and thermal reactors often cracked or melted. Later, the EPA allowed for the automakers to swap out the catalytic converters during the 50,000 miles but only after 25,000 miles. This decision was controversial both with environmentalists and with the courts, since it was unclear what incentives consumers would have to get their emissions controls serviced.57 Second, along similar lines, the manufacturers were unsure where the final lead levels in “unleaded” gasoline would be set. Everyone knew that petroleum companies would have to reduce lead levels for catalytic converters, but no consensus existed about what “unleaded” gasoline was. Many of the manufacturers used very different lead levels in their durability tests; each had its own notion of de-leaded gas. Finally, the EPA had not officially announced how or whether it would average the automakers’ emissions test results to judge whether the automakers had met the standards. Couched in this larger issue was the question of whether the EPA was going to carry out tests on the manufacturers’ assembly lines or whether it would rely on some other system to

57 Ruckelshaus proposed this changed rule on October 31, 1972. See 37 Fed. Reg. 23,778-23,779. The proposed rule contained a caveat that the consumer’s failure to get maintenance done could not improve the performance of the vehicle. Many of the automakers were building by-pass systems into their emission control plans, for cases of overheating, etc. If a system failed permanently and was by- passed, this could hypothetically improve vehicle performance as raw exhaust was vented into the air. The rule probably sought to ensure that consumer’s did not have the incentive not to get their emissions controls repaired. As I described in Chapter Four, the EPA had already dealt with automakers’ intentionally using “defeat devices” to disable emissions controls and improve driveability. The EPA did not want a replay of that controversy.

292 ensure compliance. With these three questions hanging in the balance, automakers lacked knowledge and were quite uncertain about the meaning of the law and the structure of their regulatory environment. While the EPA was sensitive the automakers’ uncertainty, the agency tried to keep the hearings focused on its own priorities.

In the hearings, Allen often handled the early questions. His goal was to address the legal issues as quickly as possible before handing the baton to Stork and

EPA’s technicians, who questioned the witnesses about technical issues.

Ruckelshaus did not want “legal fencing.” Instead, Stork and his staff staged a good, adversarial joust over engineering details. As Stork recalled, “I suppose [the debate over engineering] made a better and more credible show in terms of how deeply

EPA was really digging into what the heart of the issue was.”58 In the hearings, the

Office of Mobile Sources found an occasion to build credibility by showing off the fruits of its technical assessment program.

Still, for all of their work setting up hearings that would be focused on technological feasibility, members of EPA felt like the early discussions were dominated by industry lawyers and executives, who knew little about technical details. Stork characterized the presence of lawyers as “a dismal kind of thing.” But as Stork recalled, “We got that over with fairly quickly.”59

The hearings produced thousands of pages of transcripts, which were collected into thirteen volumes, too much to consider fully here. Instead, I will focus on the testimony of two companies—General Motors and Chrysler—which will

58 Eric Stork, NRC Transcript, 8 59 Eric Stork, NRC Transcript, 5.

293 serve to outline the contours of the debate between the automakers and EPA. On the one hand, I have chosen GM because it was, until very recently, the world’s largest carmaker, and its fortunes had large ramifications for the auto industry and, less so, for the US economy more broadly. On the other hand, one of Chrysler’s emissions control test cars, Car 333, which first came to public light during the hearings, became an important symbol and subject of debate throughout the period.

To understand the disagreement over the meaning of Car 333 is largely to understand Chrysler’s and the EPA’s diverging views on innovation. In another sense, if the two fundamental questions in the CAAA were about “feasibility” and

“good faith,” then the EPA’s debate with Chrysler focused on feasibility, while its exchange with GM centered on the question of the company’s good faith efforts.

General Motors: Trading Analogies of Good Faith

On Monday, April 17, 1972, the EPA held its first hearing in which it questioned GM. Though the EPA had already carried out six days of hearings, GM’s was the first of Big Three to appear. Expectations were high as people wondered how the United States’ domestic automakers would make their showing. The testimony of GM’s employees and the questions that the EPA asked exemplify the tension between the two parties. The EPA wondered why GM had not put more effort into developing catalyst systems; GM implicitly questioned how much authority the EPA had to force the auto companies to change the industry’s traditional industrial organization.

294

As with most companies, EPA’s early questioning of GM dealt primarily with spelling out the firm’s emissions control programs, including its relationship with external suppliers, and its most promising control system to date. GM had begun preparing for the ramifications of the CAAA even before Nixon signed it into law on

December 31, 1970. Indeed, GM President Ed Cole viewed the catalytic converter as a patentable and potentially profitable innovation, and he started a laboratory focusing on catalysts in 1968. In the summer of 1970, General Motors had surveyed the potential of six catalyst suppliers—W. R. Grace, Englehard, American Cyanamid,

Air Products, Monsanto, and OXY (Occidental Petroleum).60 Several months later,

UOP (Universal Oil Products) was added to the slate. General Motors estimated that it would require a hundred tons of catalyst a day once manufacturing operations were up and running.61 This potential market would be a huge prize to whichever company won the contract to supply GM with catalysts. GM also had an extensive and intensive internal testing program. It ran durability tests on the catalyst substrate using eighteen vehicles with catalysts from different suppliers.62 The cars were assembled in late December 1971, January 1972, and some into February of the same year.63 GM’s AC Delco Division had 128 people working on the catalysts.64

60 T. E. Hustead in Environmental Protection Agency, “Auto emissions extension: Transcript of proceedings [of public hearings on requests by five automobile manufacturers for a one-year suspension of the effective date of emission limitations applicable to 1975 model year automobiles],” vol. 6, (Washington, DC, Ace-Federal Reporters, Inc., 1972), 1356. Hereafter, “Auto Emissions Extension.” Since the document is a transcript, in the footnotes, I will attribute each claim to one of the participants. Hustead was the Head Engineer of GM’s AC Spark Plug Division. 61 Hustead, ibid., 1358. 62 R. W. Bowditch and Hustead, ibid., 1360-1. Bowditch was the Director of GM’s Automotive Emissions Control office. 63 Bowditch and Hustead, ibid.,1361. 64 Hustead, ibid., 1457. Hustead called it the “IC”: “I have 128 people at IC now.”

295

When the questioning moved beyond outlining these basic issues, the EPA essentially began to quiz GM about its “good faith” efforts, i.e., whether the company was really taking emissions control seriously. EPA officials were unimpressed by

GM’s early showing and became very frustrated because lawyers and financiers dominated the discussion on GM’s side. “We were pissed,” Eric Stork recalled. 65

People with real knowledge of the company’s technical systems hovered in the background, and only later did they become the primary respondents to EPA’s questions. This dynamic led EPA to demand that GM make a second appearance at the hearings

General Motors depended almost entirely on outside suppliers for catalyst development, but the way in which it depended on the catalyst companies without aggressively supporting research in those companies “disturbed” Ruckelshaus. He questioned whether General Motors would leave something, like an automatic transmission or power steering, that was important to its bottom line and that might give the firm “competitive advantage” to outside suppliers.66 Starkman pointed out that the auto companies often relied on external organizations to develop components that were essential to auto manufacturing. For instance, the auto companies relied on steel companies to make the alloys that formed the car’s body. This is not to say that the company did not do R&D. By the time of the hearings, GM had produced fifty-eight catalysts.67 Catalyst manufacture was a highly specialized field, however, and the catalyst suppliers had more expertise in

65 Interview with Eric Stork, January 5, 2010. 66 Ruckelshaus, ibid.,1342. 67 E. S. Starkman and Hustead, ibid., 1343. Starkman was the Vice President of GM’s Environmental Activities Staff.

296

“chemistry and chemical engineering.” One employee claimed that GM’s “area is more so engineering systems, putting the components together, if you will.”68

Ruckelshaus did not relent: “Supposing you were trying to achieve—let’s go back twenty years and you wanted—it was a question of competitive advantage over an automatic transmission. How was that developed?” A GM technician characterized automatic transmission as a “good example” of what the GM employees had been telling the panel.69 The mechanical aspects of automatic transmission, including the gears and the casting, had been relatively easy to figure out. The transmission fluid was much more difficult. “The slip factor,” he called it.

When GM first put out an Oldsmobile with an automatic transmission—“one of the first fully automatics on the market”—no fluid or oil could tolerate the heat generated. Yet, GM was not the company to develop the fluid that finally made automatic transmissions viable; petroleum companies created the fluid for GM, though GM “did the testing.”70 Thus, the development of automatic transmission shared a great deal in common with how automakers were relating to the development of catalysts. The automakers simply did not have the in-house expertise to address every problem that arose in the complex technical systems of the automobile.71 Ruckelshaus misread the companies when he assumed that they did.

68 Ibid. 69 R. C. Stempel, ibid., 1345. Stempel was a Motor Engineer in GM’s Oldsmobile Division and future president of General Motors. 70 Ibid. 71 Stork and an anonymous source have both claimed that the automakers and the SAE had been almost totally dominated by mechanical engineers until the 1970s, when chemical engineering and electrical engineering both became important to the automotive enterprise.

297

Yet, Ruckelshaus believed the issue hinged on capital. He said, “Probably with the oil companies there wasn’t a question of adequate resources as there might be here, though.” If the auto companies were willing to pour money into catalyst research, perhaps the endeavor could reach a larger scale and scope. The catalyst companies had expressed some misgivings about having enough cash to continue

R&D, and the auto companies were not supporting the research. But the GM employees claimed that “a very significant part of this development is the cost to test.”72 Therefore, by carrying out the tests, GM was sharing R&D costs. Ruckelshaus still wondered whether the auto companies had enough incentive to really push for innovation around catalysts, including supporting the research. He questioned whether the incentive was as strong “as it might be if you were to gain significant competitive advantage [through the research] as opposed to a statutory deadline.”73

GM’s representatives insisted, however, that the catalyst companies were not cash strapped and that they had plenty of incentive because becoming GM’s sole provider of catalysts would be a great boon.74

GM held that it was providing value to the catalyst manufactures by carrying out the testing and that the information about these tests was critical to the catalyst companies. Yet, an EPA staff member pointed out that, in their testimony, employees of Air Products said that they had sent catalysts to GM in January 1972, but that they had never received word from GM about how the tests went.75

72 Stempel, “Auto Emissions Extension,” vol. 6, 1346. 73 Ruckelshaus, ibid. 74 Starkman, ibid., 1347. 75 Norman Shutler, ibid., 1359. Shutler was the Director of the EPA’s Division of Mobile Source Enforcement.

298

An issue that would recur repeatedly centered on differing notions of what constituted a successful catalyst. GM had decided that a certain percentage of test vehicles had to complete the procedures successfully before the company was willing to put the control technologies on all of its vehicles.76 GM ran two concurrent testing regimens on its vehicles.77 In 1972, GM had thirteen vehicles running through the tests.78 One test revolved around the federal emissions criteria, including bulk emissions and durability. The second one was GM’s general test for all of their vehicle makes, which was based on a “matrix or statistical testing arrangement, whereby the engine transmissions, car bodies, size, are matched against the various parameters, such as emission performance, fuel economy, the various durability problems, and then on a selective basis, we can cover the high percentage or high volume applications, so we are assured that we have a proper number.”79 GM used this methodology to ensure that it could produce enough vehicles at a low enough cost to sell vehicles profitably.

Members of the EPA continued to be skeptical that GM’s general test ever limited its activity if competitive advantage was at issue. The EPA asked the panel when the test had ever become an issue during a moment of competition, not regulation.80 A GM employee pointed to the case of sports cars.81 As he outlined,

Volkswagen had brought out some sports cars that started doing well with US

76 Robert Sansom and Starkman, ibid., 1369. Sansom was the EPA’s Deputy Assistant Administrator for Planning and Evaluation. 77 Stempel, ibid., 1370-1. 78 Stempel, ibid., 1370. 79 Stempel, ibid., 1371. 80 Sansom, ibid., 1372. 81 Stempel, ibid., 1372-3.

299 consumers, and then Ford began offering the “boss Mustang, a really sporty type vehicle.”82 The horsepower wars had taken a new turn. GM began competing for this emerging consumer desire with the Oldsmobile 442. The 442 sported an open

“cold air hood” that gave it an aggressive, masculine image. Yet, these hoods had several potential problems, including “hood blow up, water entry into the engine, noise, exhaust noise involved with that system, and the durability of that hood.”83

GM submitted this system to its testing matrix. Since Oldsmobile planned only to release about 2,000 cars with this hood, the firm’s engineers were content when five cars had passed GM’s tests. The EPA pointed out that other manufacturers said that they would go forward with production if they could get five cars to pass the federal test procedures for emissions control. But the GM employee said that GM would not be so easily satisfied. Five cars passing the test was fine for a small-scale item like a muscle car, but emissions controls would be put on hundreds of thousands, and eventually millions, of vehicles. Moreover, emissions control was quite a bit more complex than a cold air hood. Although the hood required engineers to check how it affected other systems of the 442, emissions control would almost certainly require multiple new components, including new carburetors and the catalytic converter.

Each new system introduced new complexities and multiplied the difficulties of others. For this reason, GM had taken to component testing: some cars had new carburetors, while others had catalytic converters.

GM’s future plan, if it received a one-year suspension, would have two prongs. The first included further work on catalyst poisoning. Simultaneously, GM

82 Stempel, ibid., 1373. 83 Stempel, “Auto Emissions Extension,” vol. 6, 1373.

300 would work on other engine systems, such as the carburetor, the fuel mix, fuel injection, exhaust gas recirculation, manifold reactors, and others, so that the catalytic converter would not be responsible for handling as many emissions. As already noted, manifold reactors became increasingly important as engineers began to focus on the first few minutes after a cold startup, before the catalytic converter became hot enough to function at peak efficiency. During this time, emission levels were high, increasing a car’s overall emissions signature. Manifold reactors could help mitigate this problem.

General Motors returned for a second round of testimony on Wednesday,

April 26, 1972. From the beginning of the session, Allen pushed GM harder on why it had not poured resources into the catalyst companies in order to hasten the development of emission controls.84 But the GM employees stated that they were surprised even to be asked about this, as the catalyst companies had never come asking for resources. One said, “We supplied considerable aid, we think a disproportionate amount of aid to the catalyst manufacturers and their development programs by doing the testing for them, either the pilot banks testing .

. .” Allen cut him off. “You think that was disproportionate? In what sense do you mean that that was disproportionate aid to them?” he asked.85 The GM employee said that the automaker had provided the catalyst companies access to its labs and cars, something that the company normally did not do. But Allen was unsatisfied.

“Are you saying it is disproportionate because in your judgment the catalyst supplier had a higher interest in the screening of effective catalyst devices than did

84 Allen, “Auto Emissions Extension,” vol. 12, 2751. 85 Starkman and Allen, ibid., 2753.

301

General Motors?” The employee backpedaled. “Not necessarily,” he said, “I think disproportionate with respect to other parts of our product, the extent to which we will take samples of materials in various ways, paints, fabrics, plastics, metals, and test them and then say, well, they are good, bad, or indifferent I think is much less than the case of what we did in the catalyst area.”86 Another GM staff member put forward tires as an example of how GM usually organized the testing of externally developed technologies. At one point, GM had undertaken a “mutual program” with tire companies to change material from a “rayon fabric to a fiberglass type cord.”

Yet, the burden of the program had fallen on the tire manufacturers. He asserted that “the rubber companies did the initial screening and sent us in essence a tested product, if you will.”87 By furnishing this analogy, the GM staff member hoped to show that the company was making a serious effort to develop emissions controls by taking a more active role in testing. Allen was still not convinced. “Let me make .

. . let me try to put the record clear on this point: Do you contend that it would have constituted a good faith effort by General Motors to meet these standards if you had simply taken the position that you are willing to purchase from a catalyst supplier any proven devices that he can prove?” No. That wasn’t it, the GM employees insisted. All of their examples were aimed at showing that they weren’t simply sitting around waiting for external developments but were actively participating in the testing process. Finally, Allen put the situation in stark terms, “Well, I have a note here that . . . advising me that the total catalyst market is about a hundred million [dollars] per year. The automobile market is fifty billion [dollars] per

86 Starkman, ibid., 2754. 87 Stempel, ibid., 2755.

302 year.”88 But the employees insisted that the first time that GM had considered patronizing research at the catalyst companies was “last Monday,” and that they had found the very idea surprising. Robert Fri, the EPA’s Deputy Administrator, suggested that GM would at least want to consider using capital influxes to break specific, targeted bottlenecks in catalyst research.89 One GM staff member replied, “I guess what the question really comes down to is after having focused on those resources did we recognize that there was an area which might have been enhanced? I can only say in all due honesty obviously we didn’t, and I think also obviously neither did the catalyst suppliers.”90

By furnishing analogies drawn from the corporation’s history—analogies meant to explain how the industry “worked”—GM employees were simultaneously attempting to conserve the traditional organization of the industry. They were also

“educating” the EPA. The EPA, on the other hand, was trying to determine whether

GM had made “good faith” efforts by asking if good faith in this instance did not include changing traditional practices, including supporting outside research.

Chrysler: Trading Analogies about the Nature of Technical Change and “Feasibility”

The EPA and automakers clearly had differing models of technical change, which partly rested on the relationship between prescriptive knowledge about technology (knowing how to make something) and propositional knowledge about

88 Allen, ibid., 2758. 89 Fri, ibid., 2759. 90 Starkman, ibid., 2760-1.

303 it (knowing why something works).91 One EPA employee later characterized the

EPA’s understanding as being like Edison’s work on the light bulb—the automakers should take system components that worked and move forward with them.

Propositional knowledge about why that particular converter was successful was not nearly as important as carrying out systematic experiments on models that did work. The automakers, on the other hand, were focused on creating systems that worked over a wide spectrum, typically studied via statistical tests like the one GM used, and they believed that they had to understand why the systems worked in order to create robust systems. At the heart of these differing visions lay a Chrysler automobile. In the company’s emissions control tests, it was car number 333, which became “Car 333” in regulatory parlance. The vehicle had a manifold catalytic converter that had melted during testing. Because of the chemical reactions within them, converters get extremely hot, and melting was a common way that early catalyst systems failed. Normally, when the catalysts melted, the test was effectively over; the device was not going to make the cut. Yet, Car 333 was different. Though the catalyst had melted, it continued to function.

Stork’s technicians discovered the prospects of Car 333 as they mulled over the suspension submissions. In one version of the EPA’s story of Car 333,

Englehard, the company that made catalyst in the car, had tipped off the EPA that

Chrysler was going to bring the melted converter to the hearings as an object lesson

91 Joel Mokyr lays out the helpful explanatory dichotomy of prescriptive and propositional knowledge in his The Gifts of Athena: Historical Origins of the Knowledge Economy (Princeton: Princeton University Press, 2002), especially chapter one.

304 in the company’s inability to produce a viable emission control system.92 Once alerted, Stork’s technical staff examined the data on Car 333 and realized that, even though the catalyst had melted, the converter was relatively successful. If, during the hearings, a Chrysler employee held up Car 333’s melted, ugly converter, EPA staffers planned to ambush him with their findings that it met the standards.

Whether it was tipped off or not, the EPA questioned Chrysler’s employees about

Car 333’s performance.

Sydney “Syd” L. Terry, the company’s Vice President of Environmental and

Safety Relations, took the first questions after giving an opening statement. Unlike the other automakers, Terry began by saying that the company had made great headway on the catalytic converter and that it had some converters that performed above the CAAA’s requirement for tailpipe emissions. But none of the company’s efforts had come close to meeting the CAAA’s 50,000 mile durability requirements.

Terry testified that Chrysler was asking for suspension for three reasons: uncertainty about the standard’s requirements (especially about the availability of lead free gasoline, the permissibility of carrying out maintenance within the 50,000 miles the act required, and whether the EPA would do assembly-line and field testing), an inability to produce a system that met the requirements 50,000 mile, and concerns about “lead time,” that is, the amount of time required for design and planning before production could begin.93 He claimed, “Once these [areas of

92 Related by Eric Stork. 93 Syd Terry, 2065-6.

305 uncertainty] are established, we could immediately finalize engineering programs to determine and develop an optimum system.”94

Soon after the EPA asked Terry to clarify some of his points, the conversation moved to Car 333. The company had put a converter filled with “Englehard PTX,” a platinum monolithic catalyst, on the vehicle. Chrysler employees explained that, while Car 333 had fared well compared to other vehicles equipped with catalysts, it did so only because the engineers and test drivers had effectively “babied it.” They took great care not to overheat the catalytic converter, limiting it to “1400 degrees, with the occasional excursion to 1500.” They equipped the car with a “direct reading pyrometer” that allowed the driver to see the engine’s heat as he was driving.95 If the temperature spiked, he would back off the accelerator, thereby dropping the temperature. Because of this treatment, when the hearings began, Car

333 was still running.

Terry and his employees were surprised at the EPA’s interest in the car.

Terry said, “I think maybe we missed the point here somewhere. I tried to explain that the purpose of this test was to determine whether or not catalysts would live for say 40,000 or 50,000 miles, we hope, if we could keep the conditions of the catalyst the way it [i.e. the catalyst] wanted it [i.e., in ideal conditions]. So, we ran the test in such a way that it did not fail. We were trying to make it live.”96 Later,

Charles Heinen, Chrysler’s Executive Engineer of Materials, claimed that until it

“became a cause celebre, we did not realize that Car 333 was all that much fun.

94 Ibid., 2066. 95 Huebner, 2101 96 Terry, 2098-9.

306

When you fellows got into Car 333, it became the most prominent car we had ever had.”97

Beyond the issue of Chrysler’s testing Car 333 gingerly to guarantee the durability of its catalytic converter, the auto company also carried out frequent maintenance on the vehicle, which became a rub between the automakers and the agency.98 Chrysler changed the spark plugs on Car 333 every 5,000 miles, regardless of whether they had failed, to prevent overheating, which would have ruined the catalyst. Chrysler argued that this was too much maintenance to expect from consumers; the EPA didn’t think so. Indeed, the agency expected that automakers would need to communicate how the car should be used to the customer, including how it should be maintained. Stork even suggested that automakers should produce vehicles that were constructed in such a way that it would be very difficult for consumers to put tow hitches on them. Since towing trailers, boats, and other cars tended to increase the engine’s temperature and could ruin catalysts, producing cars not meant to tow made sense for an urban environment. That EPA believed regular, even frequent, maintenance was okay ran counter to long-standing traditions in the

American automobile industry, which were to make cars requiring as little maintenance as possible, and, in the case of the tow hitch, that met as many people’s demands as possible. The disagreement about maintenance and vehicle design went beyond mere quibbling over technicalities to the core of the CAAA’s standards.

97 Charlie Heinen, NRC Transcript, 4. 98 Stork, “I hope we can figure out how to make it as tough as it should be.” Stork, 2124.

307

If regular maintenance was expected, then it was more likely that the automakers would be able to meet the standards, since the catalyst would be less likely to foul or overheat.

Although little eventually came of Chrylser’s Car 333, at the time of the first hearing, it represented a difference in understanding between the EPA and the auto industry—and Chrysler in particular—about how “innovative” activity worked. On the one hand, the EPA held up Car 333 as being like Edison’s first successful light bulb. No one had any propositional knowledge about why the converter continued working even after it had melted, but in the eyes of EPA staff members, Chrysler could not have followed up on the converter in Car 333 because the company’s engineers did not view it as an example upon which to build. Chrysler, on the other hand, asserted that nothing could be made of one marginally successful case. Many more successful converters would be necessary—and converters that “succeeded” much more clearly—before the company would be making any headway. In time, the EPA came around to viewing things as the industry did, but only after the automakers took the agency to court.

The First Decision

On May 12, 1972, Ruckelshaus denied the suspension applications of all five automakers—Volvo, Ford, Chrysler, General Motors, and International Harvester.99

In his decision document, he argued that the automakers had simply not made the case that they could not produce a car that met the standard. Ruckelshaus claimed

99 Second Decision, 3

308 that the “most effective systems typically include: improved carburetion; a fast- release choke; a device for promoting fuel vaporization during warm-up; more consistent and durable ignition systems; exhaust gas recirculation; and a system for injecting air into the engine exhaust manifold to cause further combustion of unburned gases and to create an oxidizing atmosphere for the catalyst.”100

Ruckelshaus characterized systems that contained all of these sub-systems as utilizing the best “available technology,” and he believed such a system could meet the standards. Soon, however, Ruckelshaus learned that he would not have the last word.

International Harvester v. Ruckelshaus

International Harvester, General Motors, Chrysler, and Ford quickly responded to Ruckelshaus by sueing the EPA, claiming that he had not adequately supported his decision. The companies argued their case in front of the District of

Columbia Circuit Court on December 18, 1972; two months later the court ruled in the companies’ favor. Since many suits against federal bureaucracies go through the

DC Circuit Court, it has played—and continues to play—a central role in formulating administrative law, the rules that guide federal Executive Branch agencies. The court’s importance became increasingly true during the late 1960s and 1970s as lawmakers created a host of new regulations and regulatory agencies, many of which focused on managing complex technological risks. As one federal judge later pointed out, “Between 1966 and 1981, Congress enacted 182 new regulatory laws

100 From Ruckelshaus’ First Decision as quoted in the Leventhal Decision, 12.

309 and created 24 new regulatory agencies. Compare those figures with the longer period between 1946 and 1965 when only 58 new laws and 8 new agencies were brought into being.”101 Administrative law was extremely “open” and flexible at the time. Few issues were settled, and judges had immense power to shape the practices of federal agencies.

Judge Harold Leventhal wrote the court’s majority opinion in International

Harvester v. Ruckelshaus. Even prior to his time as a federal judge, Leventhal had spent considerable time thinking about business and industrial policy. He had worked as Chief of Litigation for the Bituminous Coal Division of the Department of the Interior during the New Deal (1939-1940) and spent two stints at the Office of

Price Administration during World War II (1940-3, 1946) and one at the Office of

Price Stabilization during the Korean War (1951-1952). His tour as Lieutenant

Commander in the U.S. Coast Guard Reserve during World War II and his work as a prosecutor on the staff of Justice Robert Jackson during the Nuremberg trials intervened in his work at the Office of Price Administration. After the Korean War,

Leventhal went into private practice for about fifteen years. In 1965, President

Johnson appointed him to replace Wilbur K. Miller, a long-standing judge in the DC

Circuit of the US Court of Appeals. Leventhal was confirmed by the Senate and began his tenure on April 7, 1965.

In many ways, Leventhal was a leader in the court; he set many of the doctrines that would come to dominate judicial practice. Leventhal staged what

101 Patricia M. Wald, “Thirty Years of Administrative Law in the D.C. Circuit Court,” Harold Leventhal Annual Talk at the DC Bar, July 1, 1997, http://www.dcbar.org/for_lawyers/sections/administrative_law_and_agency_practice/wald.cfm (accessed May 30, 2009).

310 one judge called the “debate of the decade” with Judge David Bazelon, the DC Court’s senior judge. The two disagreed over how deeply judges should look into agency practices, including the science that “supported” the agencies’ decisions. Bazelon argued that judges should only go so far as to ensure that agency practices would lead to “truth-finding.” Leventhal, on the other hand, contended that judges needed to examine the details of agency practices to see how fully bureaucrats were fulfilling their mandates. Leventhal’s position became known as the “hard look” approach, and it became the de facto philosophy of the court after the Supreme

Court struck down Bazelon’s view in 1978 in Vermont Yankee v. NRC.102 Though

Leventhal advocated looking hard at federal agencies, even he felt extremely restricted by his inability to master all of the technical details under consideration.

Nowhere was this more true than in dealing with environmental regulation, which often required extensive knowledge of science and technology.

Part of the Leventhal’s reservations about Ruckelshaus’ first decision came from a National Academy of Sciences’ report that seemed to contradict the reasoning behind Ruckelshaus’s decision, especially on the issue of feasibility.103

The Clean Air Act of 1970 directed the EPA to contract with the National Academy of

Sciences to complete a study on the technological feasibility of emissions control

102 Leventhal was also an early proponent of the related notion of “hybrid rulemaking.” The Administrative Procedures Act of 1946 divided rulemaking into “formal” and “informal” forms. Formal rulemaking had extensive requirements for how an agency could create rules, while informal rules could usually be promulgated through the Federal Register with little ceremony. Leventhal believed there were cases that fell between these two extremes. In these cases, some procedures, such as hearings and input from industry and public interest groups, were necessary for making good administrative judgments, though those procedures might not reach the heights of “formal” rule- making. 103 Committee on Motor Vehicle Emissions, National Academy of Sciences, Semiannual Report to the Environmental Protection Agency, January 1, 1972; idem., Interim Standards Report, April 26, 1972.

311 systems. The contract was meant partly as a check against the EPA’s internal judgment; an external “non-interested” scientific group would also determine whether the automakers could meet the standards, hopefully granting greater credibility to the process. The NAS formed an impressive Committee on Motor

Vehicle Emissions, including A. J. Haagen-Smit, the father of the science of air pollution and co-discoverer of smog in California, and the chairman of the committee, Edward L. Gintzon, an eminent electrical engineer and founding member of Varian Associates, a high-tech company at which Gintzon was working when he was on the committee. The committee first met on June 16, 1971 and came together monthly afterwards.104 The members made their recommendations on the basis of advice from eight panels of experts. Seven of the panels began work in 1971, but the catalyst panel was not assembled until “early in 1972,” when the committee realized how complex and controversial the subject was.105

The committee focused almost wholly on feasibility, with subordinate and related interests in long-term performance and cost. Unlike the automakers, the committee was especially keen to “open” the standard gasoline engine and consider alternative systems; indeed, they feared that, if pushed too aggressively, federal regulations might lock-in a sub-optimal system instead of allowing more ideal systems to emerge. This idea was a popular fear at the time. As one news article asked, “Did the Clean Air Act’s strict timetable push the auto industry down the catalyst road to the detriment of more promising antipollution technology, as Ford

104 Committee on Motor Vehicle Emissions, National Academy of Sciences, Semiannual Report to the Environmental Protection Agency, February 12, 1973, 7. Hereafter, NAS Report. 105 Ibid., 8.

312 and Chrysler have suggested?”106 Stork answered this question doubtfully, “I think there is a far greater likelihood that the automobile industry now will seek alternatives to the catalyst than it would if they didn’t have to face up to the catalyst.”107 He added that the industry was more likely to develop alternative engines if they were better than an already-existing technology, namely the Otto engine with catalytic converters in the exhaust system.108 Stork had no inkling of how locked-in the catalyst would become over time. Perhaps, the NAS committee members did. They wrote, “Four types of systems will meet the prescribed emissions standards during certification testing. These are: the modified conventional engine equipped with an oxidation catalyst, the carbureted stratified- charge engine, the Wankel engine equipped with an exhaust thermal reactor, and the diesel engine.”109 They went on to claim that automakers must be allowed one catalyst change during the 50,000 mile durability tests for catalytic converter systems and that the EPA had to create a rigorous national inspection program because no one knew how the converters would perform under daily stress. More importantly, the committee stated that automakers could produce these systems in adequate quantity to meet demand for the 1975 model year at a per vehicle cost increase of about $160 over 1973 cars.110 Yet, the NAS ultimately suggested creating

106 Donald A. Colburn, “Washington’s RX: Will the Cure Be Worse than the Disease?” Automotive Industries, February 15, 1974. 107 Ibid. 108 Ibid. 109 NAS Report, 2. 110 NAS Report, 3. The NAS also insisted that tests results should be averaged “within a vehicle and engine class.”

313 interim standards that were greater than the 1974 standards but less than the legislatively mandated 1975 standards.

With the NAS’s report in hand and his own concerns in mind, Leventhal remanded the EPA to reconsider its original decision. He placed no restraints on

Ruckelshaus other than to remind him that, as the NAS had suggested, interim standards were a possibility within the mandate of the CAAA and that they were a probable short-term solution to the disagreement between the EPA and automakers.

Second decision: Trust in Numbers, Trust in Others.

Once the DC Circuit Court remanded the decision back to EPA, Ruckelshaus and his employees began planning a second round of hearings. The format would be roughly the same as the first round. The majority opinion had ruled against

Bazelon’s desire to see cross-examination enter into the proceedings. Yet, the EPA made at least one very important change: Coming out of the Leventhal decision, the

EPA realized it had to create a more rigorous methodology for statistically averaging the companies’ test results. In many ways, the methodology that the EPA created settled the meaning of “feasibility.” The methodology came to function as a mediator between the EPA and industry; there was an agreed-upon, dependable standard that both parties understood.111 As Stork later pointed out, “We were not sued on the second decision.”112 Over time, staff members of the EPA and employees of the

111 The dynamic here strongly resembles the one that Theodore Porter describes in Trust in Numbers, where mechanical mathematical devices are substituted for interpersonal confidence. 112 Eric Stork, NRC Transcript, 12.

314 automakers came to know and trust each other, but in this early period, the methodology played an important role as arbiter. Yet, if earlier Ruckelhaus had so actively considered how to be trustworthy in the public’s eyes, the EPA’s choice to adopt the methodology—and Ruckelshaus’s second decision, which emerged from this choice—built trust with the auto industry but alienated other parties, including consumer advocates and some members of Congress.

The agency announced its preliminary methodology, which employees had developed in-house, on March 9, 1973.113 Then, during the second hearing, the EPA held an informal, off-the-record meeting with the automakers on March 17 to hear comments and to hash out a compromise on the methodology’s details.114 Stork characterized the meeting as a “bull session.” The EPA participants had debated internally whether the meeting should be recorded and transcribed and finally decided against it, probably because they believed that the industry employees would not be candid if they knew they were being recorded or, more likely, that the industry lawyers would limit what the engineers said if a tape recorder were running.115 Clarence Ditlow, the well-known consumer advocate and critic of the automobile industry, said of the decision to hold the meeting off-the-record, “That was one of our greatest public interest objections.”116

113 Second Decision, 23. Anecdotally, the core of the methodology may have been based on General Motor’s production formulas discussed earlier in the chapter. A few former and present EPA employees suggested this to me in passing. 114 Second Decision, 23. 115 Bill Pedersen, NRC Transcript, 17. Pedersen was an employee in the EPA’s General Council’s Office during the hearings. 116 Clarence Ditlow, NRC Transcript, 17.

315

The Office of Mobile Sources and the automakers met during a massive snowstorm in Ann Arbor. The storm added a sense of urgency to the proceedings.

Stork recalled the meeting as “very, very informal, noisy, and useful.” During the week following the meeting, the automakers sent in criticisms of the methodology.117 Those criticisms that the EPA found to be justified were incorporated. The finalized methodology was based upon Monte Carlo statistics, which had first been developed to solve intractable problems with nuclear weapons design. The historian of science Peter Galison has called Monte Carlo simulations an

“artificial reality.” 118 Monte Carlo simulation became a way to model the random, or stochastic, processes inherent in any real-world procedure. If Monte Carlo simulations originally modeled bomb tests, in emissions control, they represented other realities full of variability—namely variations in the performance of emission control devices and the vagaries of the shop floor. Though automakers were at this time implementing increased mechanization and automatic machine control, tolerances (and materials) were—and still are—wide enough that one car would perform differently than another of the same automotive model. As observed earlier, catalytic converters also performed over a broad spectrum, partly because of their inherent properties and partly because of the wide variability in the cars in which they were installed. The EPA’s Monte Carlo methodology became a way to model this wide range of performance. Some cars would perform better; some, worse. But once the statistically average vehicle superseded the mandated

117 Second Decision, 23. 118 Peter Galison, Image and Logic: A Material Culture of Microphysics (Chicago: The University of Chicago Press, 1997), ch. 8.

316 standard, then both the EPA and the automobile industry would know that the technology was “feasible.”

The EPA created a database for each manufacturer. Each car model was placed in an “engine category” and was tested for “best systems.” As the EPA administrator’s decision from the second hearings claimed, “a ‘best car’ analysis has been avoided and a ‘best systems’ analysis has been pursued.”119 The decision captures well the distance both the EPA and the automakers had moved since the

EPA’s staff members fixated on Chrysler’s Car 333 during the first hearings. Any single car was not a substantial enough ground on which to base a policy. By developing the methodology, the EPA had created a method of analysis that the industry could not gainsay as a rational basis of rule-making and enforcement. But in the process—and due in large part to Leventhal’s decision—the agency had moved much closer to the auto industry’s vision of innovation. The mathematization of the EPA’s decision-making was also a harbinger of things to come. Increasingly in the late 1970s and, even moreso, in the 1980s and afterwards, agencies were required to carry out cost-benefit analyses and other formal decision- analytic tools before implementing any consequential rule.

Interestingly, the mathematical methodology that EPA established quickly became unimportant. After nearly three years of searching, I have not been able to locate a copy of the methodology in the public record, the National Archives, or in private collections of present and former EPA employees. The substance of the methodology was less important than its very being. The methodology, it seems,

119 Second Decision, 24.

317 acted like what theorists call a “vanishing mediator,” an object that is necessary for the passage from one moment to another but then disappears after the transition is complete.120 The methodology only played a role in settling a disagreement between the EPA and the automakers.

On March 12, 1973, the EPA commenced over two weeks of hearings in answer both to the remanded applications for suspension—by Volvo, Chrysler, Ford,

General Motors, and International Harvester—and to American Motors Company, which had filed an application on March 2, 1973. Things moved quickly. The court had announced its decision on February 10; the hearings began a month later; and, on April 11, Ruckelshaus announced his second decision: he would grant the suspensions and set interim standards.

Five days later, Ruckelshaus, Allen, Stork, and Sansom appeared before

Senator Edmund Muskie’s Subcommittee on Air and Water Pollution, the committee that had birthed the Clean Air Act amendments of 1970. The first three men testified for three days. Muskie was not pleased. He began the hearings by noting that, after over 15 years of public research on air pollution, the Congress had passed the CAAA in 1970. Yet, now the automobile companies said they could not meet the mandated standards. He demanded, “I want to know why not. I want to know what the industry has done in the past 3 years. I want a public explanation from the industry for the path they have chosen, a course that has not been altered since

120 See Slavoj Zizek, For They Know Not What They Do: Enjoyment as a Political Factor, 2nd ed. (London: Verso, 2002), 182–187. The notion was originally spelled out in Frederic Jameson, “The Vanishing Mediator; or, Max Weber as Storyteller” in The Ideologies of Theory, vol. 2 (Minneapolis: University of Minnesota Press, 1988).

318

1969. I want to know what the industry is going to do in the coming year to overcome past failures. I want to know what commitment the industry is willing to make to the American people. And, I intend to challenge the assumptions on which the industry’s failures have been based.”121 Of course, Ruckelshaus could not answer for the industry per se, though he defended his decision to suspend the regulations for a year and to impose interim standards.

Ruckelshaus’s second decision can be interpreted in multiple ways. For some, the methodology seemed to end the controversy. Stork later pointed out that not only was the EPA not sued on the second decision, but it also had not been sued over the issue of determining feasibility since the second hearing, though the methodology was applied annually. This is a difficult claim to assess. One could easily make the argument that the automakers were appeased once the Leventhal decision effectively placed the burden of proof on the EPA’s shoulders, significantly weakening the agency against industry. More likely, however, is that the EPA learned considerably more in the months intervening between the two decisions.

Stork’s staff continued to carry out technology assessment. They had almost daily interaction with employees in the auto industry. Less visible than the public hearings but no less important to the EPA’s thinking were the technology assessments that the agency continued to produce throughout the period.

121 Senate Committee on Public Works, Decision of the Administrator of the Environmental Protection Agency Regarding Suspension of the 1975 Auto Emissions Standards, 93rd Congress, 1st sess., April 16- 18, 1973, 52.

319

Conclusion: Establishing the State of the Art

This chapter has explored the varying ways regulation produces knowledge both on the part of regulators and on the part of the regulated, and additionally for the public, including advocates of different causes. The regulators and the regulated both have to learn about the regulatory environment as defined by the law and the courts while simultaneously trying to shape that environment to fit their respective visions. The opposing bodies also learn about each other. Finally, the greatest knowledge of all is gained about the underlying science and technology that the regulations foster.

In 1969, following a Justice Department investigation into anti-competitive behavior around emissions control technologies, the major auto companies signed a consent decree, promising that they would not share information about their research and development efforts. The Justice Department held that sharing information among a corporate oligopoly led to anticompetitive behavior, and the department was watchful of any communication between the Big Three. The automakers were, in turn, nervous as they headed into the suspension hearings.

They believed that if they shared minute details of their R&D projects at the hearings, they would, in effect, violate the consent decree and incur the Justice

Department’s wrath. This possibly led to the inclusion of additional lawyers during the early hearings, which so disgusted Stork.122

The automakers regularly requested that the government drop the consent decree and its position that information should not be shared. In the first

122 Charlie Heinen, NRC Transcript, 5.

320 suspension hearings, the EPA’s George Allen surreptitiously maneuvered the automakers into a place where they had to admit that the consent decree and not sharing research was the best arrangement. The automakers themselves required catalysts makers to carry on independent research that did not include sharing research with each other. Allen asked one of GM’s employees about this arrangement, “Do you think this is desirable, this independent research by these catalyst companies?” “Yes,” the employee answered, “I think it is very desirable. I think because of this approach it has furthered the state of the art, and I think that is why there has been this great improvement over the last two years.” Allen responded, “That is the only trick question I asked you guys today, and I am glad to hear you say that because about every month Mr. Hilder [a lawyer who acted as

GM’s Assistant General Counsel] shows up and tries to get me to go to the Justice

Department Anti-trust Division to get permission for you guys to exchange information with your competitors.”123 The consent decree would stand, but the automakers were free to share information when testifying before the EPA.

After the auto companies realized that they would not face a suit from Justice, they began to share more openly at the hearings, which became an effective, yet decidedly unofficial, means for circumventing the consent decree. The companies still could not join each other in R&D projects, but they learned, in a basic way, how the technologies of their competitors were developing.124 One could understand,

123 “Auto Emissions Extension,” 1528. 124 The hearings and the EPA’s technology assessments were, by no means, the only ways that automakers and suppliers were learning about the current state of the art. The publications and conferences of the Society of Automotive Engineers (SAE) were also important venues, although

321 then, why Stork called the hearings “the continuing seminar on automotive engineering.”125 Allen characterized it this way: “The decision process itself, in that sense, undid the disabilities created by the consent decree; that is, it became a clearing house for information, and perhaps it is better that it worked that way than that the exchange of information occurred privately prior to the hearing.”126 It was better, perhaps, because the exchange happened in an open, on-the-record environment, which freed the process from suspicion. In developing a process whereby the EPA produced the truth it needed to make its decision, it had also aided the automobile firms in securing valuable markers by which they could guide their research. However unintentionally, the state had produced knowledge for which firms often pay large fees to consultants and analysts.

Following the second decision, the debate over the potential of catalytic converters changed considerably. As described in Chapter Four, by late 1973, the primary question was whether catalytic converters created sulfate particles as a by- product, a material that would be more harmful than the chemicals the converters were supposed to control. A Business Week article ran a lead line stating, “Warning:

Catalytic converters may be harmful to your health.”127 As these new issues came to the fore, the issues of what constituted “feasibility” and where the state of the art lay faded into the background. Discussions about how automakers would meet emissions standards, especially the NOx standard, continued throughout the 1970s,

there is some anecdotal evidence that SAE members were very careful about not violating the consent decree at SAE meetings. 125 Eric Stork, NRC Transcript, 4. 126 George Allen, NRC Transcript 7. 127 “Do Auto Catalysts Meet the Safety Test?”, Business Week, October 27, 1973, 116.

322 but, in many ways, the first suspension hearings accomplished their goal; they produced vital knowledge and, ultimately, established the state of the art.

Chapter 6—The Long Road to Federal Fuel Economy Standards: Capabilities- Building, Bureaucratic Self-Defense, and the Fine Art of Gamesmanship

“The real issue is that the auto industry would like to get those mothers at the EPA out of fuel economy testing. All real influence depends on the ability to obtain, assimilate, and process information.” — Eric O. Stork, Deputy Assistant Administrator, Environmental Protection Agency, Office of Mobile Source Air Pollution Control1

In his home in Arlington, Virginia—where many current and former civil servants live—Eric Stork has hung a large poster-sized drawing.2 Stork headed automotive pollution control efforts of the Environmental Protection Agency (EPA) from 1970 to 1978 and received this drawing from his staff when Stork’s new superior, David Hawkins, a new EPA Assistant Administrator appointed by

President Jimmy Carter, effectively axed him. At the center of the drawing stands a caricature of Stork, characteristically puffing on his pipe and dressed in plate and chainmail armor. He looks like a medieval knight. His foot rests on a dragon he’s slain, a sword still sticking out of its side. The dragon resembles a car. A caricature of David Hawkins—who has been drawn to look like a child holding a wooden, toy sword—stands off at the side. The tiny Hawkins squeals, “I wanna be like you when

I grow up.” A caption below the central image states, “Eric ‘The Iron Duke’ Stork”—a sobriquet that he earned from being tough on the industry—“never says ‘shame on me.’” Drawing on an old adage, Stork had added his own spin, “Screw me once, shame on you; screw me twice, shame on me; and I never say shame on me.”

1 As quoted in Douglas Williams, “EPA Admits Blunder on Mileage Figures,” Detroit Free Press, May 30, 1974. 2 A photograph of the poster is in the author’s personal collection.

323 324

Other caricatures surround the central image, each accompanied by one of

Stork’s oft-repeated phrases. For the purposes of this chapter, one of these images and phrases stands out. It pictures Stork waist deep in water. Still puffing on his pipe, he’s using a stick to beat back alligators, each of which has the name, or acronym, of a federal institution on its back—Congress, the Office of Management and Budget (OMB), Department of Energy (DOE), Department of Transportation

(DOT), and two other offices within the EPA, the Office of Planning and Evaluation

(OPE) and the Mobile Source Enforcement Division (MSED). Below it is written, “It’s hard to remember you’re supposed to drain the swamp,” part of a longer caption that concludes, “when you’re up to your ass in alligators.” In Stork’s context, this meant, “It’s hard to remember I’m supposed to be decreasing auto emissions, when

I’m fighting off all of these other federal agencies.” Clearly, bureaucratic infighting and belt-way battling had an important place in Stork’s work.

This chapter tells the story of how the EPA’s Office of Mobile Source Air

Pollution Control (hereafter, Office of Mobile Sources), out of a desire to protect itself from external criticism, built the capabilities to measure automotive fuel economy and how it then fought to retain those capabilities when other federal agencies sought to take them for their own. If bureaucratic organizations grow new capacities, sometimes they must also struggle to sustain them in the face of threats, even threats that come from other organizations within the state. In 1978, as Stork reflected on his past experience in a seminar presentation, “The Genesis of the EPA

Fuel Economy Program: An Exercise in Bureaucratic Gamesmanship,” Stork focused on the EPA’s involvement and battles over measuring and publishing data on

325 automotive fuel economy.3 Stork’s written text for this seminar provides a unique lens by which to see one such battle.

The First Fuel Economy Study: Capacity-Building as Self-Defense

Throughout the early 1970s, the automakers and some members of the public began increasingly to attack federal automotive emission standards for increasing fuel use or, put another way, decreasing fuel efficiency. Part of this trend was simple misallocation of blame by the auto industry. As Stork has often said,

“The auto industry had a long and proud tradition of producing lemons. Now, they had someone to blame it on.”4 Yet, many early (i.e., pre-catalytic converter) engineering solutions to automotive emissions control, including “enleaning” the fuel mix and retarding the “spark timing,” definitely increased fuel use.

The Office of Mobile Sources experienced consumers’ dissatisfaction with fuel efficiency mostly through letters written by citizens to their representatives in

3 Eric Stork, “The Genesis of the EPA Fuel Economy Program: An Exercise in Bureaucratic Gamesmanship.” The written text, from which Stork worked as he shared his reflections at several seminars, is in Eric Stork’s personal collection, and a copy of it is in the author’s possession. Bureaucrats often take the hatchet to a person by giving him or her a position that has a very honorable and fancy sounding name but that is basically meaningless and powerless. Hawkins made Stork his “special assistant.” Stork bridled, and he asked Hawkins to endorse his taking an Intergovernmental Personnel Act (IPA) assignment. The IPA fostered temporary movements between federal agencies, universities, and other institutions in order to strengthen institutions. As I will describe later, Stork knew professors who had used the IPA to take temporary assignments in government. Stork was considering doing the opposite, spending some time in academia. On one of his regular trips to the Office of Mobile Sources’ Ann Arbor office, Stork had dinner with David Ragone, the Dean of the University of Michigan’s Engineering School. Ragone had sat on one of Stork’s Advisory Committees while a professor at Carnegie Mellon University. He’d moved on to Dartmouth before taking the Deanship at Michigan. During dinner, Stork asked Ragone how one went about getting an IPA. Ragone told him that you contacted schools at which you were interested in being assigned. If they were interested in you, they would invite you to come give a seminar. “What is a ‘seminar’?” Stork asked. Ragone explained. Stork then wrote his presentation and gave it at five universities. He eventually decided to take his IPA assignment at Purdue University. 4 Phone interview with Eric Stork, January 5, 2010.

326

Congress, who then passed on the notes to EPA’s administrator before they found their way to Stork’s desk. Every agency must learn how to deal with these letters, which are sometimes called “Congressionals,” or face being swamped by them. For an agency dealing with automotive emissions control, the Congressionals ran a wide gamut, from worries about the health effect of automobile exhaust to complaints from consumers who had bought lemon cars. One way that Stork handled such letters was to create what he termed “Fact Sheets,” pamphlets that ran from one to a few pages and that explained some basic issue regarding automotive emissions control.5 If the Office of Mobile Sources received a number of Congressionals on a single issue, it would create a Fact Sheet on the topic. A Fact Sheet would then be sent out as a reply whenever applicable letters came in.

In late-1970 through 1971, the EPA continued receiving a rash of letters from people concerned about the effect of emission controls on fuel efficiency. Stork asked his staff members to find some way to figure out how much emission controls were affecting fuel efficiency.6 They said that there was no way to do it. Stork thought this had to be wrong; he’d been hearing claims about fuel efficiency and which cars were gas-guzzlers his whole life. The problem, his staff members said, was that there was no accepted, “standard” method for addressing fuel economy, but the staff members quickly found a solution. The men in charge of the agency’s technology assessment program, Tom Austin, Karl Hellman, and others, hit upon a preexisting method for measuring fuel economy. Created at General Motors, the

5 The Office of Mobile Sources collected most, if not all, of the Fact Sheets from the 1970s in a binder—a complete copy of which is in the author’s possession—in the library of its Ann Arbor office, the National Vehicle and Fuel Emissions Laboratory. Hereafter, cited as NVFEL Library. 6 Phone interview with Eric Stork, January 5, 2010.

327 method used hydrocarbon tailpipe emissions from dynamometer tests—that is, the same kind of procedure used in emissions control testing—to estimate the amount of fuel consumed. Although none of the Office of Mobile Sources’ publications from the period cite the methodology’s origin, it was probably Donald L. Stivender’s 1971 paper, “Development of a Fuel-Based Mass Emission Measurement Procedure.”7

Stivender was a forward-thinking engineer who worked in General Motors’

Research Laboratories. He specialized in emissions controls, electronic controls

(which were directly related to controlling emissions), and alternative power systems, especially gas turbines.8

As recounted in Chapter Two, the NHSB recycled different Society of

Automotive Engineers protocols and General Services Administration requirements to build the original crash safety standards. Similarly, the development of automotive fuel efficiency standards shows how regulation is often a contingent and incremental process. Regulators, including the scientists and technical experts who design regulatory criteria, use techniques, both human actions and technical artifacts, that reside elsewhere. They rarely invent techniques whole cloth; indeed, they avoid doing so because such work consumes precious time and resources. In this case, the staff members at the Office of Mobile Sources adapted an existing methodology for a purpose different than originally intended.

That Stivender’s article became the basis for the EPA’s fuel economy measurements is itself ironic with unintended consequences heaped upon

7 Stivender, “Development of a Fuel-Based Mass Emission Measurement Procedure,” SAE Paper 710604. The paper was delivered at the SAE’s Mid-Year Meeting, held in Montreal, Canada, June 7– 11, 1971. 8 On Stivender’s biography, see Who's Who in America - 2007, 61st Edition (2006).

328 unintended consequences. In his paper, Stivender spelled out an alternative means for measuring the mass emissions of regulated pollutants. His central concern was this: the EPA emissions test method, even the updated “constant-volume sampling” method that became standard in 1972, tended to be more accurate with traditional, four-stroke, “Otto cycle” engines. Because he was experimenting with alternative power systems, Stivender wanted to create a procedure that more accurately measured the emissions. That Stivender’s procedure also created a measurement of fuel economy was largely an ancillary point and not his primary goal. Of course,

Stivender and other venturesome engineers believed that increased fuel efficiency would come from alternative power systems. Fuel economy measurements were a by-product of Stivender’s method, not its goal.

Firms often scan the environment for new “innovative” technologies or processes; sometimes they even have staff members or offices dedicated to this task.

In economics, strategic management, and other fields such behavior is sometimes known as “technology scouting” or “technology search.” A firm’s ability to locate and incorporate new technologies can give it a competitive advantage over other companies. Conversely, a company that does a poor job at searching for new technologies will often be left behind by upstarts and other competitors and eventually blown away by, what Joseph Schumpeter called, the “winds of creative destruction.”9

9 The ability of an agency to find and assimilate external technologies has long been one of the practical upshots of R&D. In 1903, executives at Du Pont Company discussed how creating a new R&D lab would allow the company to absorb “inventions and innovations . . . brought . . . from exterior sources.” David A. Hounshell and John Kenly Smith, Jr., Science and Corporate Strategy: Du Pont R&D, 1902–1980 (Cambridge: Cambridge University Press, 1988), 29. The economists Wesley M.

329

Federal regulatory agencies also feel external pressures. In the case of fuel economy, the Office of Mobile Sources experienced the pressure of bad press and letters from constituents to Congress. Staff members searched for a new methodology that would allow the organization to create dependable, repeatable measures of fuel efficiency. By monitoring the SAE’s conferences and publications and paying close attention to recent developments, staff members grasped

Stivender’s paper, and the agency poured resources into developing the new means to measure fuel economy.

For the Office of Mobile Sources, the real insight in developing Stivender’s procedures was in realizing its wider application for the agency’s needs.10

Importantly, however, Austin, Hellman, and other members of the Office of Mobile

Sources were not only “borrowers” of available studies of fuel economy; they were also important contributors to the cutting-edge of this science. Of the twenty-two technical papers in an Society of Automotive Engineering (SAE) volume that collected the most important papers on fuel economy, Austin, Hellman, or both men

Cohen and Daniel A. Levinthal call an organization’s ability to “recognize the value of new, external information, assimilate it, and apply it to . . . ends” its “absorptive capacity.” And they claim that R&D plays an important role in building such capacity. While the Office of Mobile Sources headed away from R&D, technology assessment played a similar role in identifying worthwhile external information. Cohen and Levinthal, “Absorptive Capacity: A New Perspective on Learning and Innovation,” Administrative Science Quarterly, Vol. 35, No. 1 (March 1990), 128–152. See also, K. D. Miller and A. T. Arikan, “Technology Search Investments: Evolutionary, Option Reasoning, and Option Pricing Approaches,” Strategic Management Journal, Vol. 25, No. 5 (2004), 473–486; M. S. Brenner, “Technology Intelligence and Technology Scouting,” Competitive Intelligence Review, Vol. 7, No. 3 (1996), 20–27. 10 Here, following the language of “research and development” (R&D), I’m suggesting that Stivender’s paper constituted the actual “research,” while EPA staff members “developed” the procedure by putting it into practice, while simultaneously transforming it from its original intention.

330 wrote four, or nearly one-fifth, of them.11 The EPA’s technology assessment program grew into an important source of generating knowledge.

The EPA’s fuel economy procedures used a “carbon-balancing method”; by knowing the amount of carbon in the fuel and then measuring how much carbon the vehicle emitted (in different forms—CO, CO2, and hydrocarbons) during the test cycle, testers could calculate how much fuel the vehicles had burned. For the Office of Mobile Sources, the advantage of this method over other ones was that the only information the test required was carbon emissions. And the Office already had this information in spades. Indeed, the agency was the largest depository for such cross- firm information in the world: it had emissions data on all of the manufacturers who sold vehicles in the United States going back to the mid-1960s, when the original

Motor Vehicle Pollution Control Act took effect. The key innovation of the agency’s staff members was employing these historical data to carry out a number of studies on vehicle fuel efficiency. They could effectively use previous emissions certification tests and other in-house databases for examining historical trends in fuel efficiency.

Staff members used the pre-1968 data as the baseline. Therefore, the Office of

Mobile Sources could determine with certainty the percentage of lost efficiency due to emission controls.

The EPA had acquired the pre-1968 data through its In-Use Surveillance

Program, which had been running since the mid-1960s by the Office of Mobile

Sources’ predecessor agency in the National Air Pollution Control Administration.

11 Society of Automotive Engineering, Automotive Fuel Economy (Warrendale, PA: Society of Automotive Engineers, 1976). In my count of twenty-two papers, I am neglecting one essay, which recounts an SAE’s development of fuel economy standards but which does not generate new knowledge in the same way that the other papers do.

331

The In-Use Surveillance Program was important for the EPA’s overall air pollution control efforts because it was used to give the most accurate picture possible of how fleets of cars were emitting controlled chemicals in cities. These data were then used by cities and states to generate Air Quality Implementation Plans.12 When

Stork came to the EPA, the In-Use data had more or less stacked up without being analyzed.13 As a rule, contractors carried out the In-Use studies, but when the

Surveillance Branch requested money for further studies, which had become a matter of habit for the branch, Stork refused to authorize further contracts until the already-existing studies had been fully examined.

At about this time, a woman named Marcia Williams contacted Stork.

Williams had earned her bachelors degree in math and physics at Dickinson College, where she graduated in 1968.14 She then went to the University of Maryland to do graduate work in physics, but after seeing that all of the graduating PhD physicists were struggling to find jobs, she left after a semester. For a few years, she worked at

International Telephone and Telegraph, where she assisted in studies on ionispheric

12 The histories and structures of air quality implementation plans and National Ambient Air Quality Standards (NAAQS) are extremely complicated. To put it simply, the Clean Air Act was in part a federalist solution to the air pollution problem. While the federal government did regulate the amount of pollutants certain technologies and industries could emit, it was up to localities to regulate the overall level of ambient air pollution. For instance, the Clean Air Act might reduce certain automotive air pollutants by over ninety percent, but if enough cars—even very low polluting cars— are driving in one place at one time—say, in a city center—then there will still be an unhealthful and illegal number of pollutants in the ambient air. Localities, primarily cities, were to design programs to reduce the ambient air pollution by, for instance, limiting industry and the amount of cars on the road. The In-Use Surveillance program fit into the overall project. Generally, the NAAQS program and its federalized structure of air pollution control has been a miserable failure, with many cities’ being in constant and ceaseless violation of federal standards. On NAAQS, see Sheila Jasanoff, The Fifth Branch: Science Advisers as Policymakers (Cambridge, Mass.: Harvard University Press, 1990), 102– 104; Mark R. Powell, Science at EPA: Information in the Regulatory Process (Washington, D.C.: Resources for the Future, 1999), 90–94. 13 Email from Eric Stork, February 6, 2011. 14 Phone Interview with Marcia Williams, February 16, 2011. Marcia William’s Curriculum Vitae, in the author’s possession.

332 physics. In the fall of 1970, her husband was then accepted into a biostatistics program at the University of North Carolina, so she followed him there, where she eventually found a job at the EPA’s R&D branch at Research Triangle Park. Williams worked there for a few years, primarily doing mathematical analyses to support the agency’s ambient air quality standards. The data for these analyses came from a series of local, urban studies known collectively as the EPA’s Community Health and

Environmental Surveillance System (CHESS).

After two years, William’s husband accepted a job at the University of

Michigan. Williams contacted Stork in hopes that the Ann Arbor office might have a job for her. Stork’s response? Williams later recalled, “Eric’s comment to me . . . I can still hear him saying it . . . he said, ‘Well, if you’re good, I will find a way to help you.

If you’re not, there’s nothing I can do.’”15 Stork told her to talk to the Office of Mobile

Source’s Surveillance Branch in Ann Arbor. He let branch managers handle their own hiring, but as Stork later wrote, “Since the Branch was to [receive] no more contract money until I got their reports, the Branch Chief had incentive to hire a non-engineer who might help his program survive.”16 Williams was hired on, and over the course of the next year, she produced the first-ever emissions study of vehicles In-Use, which went on to become a regular and important series by the agency.17 John DeKany, another EPA branch chief in Ann Arbor, eventually “stole”

15 Ibid. 16 Email from Eric Stork, February 6, 2011. 17 For one such In-Use study, see Marcia Williams et al. Automobile Exhaust Emission Surveillance of the FY 72 Program, U.S. Environmental Protection Agency, Office of Mobile Source Air Pollution Control, February 1974.

333

Williams and made her his program planner.18 Austin and Hellman used these In-

Use studies as a baseline in their fuel economy study because the In-Use studies captured pre-1968 vehicles that did not have emissions controls.

In November 1972, Austin and Hellman internally published the agency’s first fuel economy study.19 They found that emissions controls had reduced fuel economy. But employees also found that emission controls were not the primary factor in fuel efficiency. The primary determinant was (and remains) vehicle weight. Emissions controls came in third place after air conditioning on the list of factors determining fuel efficiency. As Stork later characterized it, the report proved

“extremely useful” for answering “Congressionals” and other public inquiries as well as inquiries from legislators and executive branch agencies. The Office of Mobile

Sources developed a Fact Sheet—“Fuel Economy and Emission Control”—to address the issue.20 Staff members regularly sent out the Fact Sheet to anyone who made an inquiry about this relationship. Thus, the EPA created the first measures of automotive fuel economy not because they were looking to increase efficiency, nor because they particularly cared about it, but for defensive reasons, because automakers and other groups were criticizing emissions controls for increasing fuel

18 Ibid. Marcia Williams went on to have a stellar career at the EPA. By the time she left the agency in 1988, she was the Administrator of the EPA’s solid waste program. She then went to work for Browning-Ferris Industries, the nation’s largest non-hazardous waste management company and the same company that William Ruckelshaus worked for after leaving his second stint as EPA administrator in the 1980s. She now works as a consultant and expert witness, primarily on environmental issues, in Los Angeles. Phone Interview with Marcia Williams, February 16, 2011. 19 Environmental Protection Agency, “Fuel Economy and Emission Control,” November 1972. The report is rather hard to come by as it was only published internally and, therefore, was not placed in any federal depositories. The writer has a copy from the personal collection of Karl Hellman. 20 Environmental Protection Agency, “Fact Sheet 13: Automobile Fuel Economy,” NVFEL Library. There are two versions of this Fact Sheet in the library; one is an updated version of the other. Both of the Fact Sheets on Fuel Economy are undated, but they were written at least after 1978 as they make reference to changes in the EPA’s fuel economy program that occurred in that year.

334 use. The timing of this development, however, was very propitious given the coming of the OPEC Oil Embargo and the first energy “crisis.”

The Office of Mobile Sources also carried out other studies as a form of “self- defense.” Public dissatisfaction with automotive emissions control found its highest, most extreme expression in attempts to remove emissions controls from cars.21 This trend was a fascinating mix of “push” and “pull” factors. Automotive repair garages began advertising emission control removal services in local newspapers around the nation, while consumers both sought out such services and attempted to disconnect the controls on their own. While much of this behavior on the part of consumers could be attributed to simple self-interest—the desire to lower personal gasoline costs—the removal of automotive emissions controls was also an early sign of resistance to federal regulations—a resistance that future presidential candidate

Ronald Reagan would begin to feed upon in his weekly radio addresses in the mid-

1970s and would culminate in the tax revolt and deregulation movements of the late-1970s and early-1980s.

21 The historian Kevin Borg has written about the phenomenon of mechanics removing emissions controls in his book, Auto Mechanics: Technology and Expertise in Twentieth-Century America (Baltimore: The Johns Hopkins University Press, 2007). This phenomenon connects deeply with historians’ and economists’ examinations of how “users” shape technology. On users see, Ronald R. Kline and Trevor Pinch, “Users as Agents of Technological Change: The Social Construction of the Automobile in the Rural United States,” Technology and Culture, Vol. 37 (October 1996), 763–95; Nelly Oudshoorn and Trevor Pinch, eds., How Users Matter: The Co-Construction of Users and Technology (Cambridge, Mass.: The MIT Press, 2003); Ronald R. Kline, Consumers in the Country: Technology and Social Change in Rural America (Baltimore: The Johns Hopkins University Press, 2000); JoAnne Yates, Structuring the Information Age: Life Insurance and Technology in the Twentieth Century (Baltimore: The Johns Hopkins University Press, 2005); idem., “How Business Enterprises Use Technology: Extending the Demand-Side Turn,” Enterprise and Society, Vol. 7, No. 3, 422–455; Eric von Hippel, Democratizing Innovation (Cambridge, Mass.: The MIT Press, 2005).

335

The Office of Mobile Sources reacted to this problem creatively. Building on

Austin’s and Hellman’s earlier fuel economy study and other ongoing fuel efficiency studies, which I describe shortly, the Office of Mobile Sources decided to look into how tampering with emission controls affected fuel economy. To do this, it tested cars before and after taking them to automotive garages where mechanics removed the emission controls.

While the Office of Mobile Sources hoped that its study would reduce the amount of drivers taking their cars to mechanics for the purpose of removing their emissions controls, the actual direct motivation for the study was much more direct:

Less than thirty days after the beginning of the OPEC oil embargo, Ed Cole, the

President of General Motors, told the media that, as far as cars were concerned, the best way to respond to the “energy crisis” was to remove emissions controls. He even proposed getting the Senate to undo—or at least relax—the Clean Air Act.

Firing a shot over the Office of Mobile Sources’ bow, Cole told an Associated Press reporter, “We’ve got some people in Washington who are acquainting people on the

Public Works Committee with this proposition. It’s being put into perspective.”22 He claimed that the country could save “5 billion gallons of gas a year” if it removed emission controls from “40 million late-model cars” that had been sold since emission controls came into effect in 1968. Stork fired back in the same newspaper article, “All they have to do, to save fuel and have clean air, is stop building those

22 “Junk Anti-Smog Devices, Save Gas, G.M. Head Urges,” The Seattle Times, November 15, 1973. In his book, Auto Mania, the historian Tom McCarthy titled his chapter on Ed Cole and the catalytic converter, “The One Who Got It,” as in, out of all the auto executives, Ed Cole was the one who got it. Yet, considering that Cole immediately proposed removing emissions controls in the face of the OPEC oil embargo, it is simply very hard to believe that Cole “got” anything at all. McCarthy, Auto Mania: Cars, Consumers, and the Environment (New Haven: Yale University Press, 2007), chapter 9.

336 huge behemoths on which they make the most profit, and build small cars.” He went on, “Emission controls have adversely affected only the fuel consumption of heavy cars, of the type Mr. Cole builds.” Stork concluded that, “The auto industry is just using the opportunity of the energy shortage as yet another attack on the Clean Air

Act.”23

While Stork publicly battled Cole, he asked his staff members in Ann Arbor to begin the study on the fuel economy gains from removing emissions controls. That the Office of Mobile Sources could so quickly switch to a new project is a testament to the organizations’ flexibility during its early days. Later, when asked how the agency fit the fuel economy studies into its already chocked-full schedule, Karl

Hellman recounted, “Back in those days, we just did things.”24 The scientists, engineers, and technicians at the EPA’s Ann Arbor lab simply responded to the agency’s needs.

The study involved ten cars that broadly represented the “full range of typical vehicle weights encountered in the existing vehicle population.”25 (Recall that the Office of Mobile Sources had already found that vehicle weight was the most important factor in determining fuel economy.) Staff members then contacted a

“number of garages,” asking that “they do whatever they could do” to improve fuel economy.26 Surprisingly, perhaps, only a quarter of the garages contacted agreed to tamper with the emission controls. The EPA was not sure whether garages declined

23 All quotations are from, “Junk Anti-Smog Devices, Save Gas, G.M. Head Urges.” 24 Interview with Karl Hellman, August 19, 2008, NFVEL. 25 Environmental Protection Agency, “A Study of Fuel Economy Changes Resulting from Tampering with Emission Control Devices,” January 1974, NVFEL Library, 1. 26 Ibid.

337

“either for the reason that they thought such work was illegal or because they did not want to contribute to deterioration of air quality,” but eventually the staff members found the requisite ten garages needed for the study. The staff members ensured, in good social science fashion, that the garages “represented a cross- section of the automobile service industry, and included corner gas stations, commercial tune-up centers, as well as a garage that is widely advertised in the

Detroit metropolitan area as a specialist in the removal of emission control devices.”27 Some of the garage proprietors claimed boldly that they “could improve both performance and fuel economy with little impact on emissions,” while others asked the “customers” to get back to them on how they thought the performance had changed. The staff members did not tell any of the garages that they worked for

EPA, though two of the garages eventually figured out.28

The staff members then set to work tuning up the cars to the manufacturers’ specifications. They did not want the automakers claiming that any deterioration in vehicle performance, fuel efficiency, or emissions control was due to using cars that had not been properly maintained. They did not, however, tune three of the ten cars.

As their report noted, “The purpose of this deviation from the basic test approach was two fold: to ascertain whether the EPA tune-up was inadvertently assisting or biasing the garage tampering attempts and to obtain data on a real-world basis, i.e., tampering with an ‘as-received’ vehicle would represent the normal challenge to the garage.”29 In each case, the staff members tested the cars using the federal test

27 Ibid., 5. 28 Ibid., 3. 29 Ibid., 2.

338 procedures before and after taking them to the garages. The EPA staff members would then “restore [the] vehicle to [the] manufacturer’s specifications” before taking their own crack at improving a car’s fuel efficiency by tuning it to the

“manufacturer’s specifications.” The car would then be tested again. The EPA staff members got much better fuel economy gains than the garages. As the report noted,

“The EPA modification were made by engineers and technicians who have detailed knowledge of automotive emission control systems, as well as access to sophisticated test equipment. The level of skill and facilities is not ordinarily found in the average repair garage.”30 Stork would use this difference in expertise in his rhetoric against Cole.

In January of 1974, two months after Cole’s comments, the Office of Mobile

Sources finished its report, “A Study of Fuel Economy Changes Resulting from

Tampering with Emission Control Devices.” The report was first published only internally. It showed that the EPA modifications could improve fuel economy by about 9% on average by “tuning up the ‘as received’ rental vehicles to manufacturers’ specifications.”31 But the majority of auto garages actually reduced the cars’ fuel economy by 3.5% by tampering with emissions controls. The report demonstrated, therefore, that having mechanics remove emission controls, as Cole had proposed, would decrease fuel efficiency while greatly increasing automotive emissions, resulting in emissions above the nitrogen oxide standard and nearly three times greater than the standards for hydrocarbons and carbon monoxide.32

30 Ibid., 3. 31 Ibid., 4. 32 Ibid.

339

The best way to increase fuel economy, therefore, was to keep the car tuned up. The report also remarked that tampering would probably hasten engine deterioration.

Stork had pointed out one reason for this deterioration in his earlier jousting with

Cole: many of the changes that automakers had made to engines to decrease emissions controls were so basic to the engine’s design that they “cannot readily be varied independently.”33 (Remember this episode unfolded before the advent of the catalytic converter; previous to catalytic converters, emissions were reduced through “engine modifications.”) Stork gave several examples, “Carburetor air fuel ratio, ignition timing, compression ratio, and exhaust air recirculation all affect engine durability. Changes in these factors . . . can result in mechanical durability problems (such as valve and piston failures) or performance problems (automatic transmission shifting pattern, cold starting, and reduced fuel economy).”34

Furthermore, as the EPA staff members wrote in the report,

“Of significant note, the garage that advertised its expertise in emission control removal modified two cars for EPA and in both cases was unsuccessful at improving fuel economy. Of even greater interest may be the fact that this organization knew that it was participating in the EPA program. It was necessary to inform them of the EPA program because their workload schedule was such that to get an anonymous test would have required a 45-day delay. The failure of this organization to achieve improved fuel economy may be attributable to the fact that it is oriented toward improved performance (for example, acceleration and top speed). This would indicate that application of a hot rodder’s knowledge does not insure achievement of improved fuel economy.”35

This last point was the theme that Stork took up as he talked to journalists.

33 Robert W. Irvin, “U.S. Urges Keeping Anti-Smog Systems: You Won’t Save on Gas by Tinkering, Drivers Told,” The Detroit News, January 4, 1974. 34 Ibid. 35 EPA, “A Study of Fuel Economy Changes Resulting from Tampering with Emission Control Devices,” 5.

340

“The application of the hot-rodder’s skills fuel economy just doesn’t work,”

Stork told Detroit Free Press writer Douglas Williams. “It’s preposterous,” Stork went on, “to think that you can redesign 40 or 50 million cars in the garages of America and get any fuel economy gain.”36 In an earlier article, Stork had explained why he had developed the tampering study, which he described as a “little vest pocket program,” to defend his office. “We’re not exactly disappointed by the results” of the study, Stork said.37 In the end, the tampering study demonstrated how Stork and his staff members were willing to use the agency’s capacities in flexible ways, to generate useful knowledge that was helpful to the agency, and disseminate that knowledge to the media in order to fight off critics.

Building the EPA’s Fuel Efficiency Statistics and Publishing the First Fuel

Economy Numbers

On April 18, 1973, President Richard Nixon delivered a special energy message to Congress. It was his second such message. He had given an earlier one in

1971, but many critics claimed that Nixon had done little since that first communication. Nixon, however, never one to be cowed, pointed out in the 1973 message that federal energy R&D had increased by 50% since the first message.38

36 Both quotations, Douglas Williams, “Foiling Smog Gear Can Hurt Mileage, EPA Test Confirms,” Detroit Free Press, January 31, 1974. Stork’s quotations also speak to his faith in expertise at a time when, as many scholars have argued, many people in the United States were losing faith in expertise. On the decline of faith in experts, see Brian Balogh, Chain Reaction: Expert Debate and Public Participation in American Commercial Nuclear Power, 1945–1975 (Cambridge: Cambridge University Press, 1991), 17. 37 Douglas Williams, “Unhook Smog Gear? EPA Tests the Idea,” Detroit Free Press, January 18, 1974. 38 Nixon, “Special Message to the Congress on Energy Policy,” April 18, 1973. Nixon’s message is available online, as part of The American Presidency Project, at

341

As some magazines and newspapers noted, Nixon gave the second speech much later than promised.39 The advent of the “energy crisis”—really, if anything, a series of interlocking “crises,” more than a unified event—was a long time coming. Best remembered is the OPEC Oil Embargo, which set in overnight on October 20, 1973, when the organization announced an oil embargo against the United States (and soon after, other industrialized nations) for its support of Israel in the Yom Kippur

War.40 But the United States had already faced shortages of natural gas during the winter of 1972–1973, and gasoline prices were already high in the spring and summer of 1973. The country had become a net importer of petroleum—that is, its imports outweighed its exports—in 1970. The population’s use of energy, an important pillar of the “American way of life,” was outpacing both domestic production and reliable foreign exports.41 Moreover, global economic shifts also influenced the issue. In the early 1970s, the Bretton Woods monetary system came to an end after creating relative economic stability around the globe after World

War II. In August of 1971, the United States’ part in the system effectively ended when Nixon took the United States off the gold standard and “floated” the US dollar, http://www.presidency.ucsb.edu/ws/index.php?pid=3817&st=&st1= (Last accessed February 15, 2011). 39 See, for instance, “Environment: At Last, The Energy Message,” Time, April 30, 1973. 40 For a clear and broad survey of the “energy crises” of the 1970s, see Daniel Yergin, The Prize: The Epic Quest for Oil, Money, and Power (New York: Simon & Schuster, 1991), especially chapters 31–33. Another helpful volume that focuses primarily on the late-1970s but offers perspectives on the entire period is Karen R. Merrill, The Oil Crisis of 1973–1974: A Brief History of Documents (Boston: Bedford/St. Martin’s, 2007); Daniel Horowitz, Jimmy Carter and the Energy Crisis of the 1970s: The “Crisis of Confidence” Speech of July 15, 1979: A Brief History with Documents (Boston: Bedford/St. Martin’s, 2005). 41 The maximal statement of how cheap petroleum supported the “American way of life” and the “free world” is NSC 138/1, written in January 6, 1953. The National Security Council document claimed, in effect, that the Justice Department should stop seeking to bring antitrust charges against American-owned international oil companies because the country was heavily dependent on petroleum and, thus, such federal charges were a threat to national security. See, Foreign Relations of the United States, 1952–1954, Vol. 1, Pt. 2., 1317–1329.

342 greatly devaluing the currency. This act angered the members of OPEC, who un-

“pegged” the value of oil from the US dollar and re-fixed it to the gold standard, instantly increasing oil prices, especially in the United States, with its devalued dollar. In this context, people were increasingly talking about possible and real energy shortages and the need for a comprehensive energy policy. Nixon’s speech addressed all of these factors, but in the short term, his hand may have been forced when James E. Akins, a Foreign Service officer who had come to work for the White

House, published an article in Foreign Affairs titled “The Oil Crisis: This Time the

Wolf is Here.”42

Nothing in Nixon’s message was revolutionary. Writing in the July issue of

The New York Review of Books, the historian Emma Rothschild described Nixon’s speech in this way: “The presidential message achieved a tone that mixed the calm and shrill, major and minor, multinational and nationalistic, Eastern and Texan voices of different energy interests.”43 As she remarked, Nixon’s “major projects

42 James E. Akins, “The Oil Crisis: This Time the Wolf is Here,” Foreign Affairs, April 1973. For a discussion of the Akins’ role in shaping US energy policy, see Yergins, The Prize, 590–592. 43 Emma Rothschild, “What is the ‘Energy Crisis’?” The New York Review of Books, July 1973. Akin’s and Rothchild’s titles should remind us that energy issues had already been framed as a “crisis” months before the deepest troubles hit. Although the days following the OPEC Oil Embargo—which included fuel rationing, gas shortages, and the famous lines at gas stations—were very troubling and scary indeed, the idea that a “crisis” was at hand was mostly a social and rhetorical construction. The philosopher Jurgen Habermas notes that the term “crisis” originally arose from medicine, where it was used to refer to the “phase of an illness in which it is decided whether or not the organism’s self- healing powers are sufficient for recovery.” Habermas, Legitmation Crisis, trans. Thomas McCarthy (Boston: Beacon Press, 1975), 1. “Crisis,” therefore, is a biological metaphor for society, and like any theory that views society as an “organism,” it is suspect. After losing the Presidency to John F. Kennedy in 1960 and returning to legal practice in California, Richard Nixon wrote a political memoir titled Six Crises (New York: Simon & Schuster, 1962), so the later president may have been predisposed to framing problems in such terms. Nixon was also a self-described “Keynesian,” and to the degree that the notion of “crisis” is integral to John Maynard Keynes’ economic theories, Nixon have picked up the idea from there. For Keynes’ reliance on the notion of “crisis,” see, for instance, John Maynard Keynes, The General Theory of Employment, Interest, and Money (New Dehli: Atlantic Publishers & Distributors Ltd., 2006 [1936]), 17, 188, 286, 322.

343 followed most of the recommendations of the American Petroleum Institute, the

National Petroleum Council, the American Gas Association, and the other institutes and associations that had described America’s energy crisis.” In other words, Nixon was mostly towing the party line set by industry. But this could not be said of

Nixon’s relationship with the auto industry; his early policies remained a burr under

Detroit’s saddle.

One great irony of history is that Jimmy Carter is remembered for the doom and gloom of his “crisis of confidence” speech, which advocated energy conservation as a “moral equivalent of war” and in which Carter famously wore a sweater. Yet,

Nixon made similar claims that are not remembered. By June of 1973, still four months before the Oil Embargo, Nixon was asking Americans to conserve energy by driving smaller cars and, as the Edward Cowan of the New York Times put it, by

“keeping their homes warmer in summer and cooler in winter.”44 Nixon asked that businesses “reduce air-conditioning and encourage appropriate relaxation of dress standards in offices this summer.” Even in the earlier April message, Nixon had emphasized energy conservation. His policies on automotive fuel economy fit under this larger umbrella.

In the days before Nixon was to give the energy speech, his staff members were feeling around for potential policies to include in the address. Since the federal government had no centralized energy agency (the Federal Energy Office, which was another recommendation made in the speech, would not be created for several months), staff members polled a number of agencies for recommendations.

44 Edward Cowan, “President Asks the Nation to Drive Smaller Cars,” New York Times, June 30, 1973.

344

The EPA was one such agency, and Stork’s office was contacted. Stork recalls that he was on the way out the door to a meeting when Janet Lane, his “right hand man on policy,” brought up the presidential inquiry. Stork was extremely busy dealing with the suspension hearings addressed in chapter 5, and he didn’t think the Office of

Mobile Sources had anything to give the White House. But Lane suggested that the

EPA could start publishing its fuel economy numbers as a consumerist measure.45

Stork liked this idea and told her to send it up the pole. In his energy address, Nixon then “directed” the EPA (to do something that it had itself suggested) to begin publishing fuel economy data. As Nixon said,

“To provide consumers with further information, I am directing the Department of Commerce, working with the Council on Environmental Quality and the Environmental Protection Agency, to develop a voluntary system of energy efficiency labels for major home appliances. These labels should provide data on energy use as well as a rating comparing the product’s efficiency to similar products. In addition, the Environmental Protection Agency will soon release the result of its tests of fuel efficiency in automobiles.”46

Although the point was not explicit in Nixon’s speech, automobiles were also to be included in the labeling program.

The EPA published the first fuel economy statistics without ceremony or ornamentation in the Federal Register two weeks after Nixon’s energy message, on

May 2, 1973.47 It instantly created hubbub in the press and furor among the automakers. It was the first publication to compare the fuel economy of all makes and models of cars in a comprehensive manner. The Office of Mobile Sources always

45 Phone interview with Eric Stork, January 5, 2010. 46 Nixon, “Special Message to the Congress on Energy Policy.” 47 Federal Register, Vol. 38 (May 2, 1973), 10867–10908.

345 struggled to communicate the real “meaning” of fuel economy statistics (and continues to do so to this day).48 The first Federal Register notice itself warned,

“EPA must caution against attempting to compare these published fuel economy values with other values obtained under different conditions or by different techniques. Fuel economy is affected by a wide range of factors including the manner in which the vehicle is driven, type of route and terrain traveled, speeds at which the vehicle is driven, frequency of cold-starts, use of power-absorbing accessories [for instance, air-conditioning], vehicle weight, axle ratio, ambient conditions, and many others.”49

The agency made real efforts to communicate the complexity of the issue. Since the data were generated through the federal test procedure (FTP), they were best seen as comparative in nature: this car performs at this level in comparison to that car.

The FTP was an approximation of average driving patterns; thus, it could not generate an estimate of how many miles per gallon (MPG) a car would “get” for any one driver. Individual driving styles would create huge variations in a car’s MPG.

Someone who accelerated hard and slammed on the brakes at each stop would get much worse fuel economy than someone who accelerated slowly and took his or her foot off the gas well before coming to rest.50 Stork envisioned expressing fuel

48 Explaining the “meaning” of numbers, i.e., which inferences can be drawn legitimately from a body of data and which cannot, is a constant chore for statistical agencies. For instance, the Energy Information Administration (EIA) has often tried to explain the intentions of its statistical forecasts of energy prices and production and consumption levels, which are meant to be used as tools for policy-making and not as “predictions” of the future. The Bureau of Labor Statistics has had similar struggles communicating with the public about its products, including unemployment statistics. On the EIA’s efforts in public dissemination, see Lee Jared Vinsel, “’Every Week We Find a New Devil’: The Crusade for Credible Energy Information and Analysis, 1973-1981,” Forthcoming. 49 Federal Register, Vol. 38 (May 2, 1973), 10868. The term “cold-start” refers to running a car when the engine has not had time to “warm up.” The car produced many more emissions and had less fuel efficiency during these periods. For this reason, the Federal Test Procedure required periods known as “cold soaks,” in which a car would be left to sit overnight to ensure that the test held the next day would capture this cold-start phase. 50 Recently, automakers have built some cars, such as the Ford Focus, with graphic representations of fuel efficiency on the dashboard. In the Focus, this representation takes the form of five bars that increase from size from left to right, the left representing low fuel use, the right, high fuel use. Thus,

346 efficiency statistics like a batting average, as so many points out of a thousand.51

This way people would not get hung up on MPG and become upset when their cars failed to live up to that expectation. But such an alternative vision would never come to fruition. The Office of Mobile Sources hired a firm in Texas to perform

“focus groups” on fuel economy statistics. The upshot of the groups was clear: people wanted to see MPG.52 This issue would continue to plague the agency, as citizens and interest groups complained about, and the media picked up on, the lack of “accuracy” in the federal fuel economy statistics.

After publishing the fuel economy data, the Office of Mobile Sources began to work on the “voluntary” automobile-labeling program. “It was voluntary as in you will,” Stork later recalled.53 All of the automakers agreed to the program, except

American Motors Corporation. As Nixon had directed, the EPA was supposed to work with the Department of Commerce and Council on Environmental Quality, but little came from these interactions, and the EPA ended up developing the automotive labeling program on its own.54 On August 27, 1973, the EPA announced

automakers have attempted to provide drivers with a feedback mechanism by which to gauge their driving styles. The electronics involved in this process lay far outside of industry practice in the 1970s, however. NHTSA, whose role in fuel economy will become evident later in the chapter, has tried at times to educate the public about driving styles. During the period of high gasoline prices in 2008, NHTSA representatives appeared on television and used a heuristic to explain the “ideal” driving style: this ideal took the form of imagining that a cup of coffee was resting on the dashboard. The most efficient driving pattern would execute both accelerations and decelerations without spilling any coffee—clearly far from typical driving behavior. 51 Phone interview with Eric Stork, January 5, 2010. 52 Interview with Janet Auerbach, at her home, April 18, 2010. Phone interview with Eric Stork, January 5, 2010. 53 Phone interview with Eric Stork, January 5, 2010. 54 Eric Stork, “The Genesis of the EPA Fuel Economy Program.” In a Senate hearing I discuss below, it is clear from the testimony of members of the Office of Management and Budget (Roy Ash, Director; Frank Zarb, Associate Director, Energy, Science, and Natural Resources; and James Tozzi, Chief, Environmental Branch), all of whom wanted to centralize energy efficiency labeling under the

347 the details of the voluntary labeling program, including a sample of the standard format of the label that would be posted on all new cars sold in the USA.55

The EPA handled its second publication of fuel economy data completely differently. Instead of simply publishing the results in the Federal Register, Russell

Train, the new EPA administrator, unveiled the data on the 1974 model year vehicles at the National Press Club on September 18, 1973. The room was packed with journalists, and many major news media carried the story the next day.56 It was the biggest positive news the agency had ever generated, and the publication curried great favor for the Office of Mobile Sources with the powers within the EPA, including with Russell Train, who basked in the spotlight at the Press Club only four days into his new job.

The automakers were not happy, but no company was as furious as Mazda, whose Wankel engine vehicles scored badly in the EPA tests. Controversy between the automaker and the agency did not erupt until January of 1974, however.57 The research firm J. D. Power and Associates sent a news release to several west coast media establishments pointing out that there was a large discrepancy between

Mazda drivers’ perceived fuel economy—they thought they were getting between

Department of Commerce’s National Bureau of Standards (NBS), that the NBS had no hand in developing the EPA’s fuel economy labeling and information publication program. 55 Federal Register, Vol. 38, No. 165 (August 27, 1973), 22944–22947. Fitting with the EPA’s sole control of the automotive labeling program, EPA issued the announcement solely under its own name; the contact information was an address for the agency; and the announcement declared, “The Environmental Protection Agency is particularly interested in receiving comments on the effectiveness of the recommended vehicle label.” The notice also highlighted the centrality of the agency’s fuel economy data on the labels. 56 See, for instance, “Gasoline Guzzlers Unmasked by EPA Tests,” Chicago Tribune, September 19, 1973 and “Honda Civic Tops Lists in ’74 Auto Mileage Test,” New York Times, September 19, 1973. 57 Douglas Williams, “The Mazda Fuel Furor,” The Review of Southern California Journalism (May 1974), issue 11. The review was a small, student-run publication out of the California State University, Long Beach that ran from 1971–1975.

348

17 and 20 MPG—and the EPA’s numbers, which placed the Wankel—or rotary engine—cars at between 10.7 and 10.8 MPG. (Beyond “guzzling” gasoline, rotary engines were also famous for consuming large amounts of motor oil.)

Unfortunately, someone at UPI rewrote this news release, framing the story as

Mazda attacking the validity of the EPA’s fuel economy program, and the story ran on Monday, January 7, 1973.58 That same day, someone from UPI called Stork and asked him about Mazda’s ”attack.” Stork shot back, claiming that, as one UPI article paraphrased him, Mazda’s “rotary engine is an inefficient gasoline user and the

EPA’s tests on the Mazda automobile prove it.”59 C. R. “Dick” Brown, the general manager of Mazda Motors of America, returned fire. He argued that the EPA test was biased against Mazda because the test was based on an urban driving cycle while Mazda’s cars performed better on the highway. By February, he was also stating that the negative press coverage in January had already cost the company

$100 million in sales.60

In the end, the EPA developed the highway test cycle, which became a standard part of fuel economy labeling, to answer Mazda’s criticisms. In the Office of

Mobile Sources’ Ann Arbor building, a long hallway runs between the bulk of the office space and the physical lab space. Stork recalls that his staff members would go drive highway test routes around Michigan with a fifth wheel attached to the test car to measure accelerations, decelerations, and periods of stasis.61 They then lined

58 Ibid. 59 Ibid. 60 Douglas Williams, “’Gas Hog’ Tag Cost $100 Million—Mazda,” Detroit Free Press, February 25, 1974. 61 Phone Interview with Eric Stork, February 17, 2011.

349 the hallways with the test ribbons in order to examine the characteristic features of highway driving, before building a dynamometer-based driving schedule that

“represented” this average highway driving. EPA staff members then ran this test along with the federal test procedure to create the “highway” and “city” fuel economy numbers.

Yet, even after running the combined urban and highway test cycles, Mazda’s rotary engine-equipped cars fared very poorly, getting about the same fuel economy as the Ford Torino, which weighed nearly 2,000 pounds more than Mazda’s cars.62

As Douglas Williams pointed out, embarrassingly for Mazda, these new tests emboldened the EPA to create a “fact sheet” on the Mazda rotary engine, which the agency then sent out as an official statement to any interested party, including inquiring journalists.63 For all of Mazda’s protests, it was the one wearing egg in the end.

The EPA made other changes to its testing and labeling program for reasons other than automakers’ complaints. One change, however, arose from pressure from journalists and the public. Stork recalls that the Associated Press reporter Stanley

Benjamin called him, complaining that the EPA’s earlier fuel economy publications were too complex.64 It was too hard to figure out who the winners and losers were, and that was the story that mattered to the journalists—the narrative of the gas- guzzlers. The EPA made other changes as well. For instance, in the September 1973 fuel economy report, the vehicles had been divided into weight classes, and the

62 Miranda Lorraine, “EPA Issues New Mazda Data: Rotary Engine Still Rates Low,” The Washington Post, April 12, 1974. 63 Douglas Williams, “The Mazda Fuel Furor.” 64 Phone interview with Eric Stork, January 5, 2010.

350 consumer survey that EPA contracted found that these classes held no meaning for consumers.65 Few people knew how much their car weighed. So, the EPA broke the cars down into broader, easier to understand categories.

By 1974, the EPA also wanted to broaden the reach and quality of its publications, but Stork had no resources to spend on publishing. So, he got in touch with his contact at the new Federal Energy Agency (FEA), Deputy Assistant

Administrator Robert Hemphill, and made him an offer.66 If Hemphill would foot the bill for publication costs, Stork proposed, the fuel economy statistics would be released as a joint publication between the two agencies. The FEA would get a piece of the publicity around automotive fuel economy without needing to do any work.

Hemphill gladly took the bait. On September 20, 1974—once again at the National

Press Club—Russell Train, the Administrator of the EPA, and John Sawhill, the

Administrator of the FEA, announced the 1975 model year fuel economy data to great fanfare

To this day, the FEA’s successor agency, the Department of Energy, publishes the federal fuel economy statistics. The FEA’s buy-in to this program was also the first moment where federal responsibility for fuel economy—in this case, the responsibility for “publishing” the data—became divided. (As will become evident later in this chapter, the federal fuel economy program is a many-headed

65 Douglas Williams, “EPA Admits Blunder on Mileage Figures,’ Detroit Free Press, May 30, 1974; Agis Salpukas, “E.P.A. to Change Auto Fuel Data,” New York Times, May 30, 1974. 66 In response to the OPEC Oil Embarge, President Richard Nixon proposed reorganizing the bulk of federal energy functions into one central authority. He created this organization, the Federal Energy Office, through an Executive Order in December 1973. When Congress authorized the organization in May 1974, it became known as the Federal Energy Agency (FEA). The FEA handled many federal energy programs during the “ crises” of the 1970s, until President Jimmy Carter created the Department of Energy in 1977.

351 hydra, with several different agencies handling different aspects.) But by bringing the FEA in on the EPA’s fuel economy program, Stork had successfully won over the energy agency to his cause.

Fighting to Retain Fuel Economy

As Eric Stork has aptly described, federal agencies are characteristically marked by a form of “Soonerism.” The term arises from the Sooner State, Oklahoma, where people headed into federal lands before the government officially offered it up for settlement. In Washington, D.C., Soonerism takes the form of turf battles over bureaucratic jurisdiction. When a topic is “hot,” when something suddenly becomes a “problem,” every agency wants a piece of the action.67 The OPEC Oil Embargo made fuel economy and other energy issues very hot indeed. Energy prices spiked, and the federal government scrambled for any and all means for providing the country with relief. Energy issues, including fuel economy, suddenly became desirable commodities for federal agencies. Seemingly every office, agency, and bureau had its iron in the fire of energy policy. In such an environment, several organizations made a play for the EPA’s fuel economy program.

In his seminar paper, “The Genesis of the EPA Fuel Economy Program: An

Exercise in Bureaucratic Gamesmanship,” under the heading, “Designed strategy to hang on to fuel economy program,” Stork wrote, “Coopted FEA and FTC. How.

67 As mentioned in the Introduction, social “problems” are not “natural,” or are at least not wholly so. The classic work on how social problems arise or are constructed is James R. Gusfeld, The Culture of Public Problems: Drinking-Driving and the Symbolic Order (Chicago: University of Chicago Press, 1981).

352

Coopted SAE Committee. How.”68 I have already described how Stork coopted the

FEA: he offered the agency a piece of the fuel economy pie without asking them to do any work; they only had to front the cost of publication. In other cases, if one agency challenged the Office of Mobile Sources, Stork would recruit others, until he would eventually win over former challengers. He instinctively knew something that the sociologist of science Bruno Latour, who envisions sociotechnical systems as a network of interconnected actors and things, has claimed: such systems become stronger when they gain associations with and the loyalties of powerful actors.

Yet, the EPA also found a powerful adversary in the auto industry. The agency’s fuel economy publications had made the automakers unhappy. Indeed, they were furious. The Society of Automotive Engineers (SAE) was very displeased that it had been left out of the standards-making process; in this case, the EPA had introduced what was quickly becoming the standard fuel economy testing procedure by fiat when its data had been picked up by the Nixon administration.

The SAE’s biggest complaint was that the EPA was using a dynamometer test, which was at the center of the EPA’s federal test procedure for emissions control, instead of an on-road test, which the society preferred.69 Stork largely dispelled the SAE’s concerns and co-opted the organization by asking it to help the EPA revise the agency’s fuel economy test and flattering the organization in the press.70

Other influential members of the auto industry, most visibly and vocally Lee

Iacocca, applied pressure to the executive branch to get the measurement of fuel

68 Stork,“The Genesis of the EPA Fuel Economy Program.” 69 Tom Kleene, “SAE to Design EPA Mileage Test: Engineers Face June Deadline,” Detroit Free Press, March 8, 1974. 70 Eric Stork, “The Genesis of the EPA Fuel Economy Program.”

353 economy out of the EPA’s hands and into the hands of an agency more “friendly” towards industry. Thus, pushed and pulled from all sides, Eric Stork and the staff members of the Office of Mobile Sources had to fight to maintain their grip on the fuel economy program.

Iacocca was close to Roy Ash, the director of the Office of Management and

Budget under the Nixon and Ford Administrations. He visited Ash’s office in the first week of March and told Ash that he would like to see the publishing of fuel economy data moved out of the EPA into the Bureau of Standards, which was housed at the

Commerce Department.71 In April 1973, Ash told Betsy Ancker-Johnson, a scientist who was the Department of Commerce’s Assistant Secretary for Science and

Technology at the time, to take control of the fuel efficiency program and move it to the Bureau of Standards, an agency within Commerce. Employees of Ancker-

Johnson’s office began calling Stork, but he refused to take their calls. Then, they began calling the Office of Mobile Sources’ Ann Arbor office. Stork told his staff members that they should follow two rules when dealing with Ancker-Johnson’s office: first, all calls should be forwarded to him. Second, he would never take the calls. Stork later attributed Commerce’s inability to overcome this stonewall as a difference in organizational culture. “They were just too polite,” he said.72 If things had been reversed, if the EPA had been going after a program in Commerce, EPA staff members would have fought much harder than staff members at Commerce did. In the text for his 1978 seminar, Stork remarked, “if we’d had such [a] raiding

71 Rowland Evans and Robert Novak, “Fuel Economy Testing: Undercutting the EPA,” The Washington Post, May 4, 1974. 72 Phone interview with Eric Stork, January 5, 2010.

354 charter in April,” as Ash had given Commerce, “we’d have had [the] program by

July—maybe some pillage, mayhem, and plunder—but [we] would have had it.”73

Stork first learned about Ash’s promise to Iacocca to move fuel economy to the NBS from a contact at Ford, and a contact at OMB subsequently affirmed it.74 He saw it as an attempt to make the fuel efficiency information “of the industry, by the industry, for the industry.”75 Stork began pushing journalists to question Ash about the matter. On April 25, the Los Angeles Times wrote a story about the issue.76 Ash said that issue was under consideration. Stork claimed that having any other agency do the test would be inefficient because the EPA could base the data on tests it conducted anyway. The next day, the Detroit Free Press published an article that said more or less the same thing. Ash claimed that the decision regarding fuel economy would be made “in a few days.”77 Yet, the article also paraphrased and quoted a “high level civil servant close to the White House [who] said . . .the decision to move fuel economy testing to Commerce had already been made at OMB. He termed it ‘a dumb idea done for purely political reasons and one that had been hard fought within the administration.’” On that same day, April 26, Stork quickly wrote a one-page memo to EPA Administrator Russell Train, in which he outlined his vision of how the OMB was planning on taking the fuel economy program out of the EPA’s hands.78 As was typical of important, sensitive memos of the day, Stork only created

73 Eric Stork, “The Genesis of the EPA Fuel Economy Program.” 74 Eric Stork, “The Genesis of the EPA Fuel Economy Program.” 75 Phone interview with Eric Stork, January 5, 2010. 76 Dan Fisher and Paul E. Steiger, “Gas Mileage Tests by U.S. May Be Changed,” Los Angeles Times, April 25, 1974. 77 Douglas Williams, “Mileage Test Jurisdiction Up for Grabs,” Detroit Free Press, April 26, 1974. 78 Evans and Novak, “Fuel Economy Testing.”

355 three copies—one for Train, one for Train’s chief of staff, and one for his own personal files.

On Sunday, May 4, Stork rose from his bed and went downstairs. During breakfast, his wife pushed the newspaper over to him. There, in the Washington

Post, in the Rowland Evans and Robert Novak Sunday column, Stork saw his memo liberally quoted. He was dumbstruck. He had no idea how the columnists had gotten his memo, and he didn’t like it. It put him in a delicate position.

Even the conservative Evans and Novak explicitly framed Ash’s move as a pro-industry, anti-EPA policy. “If you took a poll in Detroit, no agency in

Washington would be more unpopular than EPA,” one “top-level Ash aide” told the columnists.79 Evans and Novak argued, “Behind the White House effort to strip testing from the consumer-oriented EPA is a more dangerous game which looks suspiciously like a clandestine administration effort to gut its own auto-pollution controls program.” The columnists noted that, following the Nixon administration’s

“well-established rule of cottoning big business and ignoring the consumer,” Ash was thinking of moving fuel economy either to the Commerce Department or DOT,

“both safely industry-oriented.” They then quoted Stork’s memo, dated April 26:

“What we see as really involved in the proposed move of fuel economy testing out of

EPA is an effort to emasculate EPA in any future fight over auto emission standards

[in which] . . . all real influence depends on the ability to obtain, assimilate, and use information.”80

79 Evans and Novak, “Fuel Economy Testing.” 80 Ibid.

356

Heading into work on Monday, Stork was nervous, on edge. Not long after he arrived at the office, Stork’s secretary called: Russ Train was on his car phone.

Stork had her put Train through and greeted him. Stork recalled that Train said,

“Eric, it’s about the Evans-Novak column.” “Yes, sir,” Stork replied. “I hope it didn’t embarrass you,” Train continued. “You see, Evans is an old friend, and I had him over for dinner. While I stepped out to take a phone call, I must have left your memo out on the coffee table, where Evans must have looked at it.”81 Stork’s memo had been leaked from above. Train was protecting the EPA’s and Stork’s position by tapping his high-level media contacts.

At this point, Ash backed off of moving the fuel economy program to the NBS or some other agency, such as DOT. His behavior was motivated by several factors.

The press gave the issue a fair amount of coverage, particularly after the Evans and

Novak column. The Nixon Administration also was already deeply immersed in the public fallout from the Watergate scandal. In April 1974, the Nixon administration released its first Oval Office transcripts; in May, Democrats opened impeachment hearings against him; and, finally, in July, the House voted in favor of impeachment.82 To put things mildly, things were not going well for the president, and his administration may have been avoiding further controversy. Transferring fuel economy tests from the EPA to NBS would have certainly resulted in outcry

81 Phone interview with Eric Stork, January 5, 2010. On Train’s close relationship with Rowland Evans, whom Train called “Rowly,” see J. Brooks Flippen, Conservative Conservationist: Russell E. Train and the Emergence of American Environmentalist (Baton Rouge: Louisiana State University Press, 2006), especially pg. 177, which mentions that Train regularly had dinner at Evans’ house. 82 In a videotaped oral history with the National Archives, Jim Tozzi, who was the person in charge of the EPA’s budget at OMB, said that, because the OMB was so close to the president, it felt considerable pressure during the period of Watergate. The video is archived at http://www.thecre.com/video/National_Archive.html (Last viewed February 17, 2011).

357 from environmentalists and consumer advocates. For instance, the long-time auto safety activist and critic, Clarence M. Ditlow III, who in 1974 worked for Ralph

Nader’s Public Interest Research Group, strongly supported fuel economy remaining the EPA’s hands.83

On May 17, 1974, members of the OMB and EPA appeared before the Senate

Subcommittee on Science, Technology, and Commerce, which was a part of the

Committee on Commerce. The hearing was on a proposed law, the National Fuel

Economy Testing Act of 1974, that never passed into law, though its intention would be captured in the Energy Policy and Conservation Act one year later. Ash claimed that he was still considering moving fuel economy to the NBS, but that for the time being he would leave it in the EPA’s hands. He was also forced to confess that the

NBS had not always done the best job at communicating energy conservation information when a Senator confronted him with an NBS pamphlet on air conditioners, which they both admitted was totally incomprehensible to the average consumer. Ash and the other OMB staff members insisted that the NBS should play the central role in developing fuel economy standards because it was the key agency with expertise in developing standards. That’s what the agency did. To this kind of talk, Stork and the EPA consistently replied, as Stork put it in his seminar, “because we had the only comprehensive data base, we had the best data base.”84 By the time that Russell Train and Stork testified that day, the majority of the Senators were in the EPA’s corner.

83 US Senate, Special Subcommittee on Science, Technology, and Commerce, Hearings on The National Fuel Economy Testing Act of 1974, May 17, 1974, 98. 84 Eric Stork, “The Genesis of the EPA Fuel Economy Program.”

358

The following day, the Washington Post ran an article with the title “EPA

Wins Its Fight to Keep Car-Mileage Tests in Agency.”85 That wasn’t the message that

Ash wanted to send, but it was the message. A week later, The Automotive News claimed that, “EPA Still Holds the Reins in Mileage Labeling.”86 The issue died down over the summer. In August, the General Accounting Office published a report on the issue of fuel economy programs within the federal government, after receiving a request in April for an investigation into the issue from Congressman Henry S. Reuss

(D—Wisconsin), the Chairman of the House Subcommittee on Conservation and

Natural Resources.87 While the GAO report purposely did not take sides on the issue of which agency should run the fuel economy program, it did show the EPA’s program in a rather positive light. The report highlighted how the EPA was using preexisting capabilities (the federal emissions tests) to create the fuel economy data, how the Department of Commerce would need to develop new and expensive capabilities if it wanted such a program, and how the EPA had been working hard over the past few years to improve its work. On the one hand, between information processing costs and the costs of publishing pamphlets, the EPA had spent $26,000 on fuel economy testing and labeling for the 1974 model year.88 Even the highway test cycle, which was added in 1974, was only predicted to cost the agency

85 George C. Wilson, “EPA Wins Its Fight to Keep Car-Mileage Tests in Agency,” Washington Post, May 18, 1974. 86 “EPA Still Holds Reins in Mileage Labeling,” Automotive News, May 27, 1974. 87 Comptroller General of the United States, “Review of the Automobile Fuel Economy Testing and Labeling Program,” Report to the Subcommittee on Conservation and Natural Resources, Committee on Government Operations, House of Representatives, August 15, 1974. 88 Ibid., 3.

359

$178,000.89 The NBS, on the other hand, estimated that if it took over fuel economy testing, by the 1976 fiscal year, its costs would be around $3,080,000.90 Moreover, the OMB was forced to admit that the NBS would probably have to run an emissions test before testing any car for fuel economy because emissions could not be sacrificed for efficiency. Thus, any other fuel economy program was beginning to look very redundant.

On September 20, 1974, the EPA, with the FEA, released its second major fuel economy publication at the National Press Club. The news was good: catalytic converters on GM and Ford cars had increased fuel efficiency by an estimated 13.5 percent.91 Emissions control was looking a bit better than it did at the beginning of the OPEC Oil Embargo. Stork later heard that, by November 1974, the NBS had actually put together a “budget request for the staff to plan [a fuel economy] program.” The NBS moved at a “geological pace,” he wrote.92 But the “battle [was] long over, and Ash said, forget it.” Yet, the issue was not completely “closed” until

December 1975 when Congress passed the Energy Policy and Conservation Act, which legislatively ensconced the EPA’s fuel efficiency testing in the federal automotive fuel economy standards.

89 Ibid., 17. 90 Ibid., 19. 91 Opening Statement by Russell E. Train, Administrator, U.S. Environmental Protection Agency, “1975 Fuel Economy Data,” September 20, 1974, in the personal collection of Eric Stork. The EPA knew about—and publicly predicted—these fuel economy gains early in the year. Because the companies were relying on catalytic converters, they could move away from systems and procedures, such as spark retard and “riching-up” the fuel mix, that increased fuel consumption and decreased performance. “EPA’s Chief Asserts ’75 Autos Will Use Less Fuel Than ‘74s,” Wall Street Journal, March 26, 1974. 92 Stork, “The Genesis of the EPA Fuel Economy Program.”

360

Not all perils to the EPA’s program came from outright power grabs. Threats to the agency’s credibility could be just as damaging. One such early threat came from the Federal Trade Commission (FTC). In late-1973 and all of 1974, the FTC’s

Bureau of Consumer Protection carried out an investigation into advertisements about fuel efficiency. When energy prices spiked, auto companies began making many wild and unproven claims about their cars’ frugality, their ability to squeeze miles out of a gallon of gas. Moreover, many ads tied notions of frugality to a vehicle’s perceived size and not to more important factors, such as the vehicle’s weight, that were invisible to the consumer’s gaze.93 The FTC publicly recognized the importance of the EPA labeling program, and the investigation was ostensibly focused solely on advertising claims, but the FTC also noted that the plethora of information available to the public caused “confusion.” Groups, such as Consumer

Reports and Road and Track Magazine, published numbers that varied widely from the EPA’s statistics. The popular press complained about this variation, and

Consumer Reports published an article entitle, “Gas Mileage: Whom Do You

Believe?”94 In its announcement of a public hearing, the FTC asked, “Can there be devised a single test procedure for determining fuel economy under typical driving conditions, the results of which should be required to appear in the advertising or other forms of promotion of cars?” And, “Who should actually conduct the testing of individual cars according to the procedure?”95

93 The Federal Trade Commission explained its intentions and reasoning in its first announcement of a public hearing on the matter. Federal Register, Vol. 39 (September 24, 1974), 34382–34384. 94 Consumer Reports, April 1974. 95 Both quotes from Federal Register, Vol 39., 34383.

361

These questions posed real threats to the EPA. Stork’s oft-repeated saying, which forms the epigraph of this chapter, was “All real influence depends on the ability to obtain, assimilate, and process information.”96 In other words, knowledge and information are power, and if the powers that be remain reliant on a bureau’s information, they will also be, in Max Weber’s words, “dependent” upon it. Stork knew this. If the Office of Mobile Sources was going to keep its influential place in auto regulation, it needed to retain its fuel efficiency information program.

One of the staff members at the FTC assigned to the fuel economy investigation was a young lawyer named Ernest (Ernie) Rosenberg. The son of

Polish immigrants, Rosenberg was born and reared in Brooklyn and, after attending the University of Rochester, studied law school at New York University. Near the end of his time there, his professor, Robert Pitofsky, an expert in antitrust law, was tapped by the Nixon Administration to head the FTC’s Bureau of Consumer

Protection.97 Pitofsky told Rosenberg that he should come along to the FTC, claiming that the experience would be both formative and enjoyable and that

Rosenberg would never take a pay cut and work for federal government later if he

96 This idea can be traced at least back to Max Weber, who wrote, “[E]very bureaucracy seeks to further increase [its superiority] through the means of keeping secret its knowledge and intentions.” “ Max Weber, Economy and Society: An Outline of Interpretive Sociology, Guenther Roth and Claus Wittich, eds., Vol. 2, (Berkeley: University of Chicago Press, 1978), 992. Scholars of public administration and organization studies later examined this idea more formally. See, for instance, James G. March and Herbert A. Simon, Organizations (New York: Wiley, 1958); Andrew M. Pettigrew, “Information Control as a Power Resource,” Sociology, Vol. 6, No. 2 (May 1972), 187–204; idem., “Towards a Political Theory of Organizational Intervention,” Human Relations, Vol. 28 (April 1975), 191–208; Jeffrey S. Banks and Barry R. Weingast, , “The Political Control of Bureaucracies under Asymmetric Information,” American Journal of Political Science, Vol. 36, No. 2 (May 1992), 509–524. For his part, Stork does not remember where he picked up this bit of wisdom. Most likely, the idea was simply “something in the air” in federal agencies. 97 For Robert Pitofsky’s biography, see Who's Who in America - 2011, 65th Edition (2010).

362 got used to a lawyer’s salary in a private firm. It was either now or never, Pitofsky said, and Rosenberg took the offer.

At NYU, Pitofsky and his student Rosenberg had already been studying the notion of consumer information. They were interested in the idea that producing and disseminating accurate information about consumer goods might create a more

“rational marketplace” and decrease some technological risks (by making risky technologies undesirable to consumers). But misleading advertising created a real barrier to consumer information efforts. Pitofsky’s thinking on some of these matters is contained in an article he published later, “Beyond Nader: Consumer

Protection and the Regulation of Advertising.”98 In a section titled, “Market Failure as a Cause of Consumer Abuse,” Pitofsky argued against the writings of “free market” scholars, like George Stigler and Richard Posner, who claimed that the

“market” itself would generate sufficient information for consumer choices.99

Pitofsky believed, with many economists, that advertising provided important information, writing, “Advertising substitutes for search costs by consumers by providing in a convenient and usable form information necessary for consumers to make choices among available brands, and in the process facilitates the functioning of a market economy.”100 But he also thought that there were a host of scenarios in which companies would fail to provide relevant information in ads. He pointed out that consumers did not know about the “durability of light bulbs, octane ratings for

98 Pitofsky, “Beyond Nader: Consumer Protection and the Regulation of Advertising,” Harvard Law Review, Vol. 90, No. 4 (Feb. 1977), 661–701. 99 Stigler’s classic article on this topic, which Pitofsky cites, is “The Economics of Information,” The Journal of Political Economy, Vol. 69, No. 3 (June 1961), 213–225. 100 Pitofsky, “Beyond Nader,” 663.

363 gasoline, tar and nicotine content of cigarettes, mileage per gallon for automobiles, or care labeling of textile wearing apparel” until “government intervened and required or induced disclosure.”101 Interestingly, Pitofsky attributed this lack of consumer information to “market structure,” noting, “The causes of such market failure are not particularly difficult to trace.” A monopoly, clearly, would not purposely show its own warts. More fascinating, however, was the case of oligopolies, which would fail to mention undesirable characteristics of a competitor for “fear of triggering mutually disadvantageous competition.” As Pitofsky concluded,

“By emphasizing undesirable attributes of rival products, an ‘octane war,’ a ‘tar and nicotine derby,’ or competitive safety claims for pesticides or flammable fabrics may be as destructive of mutual oligopolistic interests as a price war would be. Thus, when no one company has decisive advantage over others, all may reach the conclusion that avoidance of certain kinds of product claims is mutually advantageous.”102

Therefore, when Pitofsky and Rosenberg came to the FTC, they already had ideas about solving such “market failure” by regulating truth in advertising.

The case of fuel economy advertising was an interesting case in which to apply Pitofsky’s thoughts on consumer information. As the Federal Register notice stated frankly, “In order for consumers to be able to distinguish among competing cars on the basis of fuel economy, they need adequate, accurate information.”103

Moreover, in April of 1964, the Consumers Union, publisher of Consumer Reports, filed a petition with the FTC, requesting that the Commission begin an investigation

101 Ibid., 664. 102 Ibid., 665. 103 Federal Register, Vol 39., 34382.

364 into the fuel mileage claims of automakers, although the FTC said that it was already conducting such an investigation.104 Three months later, the FTC castigated the auto firms for making spurious claims in their ads. The automakers reacted to this criticism differently. GM caved and “agreed in a proposed consent order to drop an advertisement saying its Cadillac Eldorado had finished mileage tests ahead of 73 other cars.”105 Ford and Chrysler, however, refused to back down, arguing that the claims in their ads were “both accurate and fair.” The FTC’s Director of Consumer

Protection claimed, “We have doubts that any [tests],” including the EPA’s, “are so superior that they can be used to the exclusion of all the others.”106

During the early-to-mid 1970s, the FTC was also experiencing a change that would aid Pitofsky and his staff members in regulating consumer issues, such as advertising. For decades, the Commission had issued “guidances,” which more or less said, “If you—the producer, distributer, marketer, or seller—violate these criteria, we will sue you.” But this system was fairly onerous both in terms of money and resources for the Commission to enforce. For instance, if the Commission wanted to bring suit against a firm for misleading advertising, it would have to conduct consumer surveys and focus groups to make the case that consumers understood the ad in some way that mislead them.107 Every case had to be tried de novo. In the 1970s, a new “policy innovation” was introduced—the trade regulation rule. Under this system, the FTC only had to conduct a rule-making procedure to

104 “Consumers Union Asks FTC for Uniform Test for Cars’ Fuel Mileage,” Wall Street Journal, April 11, 1974. 105 Walter Rugaber, “Trade Commission Assails Ads By Car Makers on Fuel Savings,” New York Times, August 1, 1974. 106 Ibid. 107 Interview with Ernest Rosenberg at Rosenberg’s house, April 15, 2010.

365 establish the formal trade regulation rule. If the rule stood up to legal challenges, the

FTC could then fine someone for breaking the rule without needing to build the entire case from scratch. The Commission only had to demonstrate that the rule had been violated. Congress gave the FTC the power to create trade regulation rules through the Magnuson-Moss Warranty—FTC Improvement Act, which was passed in early 1975.108 At some point—perhaps even before the law was passed, given that they knew it was coming down the pipe—the Commissioners gave Pitofsky and his staff members permission to pursue fuel economy advertising as a trade regulation rule.

Under Pitofsky’s direction, Rosenberg and others began to study the issue of fuel economy, and this work entailed examining what groups were using what tests and analyzing which of the tests was “best.” Through this work, Rosenberg met

Stork. Since the EPA created the federal fuel economy numbers and ran the labeling program, Rosenberg contacted the agency. Stork knew that he needed Rosenberg and the FTC to find that the EPA’s numbers were the best available, or at least to allow the EPA to be the national standard setter in a field of other possibilities. The two met, and Stork answered Rosenberg’s questions. Near the end of the meeting,

Stork suggested that Rosenberg should go to Ann Arbor to visit the EPA’s testing facilities and to get a feel for the kind of work Stork’s staff members were doing.

Rosenberg thought that was a great idea, but, as a junior staff member at the FTC, he didn’t have a travel budget. Here, Stork devised his own form of “regulatory

108 See Earl W. Kintner and Christopher Smith, “The Emergence of the Federal Trade Commission as a Formidable Consumer Protection Agency,” Mercer Law Review, vol. 26 (1974-1975), 651; Caswell O. Hobbs, “Legal Issues in FTC Trade Regulation Rules,” Food, Drug, and Cosmetic Law Journal, Vol. 32 (1977), 414–422.

366 capture”; one regulator sought to capture the loyalties of another, potentially competing regulator. He offered to send Rosenberg to Michigan on the EPA’s buck.

Rosenberg took Stork up on the offer, and he was surprised what he found in

Ann Arbor. There, he spent most of his time with Austin and Hellman and also with

Charles Gray, Joe Somers, and others, who “wined and dined” him. Rosenberg found the EPA staff members to be extremely competent, committed, and smart. But more importantly perhaps, he had expected to find a bunch of stiffs who hated the auto industry. In fact, he found just the opposite. Many of the staff members at the Ann

Arbor labs were car nuts. They had hot rods and sports cars in their home garages on which they worked on the weekends. These people loved cars; they just wanted to improve them. He left Ann Arbor a true believer in the EPA’s testing methodology.

Rosenberg continued to work on fuel economy advertisements and other issues for several months before Stork contacted him. Both Rosenberg and Stork tell almost identical stories about what happened next. Stork called Rosenberg out of the blue one day and began telling him a dirty joke: One day while walking towards the corner of 4th and Pike in Seattle’s downtown, a man noticed another man who was wearing a trench coat standing across the street. Every time a woman walked by the man in the trench coat, he would walk up to her and say something.

Some women reacted with horror. Some smiled. A few even slapped him in the face.

Finally, the other man could no longer stand the mystery. He had to know what was going on, so he walked over to the man in the trench coat and asked him what he was saying to the women. “I ask them to fuck,” the man in the trench coat said. “No

367 wonder you get slapped a lot,” said the man. “Sure,” said the man in the trench coat,

“but I get laid a lot, too.” Rosenberg thought this joke was pretty good, but he had no idea why Stork was telling it to him. Stork, in his own circuitous way, was asking

Rosenberg to come work for the Office of Mobile Sources, and after some consideration Rosenberg took him up on the offer. Rosenberg’s arrival to the EPA was a win-win situation for both him and Stork.109 When William Ruckelshaus reorganized the EPA, he centralized all of the agency’s lawyers into the Office of

General Counsel (OGC), which would sometimes shut down Stork’s plans because they violated OGC’s interpretation of the letter of the law. Rosenberg could now give Stork in-house legal advice—and vice versa. As Stork later recalled, “I wanted, among other capabilities, someone who could hold up my view of what the law provided.”110 Rosenberg’s reasons for coming to the EPA were more personal, even visceral. As he recalled, “You have to understand . . . for a kid at the time [he was 27 when Stork offered him the job] . . . this was the program that was on the front page of the Washington Post and the New York Times every day, virtually. When I would go out to a bar or something like that and I would tell some girl, you know, what I did, [they would say,] ‘Really? That’s really cool. You’re beating up Detroit. You must be a great guy.’ This is the ‘70s. Everyone hated industry, and most of us hated government too, but EPA was sort of an exception.”111

On September 10, 1975, nearly a year after the original notice of public hearing, the FTC issued a ruling stating that, at least on an interim basis, all

109 Interview with Ernest Rosenberg at Rosenberg’s house, April 15, 2010. 110 Email from Eric Stork, February 8, 2011. 111 Interview with Ernest Rosenberg.

368 advertisements making fuel economy claims had to use the “appropriate EPA mileage figure.”112 The FTC claimed that certain things, especially the contentious issue of highway mileage, put the EPA’s numbers in doubt, so the Commission stopped short of making a “final trade regulation rule.” As one FTC official had earlier told the Wall Street Journal, “We aren’t proposing to announce the EPA figure as God’s word because it isn’t.”113 But for all intents and purposes, the Office of

Mobile Sources’ economy measurements were established as the “official” statistics that every advertisement on fuel economy had to mention. Stork and the EPA had won. The FTC never rescinded this decision. Rosenberg reports, however, that the

FTC’s ruling in favor of the EPA’s test procedures was not written in stone. Indeed, during his first week at the EPA, he heard that the Commission was going to decide in favor of another test. He rushed over to the FTC building and demanded to talk to the Commissioners. This action was seriously out of order for a mid-level staff member. Not only was he violating his role at his new employer, the EPA, but he was also subverting the chain of command at his old one, the FTC. Still, the

Commissioners found Rosenberg’s case convincing; he won them over. The next day,

EPA Administrator Russell Train came to visit Rosenberg in his office. Rosenberg expected that Train was going to rebuke him for violating protocol, if not fire him altogether. Instead, Train thanked him for looking out for the agency’s interests.114

Winning the FTC to EPA’s side meant more to the Office of Mobile Sources than ensuring that its numbers were included in all pertinent national advertising,

112 Federal Register, Vol. 40 (September 10, 1975), 42003. 113 Kenneth B. Bacon, “Car Makers Soon May Face Ad Standards for Fuel-Mileage Claims, FTC Staff Says,” Wall Street Journal, April 15, 1974. 114 Interview with Ernest Rosenberg, April 15, 2010.

369 although such advertising was in itself important publicity for the agency. It also meant that the EPA had won the relative loyalties of the FTC and the credibility and legitimation that came with that loyalty. In the landscape of Washington, D.C., the

FTC was a powerful actor indeed, and the EPA was considerably emboldened by gaining the Commission’s favor.

Heading towards CAFE

In December of 1973, the Senate passed Sen. Henry M. Jackson’s National

Fuel and Energy Conservation Act, which directed the Department of

Transportation to carry out a study of regulating fuel economy. The act would wait for six months to be reconciled with a House version of the bill, when it was eventually passed as the Energy Supply and Environmental Coordination Act. In the end, the law directed the Department of Transportation and the Environmental

Protection Agency to carry out a 120-day study on the possibility of improving automotive fuel efficiency. The main objective of the study was to decide whether it was possible for automakers to increase the fuel efficiency of vehicles by twenty percent by 1980.

Responsibility for the study fell to two men, one drawn from each agency to co-chair the effort. The first was EPA’s Stork. The second was Eugene Goodson, a professor at Purdue University who was spending two years as the Department of

Transportation’s Chief Scientist.115 Goodson had earned his PhD in Engineering at

Purdue University in 1963, and stayed on as a professor, earning a full professorship

115 Goodson’s resume is available online at www.bus.umich.edu/FacultyBios/CV/ggoodsn.pdf (Last accessed February 20, 2011).

370 in 1970. In 1973, he took a leave and became the DOT’s Chief Scientist. After completing his stint at the DOT, Goodson returned to Purdue until the early 1980s, when he moved into industry. During that period, he was also a Member of the

Policy Board of the famous US—Japan Study on the Future of the Automobile, co- administered by the Massachusetts Institute of Technology and the University of

Michigan.

As Stork later characterized the matter, “EPA [was] very suspicious of DOT and vice versa.” Stork wrote, “At first, [it was] like two porcupines making love.”116

But Goodson and Stork learned to work together. They assembled staff members into seven panels—Technology Assessment, Economic Analysis, Fuel Economy

Measurement, Policy Analysis, Truck-Bus Assignment, Safety Implications, and Air

Quality Implications.117 Typically, the panels’ “principal members” consisted of one person from DOT and one person from EPA—to “combine the two agencies’ expertise” no doubt but also to keep an eye on each other.

The Task Force for the 120-Day Study met at the DOT building in

Washington, D.C., on September 9–10, 1974. The report had to be submitted to the

Transportation Systems Center in Cambridge, MA, where the report would be printed. The group was under some pressure, and it had a number of issues on its slate to consider. Although the group had specifically been charged with

116 Eric Stork, “The Genesis of the EPA Fuel Economy Program.” 117 Memo, Study Task Force Co-Chairmen [Goodson and Stork] to Distribution, “120-Day Fuel Economy Study—Panel Workshop, August 28, 1974, Office of Mobile Source Pollution Control, Air Standards and Regulations, 1966–1977, Environmental Protection Agency, Record Group 412, National Archives at College Park. Hereafter, EPA Records.

371 determining whether the auto industry could improve automotive fuel economy by twenty percent by 1980, it had a number of other issues to consider as well.

For instance, several regulatory and other policy tools could be used to increase fuel economy and, thus, reduce overall automotive fuel use, and the Task

Force had to sort out which approach would work best. The group considered a wide variety of options, including “tax on gasoline [crossed out], tax on inefficient cars [gas guzzlers], raise load factors, reduce congestion, reduce auto usage— promote mass transit, and disclosure of autos’ fuel economy.”118 Time and again, however, three possibilities were put into a central position in policy discussions: first, although the federal government could require a given fuel economy for every new car, this approach was seen as undesirable because it would reduce the variety of vehicles available, including many kinds of cars that consumers—and voters— wanted. The second and third options both involved “averaging”—that is, taking the sum of all the cars a given automaker sold and figuring out the statistical mean fuel economy for all of those vehicles. In the second option, the federal government would require each automaker to increase its averaged fuel economy by a given percentage. This option had two problems, however. It was harder on manufacturers—especially foreign automakers—who already had a fuel-efficient car fleet. They would be especially pressed to find ways to squeeze the mandated percentage out of the cars’ fuel use. Moreover, from a regulatory point of view, this option was the most difficult to administer. The third option would require the average of each automakers’ fleet to meet a certain standard, a specific MPG. This

118 “Outline: 120-Day Study: Chapter 3,” an undated draft outline, EPA Records.

372 option, too, had at least one problem; namely, it favored companies that were already producing efficient vehicles. But in the broad sweep of things, this seemed like a small price to pay, and it became the dominant model.

The discussion over policies for increasing fuel economy marked an important change in the history of regulation during this period. The Study Group’s policy papers mentioned the “free market” as one possible solution to the fuel economy problem.119 As the final 120-day Study remarked, “The study does not make the judgment of whether a Federal improvement standard is needed inasmuch as a 20 percent improvement goal may be reached solely through the forces of the market. . . . [Yet,] consumer demand for better fuel economy may not be strong enough to induce manufacturers to opt for the substantial improvement possible.”120 This notion that the “market” would sort out the problem at hand was noticeably absent from discussions about auto safety and emissions control. But it would come to dominance a little over half a decade later with the election of Ronald

Reagan. The roots of neoliberalism were beginning to take hold.

Additionally and importantly, through its panel on Safety Implications, the

Task Force realized from the beginning of its studies that automotive fuel economy and automotive safety were interrelated. In some instances, they could both be positively affected. For example, the reduction of national speed limits meant to

119 For instance, one draft study, “Improved Fuel Economy for New Automobiles and Small Trucks,” mentions “free-market response” as one solution to the issue of fuel economy. Un paginated. Another paper, which was considering strengths of “Voluntary Fuel Economy Goals and Voluntary Fuel Economy Labeling,” noted that one of the “positive aspects” of such a program was that it “relies on free market, (e.g. consumers demand more efficient vehicles and goals to provide producer incentive) rather than federal intervention.” “Improvement of New Car Fuel Economy: Discussion of Four Policy Options,” August 19, 1974, 3. Both studies in EPA Records. 120 Department of Transportation and Environmental Protection Agency, “Potential for Motor Vehicle Fuel Economy Improvement,” Report to Congress, October 24, 1974, 3

373 decrease fuel usage during the “energy crisis” had also reduced highway fatalities.121

But more often they had an inverse relationship. Because the best way to improve fuel economy is to reduce vehicle weight, cars would get lighter, but also more dangerous, especially if they collided with heavier vehicles.122

The EPA always outpaced the DOT during the study. The Office of Mobile

Sources simply had greater capabilities for “obtaining, assimilating, and using information” on fuel economy and more expertise on the matter. No DOT staff members were publishing SAE papers on the topic, like Austin and Hellman were.

Moreover, the EPA had the only comprehensive database, so many of the analyses depended on the EPA’s data, and no one really had any steady ground from which to question the Office of Mobile Source’s claims. Finally, the Office of Mobile Sources’ expertise lay in technology assessment, which played a central role in the 120-Day

Study.

The 120-Day Study found that automakers could increase fuel economy by

1980. In fact, it found that they could do much better than that, suggesting that “The

121 “Effect of Speed Limits on Fuel Economy and Safety,” EPA Records. 122 This issue would become very controversial in later years, especially after petroleum prices fell in the mid-1980s, and people wondered why fuel economy standards were necessary. A classic article in this vein, published in 1989, is Robert W. Crandall’s and John D. Graham’s “The Effect of Fuel Economy Standards on Automobile Safety,” which mentions both the low fuel prices in the 1980s and the desire of many, including President Ronald Reagan, to revoke the law. Viewing federal fuel economy standards through the lens of cost-benefit analysis, the authors argued, “Obviously, the [federal fuel economy] program forced vehicle manufacturers to invest more resources in developing fuel efficiency than 1985 gasoline prices warranted.” (117) Having argued that fuel economy standards increased fatalities by fourteen percent, Crandall and Graham concluded, “When safety considerations are included, [fuel economy standards] appear to be a very costly social policy.” (118) Robert W. Crandall and John D. Graham, “The Effect of Fuel Economy Standards on Automobile Safety,” Journal of Law and Economics, Vol. 32, No. 1 (April 1989), 97–118. Others, however, pointed to the increasing presence of large consumer vehicles, namely sports utility vehicles (SUVs), in increasing fatal crashes. See, for instance, Keith Bradsher, High and Mighty: SUVs—The World’s Most Dangerous Vehicles and How They Got That Way (New York: Public Affairs, 2002).

374 full range of potential improvements . . . is from 40 to 60 percent.”123 Stork recalls

Goodson and himself sitting at Stork’s dinner table late one night, knowing that they had to send the report off the following morning. Stork sent one of his staff members, Gary Timm, to print the report, with the direction that Timm should not allow one word of the report to be changed.124 Goodson did the same with one of his own staff members. The EPA and the DOT had worked together, but they still didn’t trust each other.

Congress received the report. The report had intentionally made no recommendation for at what level a potential standard should be set, but a number began floating around, a sales-weighted fuel economy average of 28 MPG for each company, or twice the national average of 14 MPG in 1974.125 This standard found its home in the Energy Policy and Conservation Act, passed December 22, 1975, a law whose automotive portions largely captured the intentions of the earlier, failed

National Fuel Economy Testing Act of 1974. Through a compromise with the auto industry, this number was eventually reduced to 27.5, supposedly by factoring in a

0.5 MPG deduction for energy inefficiencies caused by emissions controls.126 The national fuel economy standards eventually became known as “Corporate Average

Fuel Economy,” or CAFE, Standards, though this term appears nowhere in the law.

123 Department of Transportation and Environmental Protection Agency, “Potential for Motor Vehicle Fuel Economy Improvement,” Report to Congress, October 24, 1974. 124 Phone interview with Eric Stork, January 5, 2010. 125 See “Potential for Improvement of Energy Efficiency of Automobiles and Other Major Applicances,” August 29, 1975, a report prepared by the staff of the House Energy and Power Subcommittee, which used the 120-Day Study and a report by the RAND Corporation to justify a 28 MPG level. 126 Committee on Interstate and Foreign Commerce, “Summary of Major Provisions of Energy Conservation and Oil Policy Act of 1975 (H.R. 7014) as Passed the House,” September 26, 1975, 5. “Standards are required to be adjusted downward to reflect any decrease in fuel economy in excess of ½ MPG which results from more stringent motor vehicle emission standards.”

375

(Staff members who worked for the EPA’s Office of Mobile Sources before and after the passage of the federal fuel economy standards believe that the term “CAFE” was a creation of the automakers, who meant to insinuate that federal regulators were people who spent the majority of their time hanging out in cafés. Thus, in deference,

I have avoided using the term “CAFE” throughout this chapter.)

In the US House of Representatives, the issue of fuel economy had fallen into the hands of the House Committee on Interstate and Foreign Commerce, one of the oldest and most powerful committees in Congress. The Committee had its roots in the Committee on Commerce and Manufacturers, which was created in 1795.

Furthermore, one of Committee’s subcommittees, the Subcommittee on Energy and

Power, took control of the issue of the fuel economy issue. And this Subcommittee happened to be led by a Michigan Congressman named John David Dingell, Jr.127

Dingell grew up in Michigan, where his father was the Congressman for the state’s

15th District, a position that he inherited when his father died in 1955. Dingell, Jr., was and remains a controversial figure. While exceptionally progressive on some issues—he has famously introduced a single-payer health care bill at the beginning of every session of Congress—Dingell has also come under significant criticism for doing the bidding of the auto industry, often shooting down or weakening laws that would have important repercussions for consumers because the laws threaten the

127 Dingell renamed the Committee on Interstate and Foreign Commerce the “Committee on Energy and Commerce” in 1981 in order to highlight its role in rationalizing and planning the nation’s energy policy. Ironically, Dingell made this name change at the same time Ronald Reagan was taking control of the presidency. After campaigning on a ticket that included abolishing the Department of Energy—as well as the Department of Education—Reagan believed that the best energy policy was no energy policy. The energy market would tend to its own needs. And so, the very impetus for the Committee’s new name—the desire to use federal power to “plan” the nation’s energy use—withered on the vine.

376

Big Three’s bottom lines. Dingell is an embodiment of Simon Cameron’s famous statement: “An honest politician is one who, when he is bought, will stay bought.”128

The Committee on Interstate and Foreign Commerce had two subcommittees that could have possibly overseen the fuel economy law—Dingell’s Energy and

Power Subcommittee and the Subcommittee on Public Health and Welfare, which had played a central role in passing the Clean Air Act Amendments of 1970. The division between these subcommittees also reflected a division within the executive branch, with Dingell’s Subcommittee overseeing the Department of Transportation

(DOT) and Subcommittee on Public Health and Welfare overseeing the EPA. By ensuring that the DOT controlled the standards-setting aspect of the Energy Policy and Conservation Act, Dingell guaranteed that he would have oversight.

As the way the law was written, the EPA would retain the task of measuring fuel economy. Notably, Stork began jockeying for this position very early in the

“energy crisis.” As the Detroit Free Press writer Douglas Williams remarked in an article dated January 18, 1974, “Who told the Environmental Protection Agency to concern itself about how many miles per gallon of gasoline you get in your

Brontosaurus V8? Well, nobody. But it’s a vacuum the aggressive new government agency chose to fill.” He then quoted Stork, who said, “We have no authority in law to enforce fuel economy.” But Stork still argued that the EPA was the “logical agency for the task.” As he explained, “First, we are clearly under attack by people who say

128 Dingell will one day be the subject of a fascinating biography, but no such work yet exists. Several works, however, describe Dingell’s legislative prowess at taking control of any issue that he desires during Congressional “turf wars.” Any law that affects the auto industry clearly fits the criteria of Dingell’s interests. See, for instance, David C. King, Turf Wars: How Congressional Committees Claim Jurisdiction (Chicago: The University of Chicago Press, 1997).

377 emission control has to go for the sake of fuel economy. It is our view that there is no conflict at all between clean air and transportation . . . [Second,] of all government agencies, we have far and away more data on fuel economy than anyone else in the government.”129 Ironically, Russ Train told Williams on the same day that he would begin pushing for mandatory fuel economy requirements. Asked where the standard should be set, Train said that “a reasonable one” would be around 13.5 MPG—very short of 27.5 MPG that the federal government eventually required.130 By the Office of Mobile Sources’ retention of the fuel economy testing program after the passage of federal fuel economy standards, Stork had won the battle he’d been fighting for two years.

Goodson and other members of the DOT were never happy with the 120-Day

Study, believing that it was too hard on industry. For this reason, Goodson began a self-initiated investigation, which became known as the 300-Day Study, but he left and returned to Purdue before it was completed. The 300-day Study then fell to the

DOT’s Assistant Secretary for Systems Development and Technology, Hamilton

Herman. Herman felt very passionately about the study and put a lot of work into it, but he barely finished it before the Carter administration entered office. Though the study was published, the new administration had no interest in following the lines of the report, which was more favorable to industry than the 120-Day Study.131 If

129 Douglas Williams, “EPA Jumps in on Gas Issue,” Detroit Free Press, January 18, 1974. 130 Ibid. Of course, at this early date, the EPA had not settled on the idea of requiring an “average fuel efficiency” for each firm. So, Train may have very well been talking about a minimum standard of 13.5 MPG for all regulated vehicles, which would have created a very different regulatory structure indeed. 131 “The Report by the Federal Task Force on Motor Vehicle Goals Beyond 1980,” September 2, 1976, in Eric Stork’s personal collection.

378 anything, the people whom Carter brought with him were going to tighten regulations, not loosen them.

In 1978, as he reflected on the preceding years, Eric Stork still believed that, had his agency not fought to keep the fuel economy program, automotive emissions controls may have gone out the window in 1974 or 1975, even in the midst of the

“energy crisis.” As he wrote, “Certainly not all we did was gentlemanly, or in good taste. But we thought that our adversaries were unfair [and] wanted to use [the] energy crisis to dump the Clean Air Act. Getting something done in our society takes more—much more—than having truth on your side. You’ve got to get your truth recognized and accepted. That doesn’t happen by itself [and] can’t be made to happen by laws or organization charts. But it can be done with luck, hard work, and a lot of chutzpa.”132 The strange structure of federal fuel economy standards—a structure that remains to this day—stands as a testament to Stork’s and the Office of

Mobile Source’s gamesmanship. Whether this gamesmanship ever became brinksmanship—a moment in which Stork was willing to lose it all—we may never know. The personal and organizational risks of those days have been largely covered over with the passing of time, but the EPA kept its piece, and, now, the bizarre division of federal fuel economy standards remains an ossified structure.

Stork wrote that he meant his seminar presentation as an “illustration of how things evolve.”133 Things are “disorderly, controversial, then [they] settle down.” Drawing on a famous statement about law that is often attributed to Otto von Bismarck, Stork suggested that policy was like sausage—both are things that you lose your taste for

132 Eric Stork, “The Genesis of the EPA Fuel Economy Program.” 133 This and following quotations, Stork, “The Genesis of the EPA Fuel Economy Program.”

379 if you see them being made. Above all, Stork suggested, the actual policy process is nothing like the “concepts of PPBS [Program-Planning-Budgeting-System], MBO

[Management by Objectives], ZBB [Zero-Based Budgeting], or POSDCORB

[Planning–Organizing–Staffing–Directing–Co-Ordinating–Reporting–Budgeting] would imply,” all of which were administrative management concepts dominant in the day. This does not mean that Stork and his staff members did not know how to manage their resources, just the opposite. They effectively brought all available capabilities to bear to protect their programs.

Epilogue

“Of late we’ve been told that America’s love affair with the automobile is over, that people want some other method of transportation. . . . [The] Generals in the war against the automobile [are] a “New Class” of intellectuals, journalists, burocrats [sic], and academics who are anti-materialist and opposed to the basic values of American society. . . Well, which side are you on in this “War Against The Automobile”? I’ll tell you where I stand—I’m going out to the parking lot and give my gas buggy an affectionate pat on the trunk and tell it how much I care. This is Ronald Reagan. Thanks for listening.”

Ronald Reagan, from one of his syndicated radio segments, November 29, 1977.1

The regulatory institutional developments covered in this dissertation— from the passage of the Traffic Safety Act in 1966 to the creation of fuel economy standards in the Energy Policy and Conservation Act in 1975—faced a changing political and social environment in the late-1970s and, even more so, in the

1980s. Ronald Reagan brought a new philosophy of governance with him to the

White House. The Carter Administration had already “deregulated” some industries, and there was talk of deregulating others. These deregulation efforts are more accurately described as “regulatory restructuring” because no industry went from being governed by a slew of rules to having no regulations whatsoever. Still, in industry, in academia, in the public, and in the government itself—something was shifting. Scholars often call this trend the rise of

“neoliberalism.”2

1 Kiron K. Skinner, Annelise Anderson, and Martin Anderson, eds., Reagan’s Path to Victory: The Shaping of Ronald Reagan’s Vision: Selected Writings (New York: Free Press, 2004), 237–238. Italics are from the original text. Many of Reagan’s radio speeches were written by others, so it is difficult, if not impossible, to attribute authorship. 2 Two primers on neoliberalism are David Harvey, A Brief History of Neoliberalism (Oxford: Oxford University Press, 2005) and Manfred B. Steger and Ravi K. Roy, Neoliberalism: A Very Short Introduction (Oxford: Oxford University Press, 2010). I will cite other works on neoliberalism in the substantive discussion below.

380 381

The popular account of neoliberalism describes this emergent economic and political philosophy as a series of negative policies—all aimed at ending “big government” and giving rein to the “free market”—including deregulation, decreased taxation, the reduction of government-sanctioned organized labor, and the lowering of trade barriers, including tariffs, import quotas, and capital controls. The policies were negative in the sense that they showed what people who backed them were against. If neoliberals were in favor of any positive thing, it was the “free market.” But thinkers described even the “free market” negatively in terms of the tethers that it lived without, rather than in positive terms of what it was.3 In 1989, a former International Monetary Fund advisor

John Williamson described a list of ten negative policies as the “Washington

Consensus.”4 Things had changed greatly since the time when even Republican

President Richard Nixon described himself as a follower of John Maynard

Keynes, the philosopher-king of the active state that intervened in the economy.

Yet, this popular account neglects how neoliberalism also had positive and prescriptive ideas about society. Many of these visions related to how a

3 My description here of the negative and the positive mirrors Isaiah Berlin’s discussion of negative liberty, which can be described as “freedom from” some external power, and positive liberty, which involves the freedom attained when someone reaches his or her fullest potential. A good liberal, Berlin believed that the just society should only embrace negative liberty, which left a person to live out life as he or she desired. Positive liberty, in Berlin’s estimation, always led to some form of tyranny because the powers that be would have to posit some way of being as the “good life.” See Berlin, “Two Concepts of Liberty” in The Proper Study of Mankind: An Anthology of Essays (New York: Farrar, Straus, and Giroux, 1997), 191–242. 4 Williamson coined the term “Washington consensus” in “What Washington Means by Policy Reform” in Latin American Adjustment: How Much Has Happened (Washington, D.C.: Institute for International Economics, 1990), ch. 2.

382 person should comport him or herself.5 Neoliberalism had two archetypes that taught the lesson of who a person should and should not be. The villainous personality in this morality play was the “welfare queen,” as Ronald Reagan described her. Dependent upon government, lazy, lacking ambition, uneducated, and likely addicted to drugs—the welfare queen embodied all of the unfavorable characteristics of a social vampire who fed upon the hard work of others. What’s more, in the view of neoliberals, government—especially the policies of Lyndon

Johnson’s Great Society—had produced the welfare queen by creating a permanent “safety net” that fostered dependence and sloth. The polar opposite of the welfare queen—the model to be emulated and even adored—was the

“entrepreneur,” a term that was becoming increasingly prevalent in the popular press of the day. Thus, neoliberalism aimed not only to improve impersonal conditions and strengthen the economy; it desired to create a certain kind of self.

As Congressman Jack Kemp famously said, “If you tax something, you get less of it. If you subsidize something, you get more of it. In America we tax work, growth, investment, employment, savings, productivity, initiative and ability, while subsidizing nonwork, consumption, welfare, and debt. Isn’t it time to allow people to probe, test, prove their character and sensibilities, to be inspired to

5 Michel Foucault and those who have furthered his ideas have most fully realized the insight that power has not only prohibitive but also productive functions. Foucault spells this notion out most clearly in Discipline and Punish: The Birth of the Prison, trans. Alan Sheridan (New York: Vintage Books, 1977). Foucault also explicated the idea in his histories of sexuality and his lectures at the College de France. For other works that focus on the productive aspects of power, see Nikolas Rose, Powers of Freedom: Reframing Political Thought (Cambridge: Cambridge University Press, 1999); Graham Burchell, Colin Gordon, and Peter Miller, eds., The Foucault Effect: Studies in Governmentality (Chicago: University of Chicago Press, 1991); Andrew Barry, Thomas Osborne, and Nikolas Rose, Foucault and Political Reason: Liberalism, Neo-Liberalism, and Rationalities of Government (Chicago: University of Chicago Press, 1996).

383 growth and initiative?” The role of government under neoliberalism was to foster an “enterprise society,” a place where self-starters, “innovators,” and risk- takers could simultaneously improve the world around them and reap the rewards of their labors.6

Both the positive and negative visions of neoliberalism deeply affected federal regulation of the automobile. Following the election of 1980, when

Reagan’s appointees could not destroy previously existing regulations, they smashed pending ones and eviscerated programs dedicated to safety and the environment. But during the 1980s, federal agencies also began programs based on the positive, individualist aspects of neoliberalism that had lasting and beneficial effects. In this epilogue, I briefly consider these changes with an eye towards future investigations. The changes that Ronald Reagan, his advisors, and appointees brought are best viewed in stark contrast to the previous policies of Jimmy Carter. So, I begin by considering auto regulation under the

Carter Administration

In 1977, Jimmy Carter entered the White House on a campaign that distanced itself from the corruption of Watergate and the horrors of Vietnam.

Carter intended to bring a different sensibility to the White House. Although both Nixon and Ford had been institution-builders in their own ways, creating several executive-branch agencies, they did not relish that role. In the early days of his presidency, Carter symbolized return to the pro-government, pro-planning

6 On the notion of “enterprise society”—or the more common term in England, “enterprise culture”—see Rose, Powers of Freedom, 138–145; idem., “Governing the Enterprising Self” in The Values of Enterprise Culture: The Moral Debate, eds., Paul Heelas and Paul Morris (London: Routledge, 1992), 141–163.

384 ethos of the Johnson administration. He created new agencies, including the

Department of Education and the Department of Energy.7 But Carter perhaps best signaled his new vision for government through the kinds of people he appointed to lead regulatory agencies. People had long feared that regulated industries eventually began to rule, or “capture,” the very regulatory agencies that oversaw the industries. The fox came to guard the hen house. Under

Carter’s administration, the exact opposite was true. He put consumerists in charge of consumer protection agencies; he appointed environmentalists to head environmental offices. For the four years of Carter’s presidency, the pro- regulation crowd captured the flag of bureaucratic power.

In the realm of auto regulation, a few appointments had important effects.

David Hawkins, a lawyer, became the Deputy Administrator of the EPA’s Office of

Air and Waste Management, which contained the Office of Mobile Source Air

Pollution Control. After graduating from Columbia University Law School in

1970, Hawkins came to work at the National Resources Defense Counsel (NRDC), where he worked until he came to the EPA. Hawkins played an important role as a lobbyist for environmental causes throughout his time at NRDC. Additionally, he testified in favor of strict regulations and enforcement at public hearings, such as the Clean Air Act suspension hearings described in Chapter 5. When

Hawkins came to the EPA, he had a different vision than those who preceded

7 The Department of Education was split off of the Department of Health, Education, and Welfare, the remaining portions of which became the Department of Health and Human Services. The Department of Energy was created out of several preexisting agencies, including the Federal Energy Agency, the Federal Power Commission, and the Energy Research and Development Administration.

385 him. His politics and personal management style eventually led to differences with his direct subordinate, the head of the Office of Mobile Sources, Eric Stork.

Hawkins eventually placed Stork in a symbolic role, effectively firing him, and replaced Stork with Michael Walsh, who had previously worked in the EPA’s

Office of Enforcement. Like Hawkins, Walsh was a true believer in the environmental cause. When Walsh left the EPA in disgust with the coming of the

Reagan Administration, he became the director of the Manufacturers of

Emissions Control Association (MECA), a trade organization that encouraged air pollution laws around the world. As the head of MECA, Walsh traveled the globe, delivering sermons on the benefits of clean air. After Reagan came to power,

David Hawkins returned to the NRDC, where he works to this day.

A similar story held true in auto safety. Carter appointed Joan Claybrook to head the National Highway Traffic Safety Administration (NHTSA), the successor agency of the National Highway Safety Bureau (NHSB). In the early days of the NHSB, Claybrook had acted as William Haddon’s personal assistant.

After leaving the agency, Claybrook worked for Ralph Nader’s Public Citizen.

When Claybrook became the head of NHTSA, she began an ambitious program of rule-making and enforcement as well as novel efforts in publishing consumer information. During her tenure, NHTSA began crashing cars at 35 miles per hour, not for regulatory enforcement (which was primarily based on a 30 mile per hour frontal crash test) but simply for the production of data. This information was then published so that consumers could weigh the relative safety of different vehicles when making their purchases. She also made a “five-year

386 plan” for the creation of new regulatory standards and the improvement of existing ones. Even these efforts were not enough to meet Nader’s high

(sometimes overly zealous) standards, however, and Nader and Claybrook had a falling out that lasted for over a year.8 Although Claybrook did not appease

Nader, she stood no chance of remaining as head of NHTSA in the Reagan

Administration either. Detroit hated her. And so, like Hawkins, Walsh, and many, many others, when Reagan came, Claybrook went.

Beyond his appointments, Jimmy Carter made other important interventions into the workings of the auto industry. Most important, Carter initiated the first Chrysler “bailout.”9 The company had been struggling for years, but in the recessions of the 1970s, when even the more successful Ford

Motor Company and General Motors struggled to maintain profitability and market share, Chrysler fell into bankruptcy. In 1979, the Carter administration, through a complex series of maneuvers, ensured that around $2 billion was infused into the company. The bailout was a success, and the company repaid its loans within a few years. In the process, the company’s CEO Lee Iacocca became a celebrity, and his autobiography—which now litters the shelves of seemingly every used bookstore and second-hand shop—became a national best-seller.

The bailout did nothing to endear Carter to free marketeers, however.

8 Both Nader and Claybrook discuss their falling out in the documentary An Unreasonable Man (2006). 9 Stan Luger, Corporate Power, American Democracy, and the Automobile Industry (Cambridge: Cambridge University Press, 2000), 98–104; Robert Reich and John Donahue, New Deals: The Chrysler Bailout and the American System (New York: Times Books, 1985).

387

Although Carter made this very activist move to intervene in the workings of the “market” and prop up a major failing firm, the second half of his presidential tenure was marked by a decisively rightward turn. Even the

Chrysler bailout was a move to support the auto industry itself, not the consumer and environmental groups that Carter had favored at the beginning of his term. More important, however, was the Carter administration’s embrace of deregulation. In 1977, the eminent economist Alfred E. Kahn became the chairman of the Civil Aeronautics Board. He immediately set about changing the board’s policies by increasing rate competition between the airlines, opening entry to new firms, and allowing companies to create new routes.10 In 1978,

Congress embraced Kahn’s policies and passed the Airline Deregulation Act.

Deregulations of other sectors, including the natural gas industry, would follow suit, and on the whole the Carter Administration became much more conservative (at least in the economic sense).

Additionally, because the automakers were suffering, the Carter administration significantly backed off the industry. Infamously, Carter made an

11th hour settlement in December 1980 about a safety investigation into Ford

Motor Company. Since 1977, NHTSA had been looking into whether some of

Ford Motor Company’s vehicles’ transmissions slipped from park to reverse under their own volition. If the agency had decided that the company’s cars were defective, Ford would have had to recall hundreds of thousands of cars and

10 On Kahn’s deregulation of the airline industry, see Thomas K. McCraw, Prophets of Regulation: Charles Francis Adams, Louis D. Brandeis, James M. Landis, Alfred E. Kahn (Cambridge, Mass.: The Belknap Press of Harvard University Press, 1984), ch. 7.

388 spend a great deal of money during a time when it was facing financial hardships. In the Carter-Ford settlement, the company sent out stickers to customers whose cars had problematic transmissions. The customers were to put these stickers on their steering wheels. The sticker then reminded the driver very politely that he or she should firmly put the car into park, so as to avoid any potential hazards. Warning stickers were the current notion of auto safety when

Reagan came to power.

Scholars have described the intellectual origins of neoliberalism, particularly the rise of the Chicago School of Economics, several times.11

Thinkers such as Friedrich Hayek and Milton Friedman put forward “free market” visions of society, but in the 1960s and 1970s, people, both the public and policy-makers, were mostly looking elsewhere.12 A number of changes had to take place before the Chicago School found an audience.

The “free market” ideas, including deregulation, associated with neoliberalism—and the appeal of these ideas—did not arise in a vacuum. The seemingly unstoppable economic problems of the 1970s—including

“stagflation,” or simultaneous economic stagnation and inflation—put economic consensus based on the ideas of John Maynard Keynes in serious doubt. The

11 For a popular, highly critical, if weakly researched portrait of the Chicago School, see Naomi Klein, The Shock Doctrine: The Rise of Disaster Capitalism (New York: Picador, 2007). For a more balanced but also critical account of the Chicago School and the rise of neoliberalism, see Philip Mirowski and Dieter Plehwe, eds., The Road of Mont Pèlerin: The Making of the Neoliberal Thought Collective (Cambridge, Mass.: Harvard University Press, 2009). 12 Friedrich Hayek, The Road to Serfdom (Chicago: University of Chicago Press, 1944); Milton Friedman, Capitalism and Freedom (Chicago: University of Chicago Press, 1962).

389 active state that would “stimulate” the economy and solve “market failures,” including those that led to pollution, safety, and other technological risks, was under attack.

Moreover, the Republican Party had been gaining power for some time.13

After the disastrous Barry Goldwater campaign of 1964, a rejuvenated and energetic conservatism rose up in some quarters. As Lyndon Johnson predicted while signing the Civil Rights Act into law, the Democrats “lost” the South. In

1968, Richard Nixon ran on a campaign meant to win over white voters in the

South, who despaired at the Democrat Party’s support for the civil rights. This so-called “Southern Strategy” created a profound realignment in politics in the

United States, as voting districts that had voted Democrat since the end of

Reconstruction became Republican hotbeds.14 When Reagan ran for president in 1980, he stood poised to reap these benefits that had been sown before him.

Reagan also did a better job than his predecessors at aligning the interests of

“fiscal conservatives” and “social conservatives,” two factions that do not necessarily share concerns.15 If it was partly chance that Reagan adopted neoliberal ideas his platform, in retrospect, the Republican’s rise to dominance in this period looks very likely if not absolutely necessary.

13 Scholars have written several histories of conservatism in recent years. One of the better ones is Lisa McGirr, Suburban Warriors: The Origins of the New American Right (Princeton: Princeton University Press, 2002). 14 Harry S. Dent, one of the chief architects of the “Southern Strategy,” gives a fascinating recounting of the political maneuver in his memoir, The Prodigal South Returns to Power (New York: John Wiley & Sons, 1978). 15 The best work on how the Republicans brought together these two forms of conservatism is Kiron K. Skinner, Serhiy Kudelia, Bruce Bueno de Mesquita, and Condoleezza Rice, The Strategy of Campaigning: Lessons from Ronald Reagan and Boris Yeltsin (Ann Arbor: University of Michigan Press, 2007).

390

Furthermore, an anti-expert, anti-government mindset was developing along with—although not always connected to—this growing conservatism. In

1965, former leftist intellectuals Irving Kristol and Daniel Bell formed their journal The Public Interest, which made regular sport of Johnson’s Great Society.

In their view, the experts in the Great Society were cast adrift without any real notion of how to accomplish their goals. Some members of Great Society efforts later came to agree. In 1969, Daniel Patrick Moynihan, the former Assistant

Secretary of Labor and Harvard professor, published his book, Maximum Feasible

Misunderstanding, a title that neatly summarized his views on the failures of the

Great Society’s War on Poverty.16 Experts in government, Moynihan argued, were unable to make any strides towards their goal. This disbelief in experts clearly connected to the anti-establishment rhetoric of libertarians, hippies, and the New Left. During the late 1960s and early 1970s, people developed distrust in experts, even as federal politicians continued to expand agencies based on expertise. Vietnam, the Pentagon Papers, Watergate—the news of the day cast doubt on the legitimacy of authority.17 Moreover, tax revolts were beginning to develop around the nation, particularly in California’s Proposition 13, and people worried that a federal tax revolt was next.18 And the middle class had

16 Daniel Patrick Moynihan, Maximum Feasible Misunderstanding: Community Action in the War on Poverty (New York: Free Press, 1969). 17 Mark Lilla has recently argued that this distrust of authority has found its successors both in the right (the Tea Party Movement) and in the left (for instance, New Age spiritualists who eschew immunizations and take largely unproven dietary supplements). Both sides are connected to a hard core of distrust. And although these two political factions might rhetorically distance themselves from each other, they share more than they do not. Mark Lilla, “The Tea Party Jacobins,” The New York Review of Books, May 27, 2010. 18 The best study of the California tax revolt remains David O. Sears and Jack Citrin, Tax Revolt: Something for Nothing in California (Cambridge, Mass.: Harvard University Press, 1982).

391 come to associate welfare programs with poor, non-working loafers. As the sociologist Monica Prasad has shown, neoliberal policies took off in countries where welfare was associated with the poor (such as the United States and the

United Kingdom) but failed to take hold in countries where the middle-class believed that they were receiving benefits from welfare programs (such as

France and Germany).19 In the United States the message on this issue was clear, as Reagan continued his weekly radio addresses condemning the “welfare queen.” From all sides, experts and government programs were under attack.20

Ronald Reagan was neither an economist, nor a philosopher, though he may have played one in a film. Although Reagan would award Milton Friedman, one of the central members of the Chicago School of Economics, the Presidential

Medal of Freedom in 1988, Reagan came to neoliberal policies through a more pedestrian route than reading books published by Friedman and his colleagues.

When Reagan was governor of California, he had—albeit without much excitement—backed tightening the state’s automotive emissions control regulations, which were more stringent than the federal ones. It was what the people wanted. But during the late 1970s, Reagan increasingly embraced “free market” ideas, including deregulation, that were coming into vogue. Reagan

19 Monica Prasad, The Politics of Free Markets: The Rise of Neoliberal Economic Policies in Britain, France, Germany, and the United States (Chicago: University of Chicago Press, 2006). 20 Other works that have tracked distrust in experts and the technology that experts produced include Brian Balogh, Chain Reaction: Expert Debate and Public Participation in American Commercial Nuclear Power, 1945–1975 (Cambridge: Cambridge University Press, 1991), 17–18; Thomas Parke Hughes, American Genesis: A Century of Invention and Technological Enthusiasm, 1870–1970 (New York: Viking, 1989), ch. 9; Carroll Pursell, “The Rise and Fall of the Appropriate Technology Movement in the United States, 1965–1985,” Technology and Culture, Vol. 34, No. 3 (July 1993), 629–637.

392 headed in a decidedly neoliberal direction when he took Congressman Jack

Kemp into his camp of advisors.

Kemp, a Representative from a congressional district that included

Buffalo, New York, had been considering a presidential run, but he agreed to give up his bid if he had a seat in Reagan’s inner circle. The Chicago School had deeply influenced Kemp and his friend, David Stockman, a fellow Congressman who became the head of the Office of Management and Budget under Reagan. But in his 1979 book An American Renaissance: A Strategy for the 1980s, Kemp focused not on the thoughts of Friedman, Stigler, Gary Becker, or Richard Posner, but on

Arthur Laffler, a bit player in the Chicago School who held a position at the

University of Chicago School of Business. If Laffler was not an essential member of the economic school of thought, he made up for it by playing an important part in influencing both the Ford and Reagan presidencies. Laffler’s central contribution to the discourse of the day was the so-called “Laffler Curve,” which held that lowering the tax rate could actually increase tax revenues because of economic growth and new entry by firms. Jude Wassinski, a former editor at The

Wall Street Journal, coined the phrase “Laffler Curve” after sitting in on a discussion with Laffler and others. Wassinski had also popularized the term

“supply-side economics,” a notion that flew in the face of the dominant

Keynesian philosophy of the day, which held that politicians could aid economy by spurring demand (the “demand-side”) through deficit spending and other measures. Supply-side partisans believed that economic growth would only come if politicians removed barriers to new firm entry or existing firm

393 expansion, especially taxes, labor unions, and regulations. Kemp brought these ideas into the Reagan camp. And when Reagan came to the White House, he appointed David Stockman to head the Office of Management and Budget, the agency that acts as a president’s chief tool for controlling and reining in other agencies. Stockman’s mandate was clear: undo the regulatory state and cut taxes.

Less than a month in office, on February 17, 1981, Reagan issued

Executive Order 12291, which effectively killed all pending regulations, forced agencies to review existing ones, and imposed new burdens—cost-benefit analyses, impact analyses—on any proposed rules. The political scientist Stan

Luger has estimated that this Executive Order affected more than thirty-four potential regulations of the automobile, including passive restraints, tighter emissions controls, and more stringent fuel economy standards.21

The domestic auto industry in the United States was in a horrible condition when Reagan came to power. The academic and popular press regularly compared the domestic automakers to the Japanese firms in a negative light. Academic centers around the nation, especially MIT and Harvard, began to carry out extended studies of the Japanese method of auto production that would produce important—and sometimes widely read—books later in the decade.22 Fear of Japanese dominance would eventually express itself in such

21 Luger, Corporate Power, 117. 22 See, for instance, James P. Womack, Daniel T. Jones, and Daniel Roos, The Machine That Changed the World (New York: Rawson Associates, 1990). For a popular treatment of how Japanese cars got a foothold in the United States, see David Halberstram, The Reckoning (New York: William Morrow and Company, Inc., 1986).

394 works of popular culture as the film Gung Ho (1986) and the Michael Crichton novel Rising Sun (1992). Things had not yet reached that fever pitch, however in the early 1980s. The Reagan administration often sold deregulation as

“regulatory relief,” arguing that one reason that industries, including the automobile manufacturers, were struggling was because of the burdens of regulations.

This notion had its dissenters, however. In 1978, William J. Abernathy, a professor who studied economic productivity and technological change at the

Harvard Business School, published his book, The Productivity Dilemma, which remains an important work on the history of technology in the auto industry.23

Abernathy did not believe with the supply-side economics crowd that tax policy and regulation had led to an industrial turn down. Rather, in a 1980 article co- written with a fellow Harvard Business School professor, Abernathy argued that

US firms had been, as the title proclaimed, “Managing Our Way to Economic

Decline.”24 Abernathy believed that firms had to increase long-term investment in “innovative” technologies in order to escalate their competitiveness.

Abernathy was certainly not wrong; the lack of imagination and vision on the part of managers of auto firms played a large part in their relative decline.

Ironically, however, nearly everything that Abernathy espoused, besides the occasional and potential promises of regulation, would become part of neoliberal orthodoxy, especially the focus on competitiveness, “innovation,” technology,

23 William J. Abernathy, The Productivity Dilemma: Roadblock to Innovation in the Automobile Industry (Baltimore: Johns Hopkins University Press, 1978). 24 Robert H. Hayes and William J. Abernathy, “Managing Our Way to Economic Decline,” Harvard Business Review (July-August 1980), 67–77.

395 and entrepreneurial vision. Like Kemp, Abernathy envisioned in industrial renaissance, as declared in the title of another of his co-written works, An

Industrial Renaissance: Producing a Competitive Future for America.25 Although

Abernathy distanced himself from—if not outright attacked—Reagonomics, the rising dominance of neoliberalism set, framed, and limited the terms of debate.26

Beyond issuing Executive Orders, Reagan also carried out his neoliberal vision through his political appointments. Often, he appointed undertakers— administrators whose chief role was to hamstring or tear apart the very agencies that they led. Anne M. Gorsuch, who came to head the EPA, was a prime example of this trend. Gorsuch had served a term in the Colorado House of

Representatives and was amazingly unqualified for the position. Yet, upon her arrival at the EPA, she began systematically to deconstruct it, cutting its budget by 22%, downsizing the staff, decreasing regulatory enforcement, and pulling back on important regulations, including pesticides. For the EPA staff members,

Gorsuch’s presence was extremely demoralizing. She eventually became so controversial with Congress and environmentalists that Reagan jettisoned her, bringing back William Ruckelshaus, the first EPA administrator, to his old post to restore the agency’s credibility.

25 William J. Abernathy, Kim B. Clark, and Alan M. Kantrow, Industrial Renaissance: Producing a Competitive Future for America (New York: Basic Books, 1984). 26 The rather limited debate between Abernathy and Reaganites is best viewed in contradistinction to the earlier works of Emma Rothschild and Ernest Mandel, both of whom viewed the industrial decline of the 1970s as part of a much larger experience of social and economic upheaval and adjustment. Rothschild, Paradise Lost: The Decline of the Auto-Industrial Age (New York: Vintage Books, 1974); Ernest Mandel, The Second Slump: A Marxist Analysis of Recession in the Seventies, trans. Jon Rothschild (London: New Left Books, 1978).

396

Surprisingly, every member of the EPA’s Office of Mobile Sources whom I have interviewed has told me that, by good fortune, the Ann Arbor office was rather insulated from these political upheavals. The office’s geographical isolation from Washington, D.C., might provide one explanation for this experience. One staff member told me that lawmakers also simply liked visiting the Ann Arbor office. The materiality of emissions testing impressed them and gave them a concrete sense that real work was being done there. Regardless of the reason, the Ann Arbor staff members of the Office of Mobile Sources did not sweat the coming of Reagan, unlike many of their Washington-based colleagues.27

Ironically, during this period, Eric Stork, the former head of the Office of

Mobile Sources, headed in a decidedly neoliberal direction. Throughout his work at the EPA, Stork began to become wary of the costs that regulations forced on producers and consumers. When he left the EPA in 1978 to act as a visiting professor at Purdue University, he began to become convinced of the wisdom of establishing a regulatory budget, a form of cost-benefit analyses that would cap the costs any agency could impose on society. With a regulatory budget in mind, he took aim at the certification program of the Office of Mobile Sources, the program that he had led for nearly eight years. During Stork’s stay at Purdue, he had begun looking into emissions control programs in other countries. The

27 One exception to this rule would be Charles Gray, who was discussed in Chapter 4. By the late- 1970s, Gray had been put in charge of creating new automotive emissions control and fuel regulations. He was known for being very tough on industry. When Reagan came to power, Gray moved out of this regulatory position and began his long career of doing R&D at the EPA. In the early 1980s, Gray focused primarily on fuels research, particularly the production of methanol, and he co-wrote a book on the subject. By the 1990s and the coming of the Clinton administration, Gray had moved on to the so-called Supercar project.

397 programs in Australia and Sweden left the biggest imprint on him. Neither country had a certification program to speak of, but both had automotive emissions control programs that were as successful as the one in the United

States. Stork began to believe that the certification program was misguided: it had many upfront costs, and it could not guarantee that catalytic converters could function for the required time (50,000 miles). A better and more efficient program would be not to certify vehicles, but to test them randomly as vehicles in-use, and levy the automakers with draconian fines if their vehicles failed in- use tests. In Stork’s view, this program would be smaller, cheaper, and more effective than the current certification program. Nothing came of Stork’s vision, and the certification program continues in force..

The Office of Mobile Sources may have experienced relative quiet during the Reagan administration, but nothing could have been farther from the truth for National Highway Traffic Safety Administration (NHTSA). Reagan appointed

Raymond A. Peck, Jr., to head the agency. Peck was an undertaker. Before coming to NHTSA, he had been the director of regulatory affairs for the National Coal

Association, a trade association that lobbied heavily against all forms of regulation that affected the coal industry. He spent most of his time before the coal association working as a lawyer for several companies. Peck brought

NHTSA to a standstill.

A person who worked at NHTSA during this time told me that the agency’s rule-making office became a “ghost town.”28 Since all rule-making

28 I am withholding this person’s name to protect this person’s identity.

398 efforts were halted, staff members spun their wheels, demoralized. Joan

Claybrook co-wrote two books on what was happening at NHTSA after her departure. As she and her co-authors noted, “Internal reorganizations and re- assignments have resulted in the loss of more than 200 NHTSA staff, a quarter of its workforce.”29 Moreover, in contrast to Claybrook’s consumer information campaign, under Peck, the agency fell under a “veil of secrecy.”30 Perhaps most important, since federal fuel economy standards were making cars smaller but new regulations had died on the vine, cars would be less resistant to crashes without new safety systems to defend them.31 Probably the most essential standard to fall by the wayside during Reagan years was the side impact standard, which had to wait more than a decade to become a reality.

Yet, while Claybrook’s fury against Reagan’s anti-government destruction often hit its mark, it occasionally missed. Her polemic against pro-seatbelt-use programs was perhaps the clearest example of her errors. In Reagan on the

Road: The Crash of the U.S. Auto Safety Program, Claybrook and her co-authors wrote:

“The cornerstone of the administration’s activities is a $9.6 million safety belt campaign. The agency’s ‘unofficial’ goal is a 20% increase in belt usage. Even if attained it will still leave 70% of all occupants unprotected. Automatic crash protection [air bags and other forms of so-called “passive restraints”] would achieve this

29 Joan Claybrook, Jacqueline Gillan, and Anne Strainchamps, Reagan on the Road: The Crash of the U.S. Auto Safety Program (Washington, D.C.: Public Citizen, 1982), ii. Gillan went on to become an important auto safety advocate. Strainchamps has achieved relative fame as an interviewer on Wisconsin Public Radio’s nationally-syndicated program, To the Best of Our Knowledge. 30 Ibid. 31 Ibid., iv.

399

goal in about three years and would protect all occupants in 10 years. In the meantime: • The administration is making no improvements in the reliability of safety belts or the comfort of using them despite studies showing that most people do not wear belts because they are uncomfortable and inconvenient. • No efforts are being made to enhance public awareness and use of existing passive seat belts. • Funds have been diverted from critical program areas to finance a safety belt campaign that past research, safety experts, members of Congress, and the insurance industry views as dubious.”32

Claybrook was probably right about the importance of automatic systems, like air bags, but she was dead wrong about seat belts. Her bias reflected the belief going back to the National Highway Safety Bureau’s first administrator, William

Haddon, that changing technology was more important than changing behavior.

One former NHTSA staff member, who worked for the agency from the 1970s through the 1990s, told me that the agency’s single-minded focus on technological approaches to safety in lieu of any sustained attention to driver behavior was his greatest regret.

During the 1980s, NHTSA began to focus increasingly on driver behavior.

In the early 1980s, the agency established a new office that was dedicated to behavioral issues including belt use and drunk driving. The office had a three- tiered approach: first, NHTSA, in conjunction with contractors, began grass- roots public education campaigns for seatbelt use and the famous Vince and

Larry crash test dummy public service announcements, which encouraged drivers and passengers to “buckle up.” These “spokesdummies” eventually became so popular and effective that NHTSA put them into semi-retirement,

32 Ibid., 5.

400 fearing that overexposure might diminish their effectiveness. Second, in concert with Mothers Against Drunk Driving, NHTSA also pushed for stronger drunk driving laws, which had important impacts around the nation. Third, through a work of complex political maneuvering, Elizabeth Dole, who became the

Secretary of Transportation in 1983, created a rule that set an ultimatum for industry: if states, which amounted to 2/3 of the nation’s population, had not adopted mandatory seatbelt laws by 1989, then automakers would be forced to put “passive restraints,” such as air bags or automated seatbelts, onto their vehicles.33 If the states did pass the laws, the passive restraint rule would be rescinded. This policy effectively made passing mandatory seatbelt laws a private-public endeavor. The Department of Transportation promised $40 million towards seatbelt education and lobbying for mandatory state seatbelt laws; the automakers, which did not desire mandated passive restraints, pledged millions more. Educational kits, which explained the benefits of seatbelt use, soon began entering schools and community centers. In 1984, even before Dole enunciated the policy, New York state created a mandatory seatbelt law, becoming the first state to do so.34 By 1985, nine more states had followed; by the next year, the total of states with seatbelt laws reached twenty-three. Finally, by 1990, thirty-seven states had laws. While only around 11% of people had used seatbelts in 1980, belt-use hovered around 50% by 1990. To summarize, though Joan Claybrook had disparaged NHTSA’s move towards behavioral

33 John D. Graham, Auto Safety: Assessing America’s Performance (Dover, Mass.: Auburn House Publishing Company, 1989), 180–181. 34 The following figures are from National Highway Traffic Safety Administration, Motor Vehicle Safety 1990 (Washington, D.C.: Department of Transportation, March 1992).

401 interventions, especially its focus on seatbelts, these efforts had huge beneficial effects.

In his insightful dissertation , “Systems of Restraint: Redistributing

Responsibilities for Automobile Safety in the United States since the 1960s,”

Jameson Wetmore examines how society “distributes responsibility” to different parties.35 He argues that, before the passage of the Traffic Safety Act of 1966, responsibility for auto safety fell squarely on the shoulders of drivers.

Afterwards, the law and the NHSB/NHTSA redistributed responsibility to the automakers, who became responsible for vehicle safety through federal performance standards. Later, in the 1980s, Wetmore argued, responsibility returned to drivers.36 Although Wetmore’s description is, broadly speaking, accurate, he misses the connection between these changes and larger shifts in politics and the economy. Responsibility may have fallen once again on the shoulders of drivers under the Reagan administration, but this development fit with the positive, individualist notion of neoliberalism.

Since the 1960s, the Chicago School economist Gary Becker had been framing crime as a rational choice.37 The criminal, Becker argued, was constantly weighing his or her options between certain risks and rewards. The goal of law, therefore, was to set a crime’s punishment at a high enough level

35 Jameson Michael Wetmore, “Systems of Restraint: Redistributing Responsibilities for Automobile Safety in the United States since the 1960s,” (PhD diss., Cornell University, 2003). 36 Wetmore’s accurate description of the distribution of responsibility aligns roughly with Stan Luger’s periodization of auto regulation. Luger breaks the era into Corporate Political Hegemony (1916–1966), the Politics of Compromise (1967–1978), the Resurgence of Corporate Power (1979–1981), and the Triumph of Corporate Power (1981–1988). See Luger, Corporate Power. 37 For Michel Foucault’s description of how Becker’s theory of criminality fit within neoliberalism, see Foucault, The Birth of Biopolitics: Lectures at the College de France, 1978–1979, trans. Graham Burchell (New York: Palgrave Macmillan, 2008), esp. 250–260.

402 that it would outweigh the reward of perpetrating it. Seatbelt laws fit this model.

If the act of going without a seatbelt was ruled illegal, then drivers would change their behavior. Increased criminalization accompanied the rise of neoliberalism.

But on a wider level, the seatbelt laws also fit the model of self-deterimining individuals that neoliberals tried to engender—people who would take care of themselves. The political scientist Barbara Cruikshank has described how the notion of “self-esteem,” or valuing one’s self, gained increased currency during this period.38 Similarly, the Vince and Larry crash test dummies were aimed at fostering drivers and passengers who cared for themselves.

Perhaps, the highly respected public health expert, John D. Graham, went too far then, when he wrote in 1989, “By itself [NHTSA] did not save the airbag technology. A new generation of progressive executives at Ford Motor Company followed the European lead of Mercedes-Benz and turned to the airbag in their search for long-run competitive advantage. . . . At Ford, airbags . . . are now elements of long-run corporate strategy to build Ford’s reputation as the manufacturer that supplies a quality product that customers learn to appreciate.”39 Graham’s point is that, ultimately, private industry and the

“market” for safety succeeded in getting airbags into cars, whereas government had failed. But this “market” was not virgin land. NHTSA had been trying for several years to build safety-consciousness through seatbelt laws, public

38 Barbara Cruickshank, “Revolutions Within: Self-Government and Self-Esteem” in Foucault and Political Reason: Liberalism, Neo-Liberalism, and Rationalities of Government, Andrew Barry, Thomas Osborne, and Nikolas Rose, eds. (Chicago: The University of Chicago Press, 1996), 231– 252. 39 John D. Graham, Auto Safety: Assessing America’s Performance (Dover, Mass.: Auburn House Publishing Company, 1989), 228.

403 information campaigns, and the Vince and Larry public service announcements.40 Historians of business and advertising are always cautious to move from “prescriptive” objects, including advertisements, etiquette books, pamphlets that producers send to consumers in order to inform them, and the like, to descriptions of behavior—and for good reason. We have very little idea how these instruments of suggestion, prescription, and instruction actually affect people’s actions, at least at the societal, “in-the-wild” level.41 But we can certainly suggest NHTSA’s campaigns played some role in generating consumer demand—enough of a role that the rise of interest in the airbag is more complex than the efforts of a few “progressive executives.”

Today, the automobile may be the most heavily regulated consumer product in the United States. Some may question if there is much more to do to the automobile to make it better. Many cars now have between four and six airbags; some have more than ten. How much more safety can we want from a car? As Keith Bradsher has described, the rise of the SUV (sports utility vehicle) in the late-1980s and 1990s made this question important, given SUVs greatly damaged smaller cars in crashes and that, as studies suggested, SUV drivers were more likely to take part in risky driving. The question also had an important class dimension because the affluent bought SUVs, while the poor

40 Moreover, Graham fails to account for the general increased perception of risk and safety (which is extremely difficult to account for) that some call “risk society.” See Ulrich Beck, Risk Society: Towards a New Modernity, trans. Mark Ritter (London: Sage Publications, 1992). 41 Of course, the fields of psychology, marketing, and, more recently, behavioral economics have carried out laboratory experiments on the issue (mostly on captive undergraduate students).

404 tended to own smaller cars. In automotive emissions control, after many delays, the automakers finally reached the 1970 Muskie Act’s 90% reduction standards in 1981. The Clean Air Act amendments passed under George H. W. Bush in

1990 made emissions controls even more stringent and required new vapor recovery systems that captured evaporative emissions during refueling. (Before vapor recovery systems were installed on vehicles and gasoline pumps, evaporative emissions would simply release into the ambient air whenever someone refueled. Studies found that these evaporative emissions made significant contributions to overall air pollution and threatened to wipe out gains from catalytic converters.) Still, some former employees of EPA’s Office of

Mobile Sources and NHTSA have wondered aloud to me why the agencies still exist in their full capacity. Is the work of these agencies done?

One former NHTSA employee told me that the science of crashworthiness is very well understood, and though we will continue to make incremental improvements in vehicle crashworthiness, the real future of automotive safety is in the science of human perception and attention. Different automakers have been experimenting with and implementing systems that set off an alarm when a vehicle is crossing a solid painted line, when a vehicle is in reverse and something is in its path, and when a vehicle is cruising and other cars in front of it are beginning to brake. NHTSA has considered mandating some of these systems.

In automotive emissions control, the future likely rests in reducing greenhouse gas emissions (GHGs), primarily carbon dioxide (CO2). The most

405 probable tool for attaining reduced GHGs in cars is federal fuel economy standards. Because reducing overall automotive fuel use necessarily entails a reduction of carbon emissions, fuel economy standards will play a likely role in any significant climate change policy, if one is ever passed. Many foresee electric automobiles as the solution to carbon emissions. But battery-powered cars will pose their own safety risks, including poor crashworthiness. More distant, hydrogen-powered automobiles will surely require crashworthiness standards and regulation.

Regardless of the future of automotive regulation, one thing is sure: the performance standard as a means for reducing targeted technological risks will remain in policymakers’ toolbox. As policymakers continue to consider ways of averting the potentials of disastrous climate change, the performance standard suggests itself for several uses. When we reflect upon the history of automotive regulations and the performance standards that undergirded them—in ways both positive and negative—we can learn from the varied chapters in the history of the US government’s regulatory management of the automobile, 1966–1988.

Bibliography

Primary Sources

Archives and Other Depositories

National Archives and Records Administration, National Archives at College Park, MD (NARA II).

Records of the Environmental Protection Agency (Record Group 412). Records of the General Services Administration (Record Group 269). Records of the National Highway Traffic Safety Administration (Record Group 416).

Library of the Department of Transportation, Department of Transportation Building, Washington, D.C.

Reading Room of the National Highway Traffic Safety Administration, Department of Transportation Building, Washington, D.C. (The Reading Room is now defunct and has been merged with the Department of Transportation Library.)

Library of the National Vehicle and Fuel Emissions Laboratory, Environmental Protection Agency, Office of Transportation and Air Quality, Ann Arbor, MI.

Private Papers of Eric O. Stork, Arlington, VA.

Private Papers of Karl Hellman, in the author’s possession.

Government Documents

Bureau of Public Roads and Aerospace Research Associates, “A Reusable Energy Absorbing Highway Protective System for Median Areas,” Bureau of Public Roads Report #96, 1968.

Committee on Interstate and Foreign Commerce, “Summary of Major Provisions of Energy Conservation and Oil Policy Act of 1975 (H.R. 7014) as Passed the House,” September 26, 1975.

Committee on Motor Vehicle Emissions, National Academy of Sciences, Semiannual Report to the Environmental Protection Agency, January 1, 1972; idem., Interim Standards Report, April 26, 1972.

Committee on Motor Vehicle Emissions, National Academy of Sciences, Semiannual Report to the Environmental Protection Agency, February 12, 1973.

406 407

Comptroller General of the United States, “Review of the Automobile Fuel Economy Testing and Labeling Program,” Report to the Subcommittee on Conservation and Natural Resources, Committee on Government Operations, House of Representatives, August 15, 1974.

Department of Health, Education, and Welfare, Report of the Secretary’s Advisory Committee on Traffic Safety, February 29, 1968.

Department of Transportation and Environmental Protection Agency, “Potential for Motor Vehicle Fuel Economy Improvement,” Report to Congress, October 24, 1974.

Emission Control Technology Division, Office of Mobile Sources, “Automobile Sulfuric Acid Emission Control—The Development Status as of December 1975: A Report to the Administrator, EPA.” http://nepis.epa.gov/EPA/html/newtitles.html. (Last accessed February 24, 2011).

Environmental Protection Agency, “Evaluation of the LaForce-Modified AMC Hornet,” December 1974.

Environmental Protection Agency, “Fuel Economy and Emission Control,” November 1972.

Environmental Protection Agency, “Health Consequences of Sulfur Oxides: A Report from Chess, 1970–1971,” May 1974.

Environmental Protection Agency, “A Study of Fuel Economy Changes Resulting from Tampering with Emission Control Devices,” January 1974.

Federal Highway Administration, Press Release, December 22, 1969.

Government Accounting Office, “Examination into the Adequacy of the Environmental Protection Agency’s Motor Vehicle Certification Activities,” June 12, 1972.

National Academy of Sciences, Semiannual Report by the Committee on Motor Vehicle Emissions of the National Academy of Sciences to the Environmental Protection Agency, January 1, 1972.

National Center for Health Statistics, Weight, Height, and Selected Body Dimensions of Adults, United States, 1960–1962 (Washington: U.S. Dept. of Health, Education and Welfare, Public Health Service, 1965).

408

National Highway Traffic Safety Administration, Motor Vehicle Safety 1990 (Washington, D.C.: Department of Transportation, March 1992).

Senate Committee on Public Works, Decision of the Administrator of the Environmental Protection Agency Regarding Suspension of the 1975 Auto Emissions Standards, 93rd Congress, 1st sess., April 16-18, 1973.

Senate Committee on Public Works, Highway Safety Act: Hearings on S. 3052, 89th Cong., 2nd sess., 1966.

United States Department of Commerce, National Traffic Safety Agency, “Report on the Development of the Initial Federal Motor Vehicle Safety Standards Issued January 31, 1967,” March 17, 1967, NHTSA Microfilm, Reel One, ii.

U.S. Department of Health, Education, and Welfare, Report of the Secretary’s Advisory Committee on Traffic Safety (Washington, DC: US Government Printing Office, 1968).

United States Senate, Special Subcommittee on Science, Technology, and Commerce, “Hearings on The National Fuel Economy Testing Act of 1974,” May 17, 1974.

Williams, Marcia, et al. Automobile Exhaust Emission Surveillance of the FY 72 Program, U.S. Environmental Protection Agency, Office of Mobile Source Air Pollution Control, February 1974.

Federal Register Notices

Federal Register, Vol. 30 (1965), 8319;

Federal Register, Vol. 31 (1966), 4088.

Federal Register, Vol. 32 (February 3, 1967), 2408–2414.

Federal Register, Vol. 32 (November 5, 1969), 8808.

Federal Register, Vol. 38 (May 2, 1973), 10867–10908.

Federal Register, Vol. 38, No. 165 (August 27, 1973), 22944–22947.

Federal Register, Vol. 39 (September 24, 1974), 34382–34384.

Federal Register, Vol. 40 (September 10, 1975), 42003.

409

Articles, Conference Proceedings, Dissertations, Theses, and Books

Akins, James E. “The Oil Crisis: This Time the Wolf is Here,” Foreign Affairs, Vol. 51, No. 3 (April 1973), 462–490.

Barth, Delbert S. “Federal Motor Vehicle Emission Goals for CO, HC, and NOx Based on Desired Air Quality Levels,” Air Pollution Control Administration Journal, Vol. 20, No. 8 (August 1970), 519–523.

Bazzell, R. C. “The SAE Manikin Interior Space Evaluator,” a paper given at the National Automobile Meeting, Detroit, MI, March 19-21, 1963, SAE Paper 630176.

Boelter, Llewellen M. K. “Comments on the Theory of Planning,” Paper presented at the National Conference of American Society of Civil Engineers, Los Angeles, Feb. 1959.

Brenner, Robert. “Analysis of Traffic Flow Using ‘Time’ as the Dependent Value” (PhD. diss., University of California Los Angeles, 1962).

Brenner, Robert. “The Future of Highway Research,” Highway Research Record, No. 79, 1965, 58-9.

Brenner, Robert. “Automobile Design and Safety—Technology Tradeoffs and Transfer Processes” in Space Technology Transfer to Community and Industry, eds. Ralph H. Tripp and John K. Stotz, Jr. (Tarzana, CA: AAS Publications Office, 1972), 173.

Claybrook, Joan, Jacqueline Gillan, and Anne Strainchamps. Reagan on the Road: The Crash of the U.S. Auto Safety Program (Washington, D.C.: Public Citizen, 1982).

Coordinating Research Council. “A Survey of Average Driving Patterns in the in the Six Urban Areas of the United States: Summary,” CRC Study CAPE-10.

Esposito, John C. Vanishing Air: Ralph Nader’s Study Group Report on Air Pollution (New York: Grossman Publishers, 1970).

Franchini, Enzo. “Crash Testing Evolution at Fiat,” February 1, 1966, SAE Paper 660165.

Gray, Charles L. and Jeffrey A. Alson. Moving America to Methanol: A Plan to Replace Oil Imports, Reduce Acid Rain, and Revitalize Our Domestic Economy (Ann Arbor: University of Michigan Press, 1985).

410

Gurdjian, E. S. “Experiences in Head Injury and Skeletal Research,” Proceedings of the Tenth Stapp Car Crash Conference (November 8–9, 1966), 145-157.

Haagen-Smit, A. J. “The Air Pollution Problem in Los Angeles,” Engineering and Science, Vol. 14 (December 1950), 1–7.

Haddon, William, Edward E. Suchman, and David Klein. Accident Research: Methods and Approaches (New York: Harper and Row, Publishers, 1964).

Haddon, William. “The Changing Approach to the Epidemiology, Prevention, and Amelioration of Trauma: The Transition to Approaches Etiologically Rather than Descriptively Based,” American Journal of Public Health, Vol. 58. No. 8, (August,1968), 1433.

Hansen, A. M. ”SAE Test Procedure for Instrument Panels: An Impact Pendulum for Testing Safety Characteristics of Padded Instrument Panels and Other Passenger Compartment Components,” October 10, 1965, SAE Paper 650963.

Hanson, Harlan L. “Energy Absorbing Steering Column: A Case History,” Proceedings of the Tenth Stapp Car Crash Conference (November 8–9, 1966).

Hass, G. C., M. P. Sweeney, and J. N. Pattison. “Laboratory Simulation of Driving Conditions in the Los Angeles Area,” SAE Paper 660546, presented at SAE West Coast Meeting, Los Angeles, August 1966.

Howe, R. T. “Can the Urban Transportation Problem Be Solved?” Traffic Quarterly Vol. 14 (Jan. 1960), 85-95.

Huelke, Donald F. “Biomechanical Studies on the Bones of the Face,” Impact Acceleration Stress, National Academy of Sciences, Space Science Board, Man in Space Committee, ed. (Washington, D.C.: National Academy of Sciences, 1962), 131-133.

Huls, Thomas A. “Evolution of Federal Light-Duty Mass Emission Regulations,” February 1, 1973, SAE Paper 730554.

Kaptur, Vincent D. and Michael C. Myal. “The General Motors Comfort Dimensioning System,” paper presented at SAE International Congress of Automotive Engineering, Detroit, January 1961, SAE Paper 610174.

Kruse, Ronald E. and Thomas A. Huls. “Development of the Federal Urban Driving Schedule,” February 2, 1973, SAE Paper 730553.

Moynihan, Daniel Patrick. “Epidemic on the Highways,” The Reporter, Vol. 20, No. 9 (April 1959), 16–23.

411

Nader, Ralph.“Taming the Corporate Tiger” in The Ralph Nader Reader (New York: Seven Stories Press, 2000), 136.

Nader, Ralph. Unsafe at Any Speed (New York: Pocket Books, 1966).

Nixon, Richard M. “Special Message to the Congress on Energy Policy,” April 18, 1973. Available online, as part of The American Presidency Project, at http://www.presidency.ucsb.edu/ws/index.php?pid=3817&st=&st1 (Last accessed February 15, 2011).

Pitofsky, Robert. “Beyond Nader: Consumer Protection and the Regulation of Advertising,” Harvard Law Review, Vol. 90, No. 4 (Feb. 1977), 661–701.

Riesman, David, Reuel Denney, and Nathan Glazer, The Lonely Crowd: A Study of the Changing American Character (New Haven: The Yale University Press, 1950).

Rothschild, Emma. Paradise Lost: The Decline of the Auto-Industrial Age (New York: Vintage Books, 1974).

Society of Automotive Engineers. Vehicle Emissions (Selected SAE Papers), (New York: SAE, 1964).

Society of Automotive Engineers. Automotive Fuel Economy (Warrendale, PA: Society of Automotive Engineers, 1976).

Stahman, Ralph C., Kenneth D. Mills, and Merrill W. Korth. “Federal Air Pollution Efforts, the Early Years,” February 1, 1989, SAE Paper 890814.

Stevens, R. K. , P. J. Lamothe, W. E. Wilson, J. L. Durham, and T. G. Dzubay, eds. The General Motors/Environmental Protection Agency Sulfate Dispersion Experiment (Research Triangle Park, NC: Environmental Protection Agency, 1976).

Stivender, Donald L. “Development of a Fuel-Based Mass Emission Measurement Procedure,” SAE Paper 710604. The paper was delivered at the SAE’s Mid- Year Meeting, held in Montreal, Canada, June 7–11, 1971.

Stork, Eric O. “The Genesis of the EPA Fuel Economy Program: An Exercise in Bureaucratic Gamesmanship.”

Stork, Eric O. “Living with Regulation,” Automotive Emissions Seminar, Environmental Activities Week, General Motors Corporation, Warren, Michigan, June 19, 1973.

412

Stork, Eric O. “The United States Experience with Imposing Automobile Emission Standards,” A Paper for the Australian Society of Automotive Engineers, Perth, Australia, September 22, 1976. von Braun, Wernher. “Earth Benefits from Space and Space Technology” in Space Technology Transfer to Community and Industry, eds. Ralph H. Tripp and John K. Stotz, Jr. (Tarzana, CA: AAS Publications Office, 1972).

Whyte, William H. The Organization Man (New York: Simon and Schuster, 1956).

William E. Wilson et al., “General Motors Sulfate Dispersion Experiment: Summary of EPA Measurements,” Journal of the Air Pollution Control Association, Vol. 27, No. 1 (January 1977), 46–51.

Wilson, Sloan. The Man in the Gray Flannel Suit (New York: Simon and Schuster, 1955).

Periodicals

Automotive News Automotive Industries Business Week Chicago Tribune Consumer Reports The Detroit News Environmental Health Letter Environmental Science and Technology IIHS Status Report Los Angeles Times The New York Review of Books New York Times Palo Alto Times Seattle Times SEC News Digest Time Wall Street Journal The Washington Post

Periodical Articles

“Auto Makers Sue on Safety Rules,” New York Times, April 1, 1967.

Bacon, Kenneth B. “Car Makers Soon May Face Ad Standards for Fuel-Mileage Claims, FTC Staff Says,” Wall Street Journal, April 15, 1974.

413

Camp, Charles and Walter Mossberg. “Sniffing at the Tailpipe: Fuss at Ford Shows How Weak Enforcers Are,” Wall Street Journal, June 6, 1972.

“Car Safety Rule Termed Illegal,” New York Times, Mar. 7, 1967.

Colburn, Donald A. “Washinton’s RX: Will the Cure Be Worse than the Disease?”, Automotive Industries, February 15, 1974.

“Consumer Reports Says Chevelle Failed to Pass Road Tests,” Wall Street Journal, January 23, 1968.

“Consumers Union Asks FTC for Uniform Test for Cars’ Fuel Mileage,” Wall Street Journal, April 11, 1974.

Cowan, Edward. “President Asks the Nation to Drive Smaller Cars,” New York Times, June 30, 1973.

“Do Auto Catalysts Meet the Safety Test?” Business Week, October 27. 1973, 116.

“Edward P. LaForce Convicted,” SEC News Digest, November 28, 1978.

“Environment: At Last, The Energy Message,” Time, April 30, 1973.

“Environmental Currents,” Environmental Science and Technology, Vol. 4, No. 2, 95.

“EPA’s Chief Asserts ’75 Autos Will Use Less Fuel Than ‘74s,” Wall Street Journal, March 26, 1974.

“EPA Still Holds Reins in Mileage Labeling,” Automotive News, May 27, 1974.

“EPA Tests Refute LaForce Engine Claim,” Chemical Engineering News, Vol. 52, No. 51 (1974), 5.

“EPA Would Rather Switch than Fight, Substitutes Stork for Moran,” Environmental Health Letter, July 15, 1975.

Evans, Rowland and Robert Novak, “Fuel Economy Testing: Undercutting the EPA,” The Washington Post, May 4, 1974.

“Fiat Will Pay U.S. $100,00 in Settlement of Civil-Penalty Claim,” New York Times, February 23, 1971.

Fisher, Dan and Paul E. Steiger. “Gas Mileage Tests by U.S. May Be Changed,” Los Angeles Times, April 25, 1974.

414

Flint, Jerry M. “Auto Makers Told to Remove Device,” New York Times, December 16, 1972.

Flint, Jerry M. “Test Fault Perils Sale of ’73 Fords,” New York Times, May 23, 1972.

“Ford Reassigns 4 in Faulty Tests,” New York Times, May 24, 1972.

“Ford ‘Reassigns’ Employees in Test Snarl, Hopes to Market ’73 Models on Schedule,” Wall Street Journal, May 24, 1972.

“Gasoline Guzzlers Unmasked by EPA Tests,” Chicago Tribune, September 19, 1973.

“Honda Civic Tops Lists in ’74 Auto Mileage Test,” New York Times, September 19, 1973.

Insurance Institute for Highway Safety, “NHTSA Advocate Dead at 37,” Status Report, Vol. 9, No. 14 (July 26, 1974), 6. The vast majority of the Insurance Institute’s Status Report newsletters are available online: http://www.iihs.org/sr/default.aspx (Last accessed December 20, 2010).

Irvin, Robert W. “U.S. Urges Keeping Anti-Smog Systems: You Won’t Save on Gas by Tinkering, Drivers Told,” The Detroit News, January 4, 1974.

“Junk Anti-Smog Devices, Save Gas, G.M. Head Urges,” The Seattle Times, November 15, 1973.

Kleene, Tom. “SAE to Design EPA Mileage Test: Engineers Face June Deadline,” Detroit Free Press, March 8, 1974.

Koshetz, Herbert. “Fiat Spurs Trade,” New York Times, June 27, 1971.

Lorraine, Miranda. “EPA Issues New Mazda Data: Rotary Engine Still Rates Low,” The Washington Post, April 12, 1974.

MacKay, Ned. “The EPA Sniffs Sulfuric Acid,” Palo Alto Times, March 12, 1975.

“Nixon’s Car Safety Chief Ready to Quit,” New York Times, July 9, 1970.

Rothschild, Emma. “What is the ‘Energy Crisis’?” The New York Review of Books, Vol. 20, No. 12 (July 1973).

Rugaber, Walter. “Trade Commission Assails Ads By Car Makers on Fuel Savings,” New York Times, August 1, 1974.

“Safety Standards on ’68 Cars Eased,” New York Times, August 13, 1967.

415

Salpukas, Agis. “Court Fines Ford $7-Million in Suit on Pollution Test,” New York Times, February 14, 1973.

Salpukas, Agis. “E.P.A. to Change Auto Fuel Data,” New York Times, May 30, 1974.

Walsh, Bryan. “California’s Clean-Air Slapdown,” Time, December 20, 2007.

Williams, Douglas. “EPA Admits Blunder on Mileage Figures,” Detroit Free Press, May 30, 1974.

Williams, Douglas. “EPA Jumps in on Gas Issue,” Detroit Free Press, January 18, 1974.

Williams, Douglas. “Foiling Smog Gear Can Hurt Mileage, EPA Test Confirms,” Detroit Free Press, January 31, 1974.

Williams, Douglas. “Ford Asks EPA Approval to Use Disputed Device,” Detroit Free Press, December 22, 1972.

Williams, Douglas. “’Gas Hog’ Tag Cost $100 Million—Mazda,” Detroit Free Press, February 25, 1974.

Williams, Douglas. “The Mazda Fuel Furor,” The Review of Southern California Journalism (May 1974).

Williams, Douglas. “Mileage Test Jurisdiction Up for Grabs,” Detroit Free Press, April 26, 1974.

Williams, Douglas. “Unhook Smog Gear? EPA Tests the Idea,” Detroit Free Press, January 18, 1974.

Wilson, George C. “EPA Wins Its Fight to Keep Car-Mileage Tests in Agency,” Washington Post, May 18, 1974.

Secondary Sources

Books and Articles

Abernathy, Kim B. Clark, and Alan M. Kantrow, Industrial Renaissance: Producing a Competitive Future for America (New York: Basic Books, 1984).

416

Abernathy, William J. The Productivity Dilemma: Roadblock to Innovation in the Automobile Industry (Baltimore: Johns Hopkins University Press, 1978).

Amidei, Federico Barbiellini. “Italian Investments in the United States— Contributions to a History,” European Business History Association, Université de Genève, 13-15 September 2007. Available at www.unige.ch/ses/istec/EBHA2007/papers/BarbielliniGoldstein.pdf (Last accessed March 10, 2011).

Arendt, Hannah. Eichmann in Jerusalem: A Report on the Banality of Evil (New York: The Viking Press, 1963).

Arora, Ashish, Andrea Fosfuri, and Alfonso Gambardella, Markets for Technology: The Economics of Innovation and Corporate Strategy (Cambridge, Mass.: The MIT Press, 2001).

Arrow, Kenneth. The Limits of Organization (New York: Norton, 1974).

Balogh, Brian. Chain Reaction: Expert Debate & Public Participation in American Commercial Nuclear Power, 1945–1975 (Cambridge: Cambridge University Press, 1991).

Banks, Jeffrey S. and Barry R. Weingast, , “The Political Control of Bureaucracies Under Asymmetric Information,” American Journal of Political Science, Vol. 36, No. 2 (May 1992), 509–524.

Basalla, George. The Evolution of Technology (Cambridge: Cambridge University Press, 1988).

Baumol, William J. “Entrepreneurship: Productive, Unproductive, and Destructive,” The Journal of Political Economy, Vol. 98, No. 5, Part 1. (Oct., 1990), 893-921.

Bentley, Arthur F. The Process of Government: A Study of Social Pressures (Chicago: University of Chicago Press, 1908).

Borg, Kevin. Auto Mechanics: Technology and Expertise in Twentieth-Century America (Baltimore: The Johns Hopkins University Press, 2007).

Bradsher, Keith. High and Mighty: SUVs—The World’s Most Dangerous Vehicles and How They Got That Way (New York: Public Affairs, 2002).

Brenner, M. S. “Technology Intelligence and Technology Scouting,” Competitive Intelligence Review, Vol. 7, No. 3 (1996), 20–27.

Brown, Mark B. Science in Democracy: Expertise, Institutions, and Representation (Cambridge, Mass.: The MIT Press, 2009).

417

Brownlow, Louis. The Passion for Anonymity: The Autobiography of Louis Brownlow (Chicago: University of Chicago Press, 1955).

Burke, James. Connections (Boston: Little, Brown, and Company, 1978).

Callon, Michel. “Society in the Making: The Study of Technology as a Tool for Sociological Analysis,” in The Social Construction of Technological Systems, Wiebe E. Bijker, Thomas P. Hughes, and Trevor Pinch (Cambridge, Mass.: The MIT Press, 1987).

Cavell, Stanley. “Austin at Criticism” in Must We Mean What We Say?: A Book of Essays (Cambridge: Cambridge University Press, [1968] 2002), 103.

Cohen, Wesley M. and Daniel A. Levinthal, "Absorptive capacity: A new perspective on learning and innovation," Administrative Science Quarterly, Vol. 35, No.1 (1990), 128-152.

Cohen, Wesley M. and Steven Klepper, “A Reprise of Size and R & D,” The Economic Journal, 106 (July, 1996), 925-951.

Crandall, Robert W. and John D. Graham, “The Effect of Fuel Economy Standards on Automobile Safety,” Journal of Law and Economics, Vol. 32, No. 1 (April 1989), 97–118.

Cruickshank, Barbara. “Revolutions Within: Self-Government and Self-Esteem” in Foucault and Political Reason: Liberalism, Neo-Liberalism, and Rationalities of Government, Andrew Barry, Thomas Osborne, and Nikolas Rose, eds. (Chicago: The University of Chicago Press, 1996), 231–252.

Dewey, Scott Hamilton. Don’t Breathe the Air: Air Pollution and U.S. Environmental Politics, 1945–1970 (College Stattion: Texas A&M Press, 2000).

Dupree, A. Hunter. Science in the Federal Government: A History of Policies and Activities to 1940 (Cambridge, Mass.: The Belknap Press of Harvard University Press, 1957).

Edgerton, David. The Shock of the Old: Technology and Global History Since 1900 (Oxford: Oxford University Press, 2007).

Ezrahi, Yaron. The Descent of Icarus: Science and the Transformation of Contemporary Democracy (Cambridge, Mass.: Harvard University Press, 1990).

418

Foucault, Michael. The Birth of Biopolitics: Lectures at the College de France, 1978– 1979, trans. Graham Burchell (New York: Palgrave Macmillan, 2008), esp. 250–260.

Foucault, Michel. Discipline and Punish: The Birth of the Prison, trans. Alan Sheridan (New York: Vintage Books, 1977).

Galison, Image and Logic: A Material Culture of Microphysics (Chicago: The University of Chicago Press, 1997).

Gerard, David and Lester Lave. “Experiments in Technology Forcing: Comparing the Regulatory Processes of US Automobile Safety and Emissions Regulations,” International Journal of Technology, Policy, and Management, Vol. 7, No. 1 (2007), 1–14.

Giddens, Anthony. Constitution of Society: Outline of a Theory of Structuration (Berkeley: University of California Press, 1986).

Giddens, Anthony. “Risk and Responsibility,” The Modern Law Review, Vol 62, No 1 (January 1999), 1–10.

Giedion, Siegfried. Mechanization Takes Command: A Contribution to Anonymous History (New York: W. W. Norton & Company, Inc., [1948] 1969).

Graham, John D. Auto Safety: Assessing America’s Performance (Dover, Mass.: Auburn House Publishing Company, 1989), 180–181.

Gusfeld, James R. The Culture of Public Problems: Drinking-Driving and the Symbolic Order (Chicago: University of Chicago Press, 1981).

Habermas, Jürgen. Legitmation Crisis, trans. Thomas McCarthy (Boston: Beacon Press, 1975).

Hilgartner, Stephen. Science on Stage: Expert Advice as Public Drama (Stanford: Stanford University Press, 2000).

Hirschmann, Albert O. The Strategy of Economic Development (New Haven: Yale University Press, 1958).

Hobbs, Caswell O. “Legal Issues in FTC Trade Regulation Rules,” Food, Drug, and Cosmetic Law Journal, Vol. 32 (1977), 414–422.

Holsworth, Robert D. Public Interest Liberalism and the Crisis of Affluence: Reflections on Nader, Environmentalism, and the Politics of Sustainable Society (Boston: G. K. Hall & Co., 1980).

419

Horkheimer, Max and Theodor Adorno. Dialectic of Enlightenment: Philosophical Fragments, trans. Edmund Jephcott (1944, Standford: Stanford University Press, 2002).

Horowitz, Daniel. Jimmy Carter and the Energy Crisis of the 1970s: The “Crisis of Confidence” Speech of July 15, 1979: A Brief History with Documents (Boston: Bedford/St. Martin’s, 2005).

David Hounshell, “Planning and Executing ‘Automation’ at Ford Motor Company, 1945–65: The Cleveland Engine Plant and Its Consequences” in Fordism Transformed: The Development of Production Methods in the Automobile Industry, Haruhito Shiomi and Kazuo Wada, eds. (Oxford: Oxford University Press, 1995), 49–86.

Hounshell, David A. and John Kenly Smith, Jr. Science and Corporate Strategy: Du Pont R&D, 1902–1980 (Cambridge: Cambridge University Press, 1988).

Hughes, Thomas Parke, ed. Changing Attitudes Toward American Technology (New York: Harper Row Publishers, 1975).

Hughes, Thomas Parke. American Genesis: A Century of Invention and Technological Enthusiasm, 1870–1970 (Chicago: University of Chicago Press, 2004).

Hunter, Louis C. Steamboats on the Western Rivers: An Economic and Technological History (Cambridge: Harvard University Press, 1949).

Jameson, Frederic. “The Vanishing Mediator; or, Max Weber as Storyteller” in The Ideologies of Theory, vol. 2 (Minneapolis: University of Minnesota Press, 1988).

Jasanoff, Sheila. Designs on Nature: Science and Democracy in Europe and the United States (Princeton: Princeton University Press, 2005).

Jasanoff, Sheila. The Fifth Branch: Science Advisers as Policymakers (Cambridge, Mass.: Harvard University Press, 1990), 102–104.

Jasanoff, Sheila, ed. States of Knowledge: The Co-Production of Science and the Social Order (New York: Routledge, 2004).

Jones, Charles O. “The Limits of Public Support: Air Pollution Agency Development,” Public Administration Review, Vol. 32, No. 5 (September/October 1972), 502– 508.

Kahn, Jonathan. Budgeting Democracy: State Building and Citizenship in America, 1890–1928 (Ithaca: Cornell University Press, 1997).

420

Kevles, Daniel J. The Physicists: The History of a Scientific Community in Modern America (New York: Alfred A. Knopf, 1978).

Keynes, John Maynard. The General Theory of Employment, Interest, and Money (New Dehli: Atlantic Publishers & Distributors Ltd., 2006 [1936]).

King, David C. Turf Wars: How Congressional Committees Claim Jurisdiction (Chicago: The University of Chicago Press, 1997).

Kintner, Earl W. and Christopher Smith. “The Emergence of the Federal Trade Commission as a Formidable Consumer Protection Agency,” Mercer Law Review, vol. 26 (1974-1975), 651–688.

Kline, Ronald R. Consumers in the Country: Technology and Social Change in Rural America (Baltimore: The Johns Hopkins University Press, 2000).

Kline, Ronald R. and Trevor Pinch, “Users as Agents of Technological Change: The Social Construction of the Automobile in the Rural United States,” Technology and Culture, Vol. 37 (October 1996), 763–95.

Krier, James E. and Edmund Ursin, Pollution and Policy: A Case Essay on California and Federal Experience with Motor Vehicle Air Pollution, 1940–1975 (Berkeley: University of California Press, 1977).

Kubler, George, The Shape of Time: Remarks on the History of Things (New Haven: Yale University Press, 1962).

Lakwete, Angela. Inventing the Cotton Gin: Machine and Myth in Antebellum America (Baltimore: Johns Hopkins University Press, 2003).

Latour, Bruno. Aramis, or the Love of Technology (Cambridge, Mass.: Harvard University Press, 1996).

Lee, Jaegul and Francisco Veloso, “Inter-Firm Innovation under Uncertainty: Empirical Evidence for Strategic Knowledge-Partitioning,” Journal of Product Innovation Management, Vol. 25., Issue 5 (2006), 418–435.

Lee, Jongmin. “Regulatory Engineering in Pollution and Health Research: U.S. Environmental Protection Agency’s CHESS, 1969–1977,” presented at the annual meeting of the Society for the History of Technology, Tacoma, Washington, October 2, 2010.

Levi-Strauss, Claude. The Savage Mind (Chicago: University of Chicago Press, 1966).

Light, Jennifer S. From Warfare to Welfare: Defense Intellectuals and Urban Problems in Cold War America (Baltimore: Johns Hopkins University Press, 2003).

421

McCarry, Charles. Citizen Nader (New York: Saturday Review Press, 1972).

McCarthy, Tom. Auto Mania: Cars, Consumers, and the Environment (New Haven: Yale University Press, 2007).

McCraw, Thomas K. Prophets of Regulation: Charles Francis Adams, Louis D. Brandeis, James M. Landis, Alfred E. Kahn (Cambridge, Mass.: The Belknap Press of Harvard University Press, 1984).

Machery, Edouard. Doing without Concepts (Oxford: Oxford University Press, 2009).

Mandel, Ernest. The Second Slump: A Marxist Analysis of Recession in the Seventies, trans. Jon Rothschild (London: New Left Books, 1978).

March, James G. and Herbert A. Simon, Organizations (New York: Wiley, 1958).

Marcuse, Herbert. One-Dimensional Man: Studies in the Ideology of Advanced Industrial Society (Boston: Beacon Press, 1964).

Marcuse, Herbert. Eros and Civilization: A Philosophical Inquiry into Freud (Boston: Beacon Press, 1955).

Mashaw, Jerry L. and David L. Harfst, The Struggle for Auto Safety (Cambridge, Mass.: Harvard University Press, 1990).

Menand, Louis. The Metaphysical Club: A Story of Ideas in America (New York: Farrar, Straus, and Giroux, 2001).

Merrill, Karen R. The Oil Crisis of 1973–1974: A Brief History of Documents (Boston: Bedford/St. Martin’s, 2007).

Merton, Robert K. “The Normative Structure of Science” in The Sociology of Science: Theoretical and Emirical Investigations (Chicago: University of Chicago Press, 1973), 267-280.

Miller, K. D. and A. T. Arikan. “Technology Search Investments: Evolutionary, Option Reasoning, and Option Pricing Approaches,” Strategic Management Journal, Vol. 25, No. 5 (2004), 473–486.

Mokyr, Joel. The Gifts of Athena: Historical Origins of the Knowledge Economy (Princeton: Princeton University Press, 2002).

Nelson, Richard R. and Sidney G. Winter. An Evolutionary Theory of Economic Change (Cambridge, Mass.: Belknap Press of Harvard University Press, 1985).

422

Neustadt, Richard E. and Ernest R. May. Thinking in Time: The Uses of History for Decision-Makers (NewYork: The Free Press, 1986).

Oudshoorn, Nelly and Trevor Pinch, eds. How Users Matter: The Co-Construction of Users and Technology (Cambridge, Mass.: The MIT Press, 2003).

Pettigrew, Andrew M. “Information Control as a Power Resource,” Sociology, Vol. 6, No. 2 (May 1972), 187–204.

Pettigrew, Andrew M. “Towards a Political Theory of Organizational Intervention,” Human Relations, Vol. 28 (April 1975), 191–208.

Pinch, Trevor and Wiebe E. Bijker. “The Social Construction of Facts and Artifacts: Or How the Sociology of Science and the Sociology of Technology Might Benefit Each Other” in The Social Construction of Technological Systems: New Directions in the Sociology and History of Technology, eds. Wiebe E. Bijker, Thomas P. Hughes, and Trevor Pinch (Cambridge, Mass.: The MIT Press, 1987).

Porter, Theodore M. Trust in Numbers: The Pursuit of Objectivity in Science and Public Life (Princeton: Princeton University Press, 1996).

Powell, Mark R. Science at EPA: Information in the Regulatory Process (Washington, D.C.: Resources for the Future, 1999).

Price, Don K. The Scientific Estate (Cambridge, Mass.: Harvard University Press, 1965).

Roach, Mary. “I Was a Human Crash-Test Dummy,” Salon.com, November 19, 1999. http://www.salon.com/health/col/roac/1999/11/19/crash_test/print.html (Last accessed April 9, 2011).

Sandage, Scott A. Born Losers: A History of Failure in America (Cambridge, Mass.: Harvard University Press, 2005).

Schumpeter, Joseph A. Capitalism, Socialism, and Democracy (New York: Harper & Brothers, 1942).

Shapin, Steven and Simon Schaffer. Leviathan and the Air-Pump: Hobbes, Boyle, and the Experimental Life (Princeton, NJ: Princeton University Press, 1985).

Simon, Herbert. Administrative Behavior: A Study of Decision-Making Processes in Administrative Organizations (New York: MacMillan, 1947).

423

Skowronek, Stephen. Building a New American State: The Expansion of National Administrative Capacities, 1877–1920 (Cambridge: Cambridge University Press, 1982).

Slaton, Amy E. Reinforced Concrete and the Modernization of American Building, 1900-1930 (Baltimore: Johns Hopkins University Press, 2001).

Stigler, George. “The Economics of Information,” The Journal of Political Economy, Vol. 69, No. 3 (June 1961), 213–225.

Stokes, Donald E. Pasteur’s Quadrant: Basic Science and Technological Innovation (Washington, D.C.: Brookings Institute, 1997).

Taylor, Margaret R., Edward S. Rubin, and David A. Hounshell. “Regulation as the Mother of Innovation: The Case of SO2 Control,” Law & Policy, Vol. 27, No. 2 (April 2005), 348-78.

Timoshenko, Stephen. A History of the Strength of Materials (New York: McGraw-Hill, 1953).

Truman, David B. The Governmental Process: Political Interests and Public Opinion (New York: Alfred A. Knopf, Inc., 1951).

Usher, Abbot Payson. A History of Mechanical Invention (New York: McGraw-Hill, 1929).

Vinsel, Lee Jared. “’Every Week We Find a New Devil’: The Crusade for Credible Energy Information and Analysis, 1973-1981,” Forthcoming. von Hippel, Eric. Democratizing Innovation (Cambridge, Mass.: The MIT Press, 2005). von Hippel, Eric. “Economics of Product Development by Users: The Impact of ‘Sticky’ Local Information,” Management Science, Vol. 44, No. 5 (May 1998), 629–644.

Wald, Patricia M. “Thirty Years of Administrative Law in the D.C. Circuit Court,” Harold Leventhal Annual Talk at the DC Bar, July 1, 1997, http://www.dcbar.org/for_lawyers/sections/administrative_law_and_agenc y_practice/wald.cfm (accessed May 30, 2009).

Weber, Max. Economy and Society: An Outline of Interpretive Sociology, Guenther Roth and Claus Wittich, eds., Vol. 2, (Berkeley: University of Chicago Press, 1978).

424

Wells, H. G. The Works of H. G. Wells, Vol. 4, “Anticipations and Other Papers” (New York: Scribner, 1904).

Who's Who in America - 2007, 61st Edition (2006).

Who's Who in America - 2011, 65th Edition (2010).

Who's Who in American Politics, 22nd Edition (2009).

Who's Who in Science and Engineering - 2000-2001, 5th Edition (1999).

Winter, Langdon. “Do Artifacts Have Politics?”, Daedalus, Vol. 109, No. 1, Modern Technology: Problem or Opportunity? (Winter, 1980),121-136.

Yates, Joanne. Structuring the Information Age: Life Insurance and Technology in the Twentieth Century (Baltimore: Johns Hopkins University Press, 2005).

Yates, Joanne. “How Business Enterprises Use Technology: Extending the Demand- Side Turn,” Enterprise and Society, Vol. 7, No. 3, 422–455.

Yergin, Daniel. The Prize: The Epic Quest for Oil, Money, and Power (New York: Simon & Schuster, 1991).

Zeigler, Harmon. Interest Groups in American Society (Englewood Cliffs, NJ: Prentice Hall Inc., 1964).

Zizek, Slavoj. For They Know Not What They Do: Enjoyment as a Political Factor, 2nd ed. (London: Verso, 2002).

Theses and Dissertations

Albert, Daniel Marc. “Order Out of Chaos: Automobile Safety, Technology, and Society, 1925 to 1965,” (Ph.D. Diss., University of Michigan, 1997).

Jardini, David R. “Out of the Blue Yonder, The RAND Corporation’s Diversification into Social Welfare Research, 1946-1968 (PhD diss., Carnegie Mellon University, 1996).

Taylor, Margaret R. “The Influence of Government Actions on Innovative Activities in the Development of Environmental Technologies to Control Sulfur Dioxide Emissions from Stationary Sources” (PhD diss., Carnegie Mellon University, 2001).

425

Wetmore, Jameson Michael. “Systems of Restraint: Redistributing Responsibilities for Automobile Safety in the United States Since the 1960s,” (PhD diss., Cornell University, 2003).