Population and Development Review, Volume 26, Number 1

POPULATION AND DEVELOPMENT REVIEW

Kenneth Prewitt The US decennial

VOLUME 26 NUMBER 1 census: Political questions, scientific M A R C H 2 0 0 0 answers Melissa Leach and James Fairhead Challenging neo-Malthusian deforestation analyses in West Africa Simon Szreter and Eilidh Garrett Reproduction, compositional demography, and economic growth in early modern England Margaret E. Greene and Ann E. Biddlecom Demographic accounts of male reproductive roles Notes and Commentary J.C. Caldwell on rethinking the African AIDS epidemic Data and Perspectives R. Lee on long-term population projections and the US Social Security System Archives Emile Zola against Malthusianism Book Reviews Review essays by P. Streeten and V. Smil; reviews by T.J. Espenshade, D.G. Papademetriou, J. Horn, T. Waters, J. Glass, and others Documents The Census Bureau on prospects for US population growth in the twenty-first century Population and Development Review seeks to advance knowledge of the interrelationships between population and socioeconomic development and provides a forum for discussion of related issues of public policy.

EDITOR Paul Demeny

MANAGING EDITOR Ethel P. Churchill

EDITORIAL COMMITTEE Paul Demeny, Chair Geoffrey McNicoll Ethel P. Churchill Michael P. Todaro Susan Greenhalgh

EDITORIAL STAFF Robert Heidel, Production Editor Y. Christina Tse, Production/Design Margaret A. Knoll, Circulation Sura Rosenthal / Susan Rowe, Production

ADVISORY BOARD Gustavo Cabrera Milos˘ Macura John C. Caldwell Carmen A. Miró Mercedes B. Concepción Asok Mitra Richard A. Easterlin Samuel H. Preston Akin L. Mabogunje

Signed articles are the responsibility of the authors. Views expressed in the Review do not necessarily reflect the views of the Population Council. Direct manuscripts, comments on articles, and correspondence to: Population and Development Review Population Council One Dag Hammarskjold Plaza New York, New York 10017 USA Subscription information appears on the inside back cover. Support from the United Nations Population Fund is gratefully acknowledged. Volumes are available on microfilm from University Microfilms, Inc., 300 North Zeeb Road, Ann Arbor, Michigan 48106. The website for Population and Development Review is www.popcouncil.org/publications/pdr/default.htm The full contents of Volumes 1–22 (1975–96) are available through participating libraries from JSTOR at www.jstor.org/journals/00987921.html Population and Development Review (ISSN 0098-7921) is published quarterly in March, June, September, and December by the Population Council, One Dag Hammarskjold Plaza, New York, NY 10017 USA. Periodicals postage paid at New York, NY, and additional mailing offices. POSTMASTER: Send address changes to Population and Development Review, One Dag Hammarskjold Plaza, New York, NY 10017. © 2000 by The Population Council, Inc. All rights reserved. Printed in the United States of America. Library of Congress Catalog Card Number: 75-27311 ISSN 0098-7921 Information for Subscribers Population and Development Review is available on a paid subscription basis at the following rates: One year (4 issues) US$36.00 Two years (8 issues) US$60.00 To enter a subscription, send payment by check or money order drawn on a US bank, payable to the Population Council, or by Visa or MasterCard (provide card number and expiration date), together with mailing address to: Population and Development Review Population Council One Dag Hammarskjold Plaza New York, New York 10017 USA Credit card orders may also be placed by phone to (212) 339-0514, by fax to (212) 755-6052, or by e-mail to [email protected]

Automatic renewal For credit card payment only, standing orders are available at US$30.00 per year, with automatic renewal each year until subscription is cancelled by subscriber. Subscriber’s credit card will automatically be charged annually in December for the next year’s volume.

Complimentary subscriptions Complimentary subscriptions are available to qualified applicants residing in developing countries. Individuals and institutions wishing to apply for complimentary subscriptions should send requests on letterhead stationery to the above address stating the nature of professional involvement in development- and population-related issues.

Back issues Casebound (hardcover) volumes are available at $45.00 each. Single back issues are available at $8.00 per issue. Please designate copies desired by volume and issue number. Orders may be placed by mail, phone, fax, or e-mail as indicated above. Complimentary copies are available to qualified individuals residing in developing countries. A twenty-year cumulative index to Volumes 1–20 (1975–94) is available at no charge to subscribers to the Review.

Printed on recycled paper in the USA. POPULATION AND DEVELOPMENT REVIEW VOLUME 26 NUMBER 1 M A R C H 2 0 0 0

ARTICLES

The US Decennial Census: Political Questions, Scientific Answers 1

KENNETH PREWITT

Challenging Neo-Malthusian Deforestation Analyses in West Africa’s Dynamic Forest Landscapes 17

MELISSA LEACH

JAMES FAIRHEAD

Reproduction, Compositional Demography, and Economic Growth: Family Planning in England Long Before the Fertility Decline 45

SIMON SZRETER

EILIDH GARRETT

Absent and Problematic Men: Demographic Accounts of Male Reproductive Roles 81

MARGARET E. GREENE

ANN E. BIDDLECOM

NOTES AND COMMENTARY

Rethinking the African AIDS Epidemic 117

JOHN C. CALDWELL DATA AND PERSPECTIVES

Long-Term Population Projections and the US Social Security System 137

RONALD LEE

ARCHIVES

Emile Zola Against Malthusianism 145

BOOK REVIEWS

Freedom and Welfare: A Review Essay on Amartya Sen, Development as Freedom

PAUL STREETEN 153

Rocky Mountain Visions: A Review Essay on Paul Hawken, Amory Lovins, and L. Hunter Lovins, Natural Capitalism: The Next Industrial Revolution

VACLAV SMIL 163

George J. Borjas, Heaven’s Door: Immigration Policy and the American Economy

THOMAS J. ESPENSHADE 177

Douglas S. Massey, Joaquín Arango, Graeme Hugo, Ali Kouaouci, Adela Pellegrino, and J. Edward Taylor, Worlds in Motion: Understanding International Migration at the End of the Millennium

DEMETRIOS G. PAPADEMETRIOU 180

Alison Games, Migration and the Origins of the English Atlantic World

JAMES HORN 182

Joseph P. Ferrie, Yankeys Now: Immigrants in the Antebellum United States 1840–1860

TONY WATERS 183 Francine M. Deutsch, Halving It All: How Equally Shared Parenting Works

JENNIFER GLASS 185

Short Reviews 187

DOCUMENTS

The Census Bureau on Prospects for US Population Growth in the Twenty-First Century 197

ABSTRACTS 201

AUTHORS FOR THIS ISSUE 208

The US Decennial Census: Political Questions, Scientific Answers

KENNETH PREWITT

WHY HAVE PREPARATIONS for the US population and housing census in 2000 been the target of sharp partisan battles, battles that will also affect how the census counts will be used? The answer takes us, first, back to the political origins of the decennial census.

The census in the US Constitution

The US Constitution, written in 1787, includes a provision (in Article I, Sec- tion 2) for a decennial census:

Representatives and direct Taxes shall be apportioned among the several States which may be included within this Union, according to their respective Num- bers. . . . The actual Enumeration shall be made within three Years after the first Meeting of the Congress of the United States, and within every subsequent Term of ten Years, in such Manner as they shall by Law direct.

America’s founding political elites crafted this paragraph with two nation-building purposes in mind. The bold new Constitution they were writing would replace a weak confederacy with a centralized and stronger federal government. But as with any engineering of a new government, the first and difficult task is to build consensus on how power will be allocated. Geography, not estates or social classes, was the building block. The confederacy that was being pushed aside had given each of the original 13 states equal voting power. This favored smaller states; their leaders at the Consti- tutional Convention naturally insisted that the principle of state equality be upheld in the new government. The more populous states, such as New York and Virginia, balked at such power sharing, arguing the fairness of

POPULATION AND DEVELOPMENT REVIEW 26(1):1–16 (MARCH 2000) 1

Click to print article Click to return to Table of Contents 2 T HE US DECENNIAL CENSUS allocating political power proportionate to population size. A bicameral leg- islature offered the compromise. In one branch, the Senate, each state would be equal in its voting power; in the other, the House of Representatives, voting would be allocated proportionate to population. Of course “proportionate to population” meant that there must be a count of the population; consequently, the Constitutional innovation of providing for a census. But why a census every ten years? Because of the second nation-building task. Geographic expansion was much on the mind of the new nation. Its restless people were already crossing the Appalachian mountain range, spreading westward into the Ohio Valley and down the Mississippi River to the Gulf of Mexico. Some, among them drafters of the new Constitution, even imagined a nation reaching across the continent to the Pacific Ocean. A vast territory would thus be added to the new nation. But what was to be the status of these soon-to-be-acquired territories? Would they be annexed as colonies or join the nation on equal footing with the original 13 states? The prospect of an empire with colonies did not sit well with the principles of equality for which the war of independence had been waged. New and equal states it was to be, and the nation added states steadily throughout the nineteenth century. The decennial census measured population growth and its geographic dispersion and thereby served to regulate the pace at which the western and southern territories were added as new states to the Union.1 In serving these two Constitutional purposes—reallocation of power as the population grows and regulating the expansion of the union—the decennial census was intended to serve state-building, that is, “political” purposes; no sense can be made of current census controversies without appreciating this basic fact.

A troubling shift in the politics of the census

The decennial mechanism worked as anticipated by the founders. Every ten years, with one exception, seats in the House of Representatives have been reallocated as the population increased and moved westward. The exception is 1920. Southern, rural members of Congress did not welcome the 1920 census results that documented the massive wartime population movement from the rural south to northern cities. Reapportionment was delayed until after the 1930 census.2 The 1920s debate about census results was intensely partisan; and this was not the first political battle over the census. George Washington rejected a formula designed by New York’s Alexander Hamilton for allocating seats after the 1790 census, a step that led to a substitute formula more favorable to agricultural interests proposed by Thomas Jefferson of Virginia. Partisan concerns dominated during the pre–Civil War period as the acceptance of new states to the Union was carefully paced to maintain the balance between pro- and antislavery forces. For the most part, however, across

Click to return to Table of Contents K ENNETH PREWITT 3 nearly two centuries the partisan battles over the decennial census focused on how the census counts were to be used, not on how the census was to be conducted. Although there were grumbles about the completion and accuracy of counts as early as 1790, the counts were accepted as the best approximation available, and politics intruded at the point of interpretation and application. This circumstance has changed in the last 15 years, with far-reaching and troubling consequences for the US decennial census. Political leaders now argue that the census can be designed to predetermine the apportionment outcome in a way that favors one or the other of the political parties. This invites partisan disagreement not just about how the census count should be interpreted or applied, but how the census should be designed. This shift in political thinking began with legal action focused on the 1980 census and gradually gained momentum throughout the decade. It was brought into sharp focus when, in 1987, political appointees of the Re- publican administration in the Department of Commerce (the governmental home of the Census Bureau) overrode the professional judgment of senior Census Bureau staff, who are nonpolitical civil servants, about a key design element for the 1990 census. There was concern among a number of congressional members and mayors of large cities, mostly Democrats, that the Commerce Department had based its ruling less on technical or budget- ary grounds, as it claimed, than on partisan calculations. The Department’s decision was litigated, eventually leading to a court-enforced agreement that the Census Bureau could implement a modification of its recommended design for the 1990 census, but leaving unresolved how the count generated by that design might be used. In this agreement the court was protecting two centuries of census practice: the collection of the census information should be the responsibility of a professional Census Bureau, but the application of the numbers would be decided in the legal/political arena.3 Following the completion of the 1990 census, it fell to the Secretary of Commerce to decide whether to use the census results that had been produced by the disputed procedure. The Secretary, rejecting the recommendation of the Census Bureau Director, decided not to use those results, reasoning that:

[T]he choice of the adjustment method selected by the Bureau officials can make a difference in apportionment, and the political outcome of that choice can be known in advance. I am confident that political considerations played no role in the Census Bureau’s choice of an adjustment model for the 1990 census. I am deeply concerned, however, that adjustment would open the door to political tampering with the census in the future.4

This passage, as best I can determine, is the first instance in American political history in which a high government official gives voice to the speculation that the nonpartisan, professionally managed Census Bureau might

Click to return to Table of Contents 4 T HE US DECENNIAL CENSUS choose a data-collection methodology so as to favor one political party over another. Within a few years the partisan concern over what the Census Bureau might do had become the facile accusation that the proposed methodology for Census 2000 was in fact designed to achieve a given partisan outcome. A widely circulated memo to this effect was sent by Jim Nicholson, writing as Chairman of the Republican National Committee, to the state Republi- can Party leaders on 20 May 1997:

I am contacting you to recruit your assistance in addressing an issue of unusual importance to the future of the Republican Party. At the heart of the matter is one of the federal government’s most fundamental Constitutional functions: the United States census. At stake is our GOP majority in the House of Representatives, as well as partisan control of state legislatures nationwide.

The Clinton Administration is implementing a radical new way of taking the next census that effectively will add nearly four and one-half million Democrats to the nation’s population. This is the political outcome of a controversial Executive decision to use a complex mathematical formula to estimate and “adjust” the 2000 census. . . .

The GOP would suffer a negative effect in the partisan makeup of 24 Con- gressional seats, 113 State Senate seats and 297 State House seats nationwide. . . . An adjusted census could provide Democrats the crucial edge needed to prevail in close contests to control several state legislative chambers.5

Readers will seriously underestimate the importance of this memo if they conclude that it is only Republican Party officials who have concluded that the Census Bureau could and would propose a data-collection methodology with a given partisan outcome in mind. Census methodology has become fair game across the political spectrum—a deeply disturbing development. What, then, brought us to this point?

The stakes are raised: Social justice and federal funds

To understand the current controversy over the decennial census, we first take note of developments in the years following World War II that substantially raised the political stakes. The 1960s introduced notions of “social justice” to the politics of reap- portioning seats in the House of Representatives following each decennial count. A 1962 Supreme Court ruling advanced the “one person, one vote” principle, which prohibited legislative districts of unequal population size. Prior to this ruling the districts from which members of Congress were selected could have hugely different numbers of voters resident, in effect di-

Click to return to Table of Contents K ENNETH PREWITT 5 luting the voting power of the mostly urban districts of high population and enhancing the voting power of the mostly rural districts of low population. The politics behind this court ruling continue the long tradition of partisan battles over census results that primarily affect geographic interests— rural versus urban, southern versus northern. Partisan differences in this period also took into account differing demographic divisions in American society that partially overlapped the traditional geographic battle lines. Here the key legislative initiative was the Voting Rights Act of 1965, designed to allow equal access by racial minorities to the voting booth. Among other provisions, federal law now required that legislative district boundaries be drawn in a nondiscriminatory manner. To enforce one person, one vote and nondiscriminatory voting districts, the Department of Justice requires data in rich geographic detail. These data are generated by the decennial census. A breakdown of the population by race and ethnicity is now provided at the block level, and these data are used to determine whether various voting districts have been designed to dilute the voting power of racial minorities. Additional social justice responsibilities adhered to the decennial census in the wake of ongoing civil rights and equal employment legislation. The census provides the denominator against which to determine whether racial, gender, and ethnic patterns in employment, housing, education, and other social services reflect discrimi- natory intent. Federal departments making use of census data in this connection include, especially, the Departments of Labor, Housing and Urban Development, Education, and Health and Human Services. The postwar decades added yet a further task to the decennial census: allocation of federal funds. The relationship of census counts to resources was an on-and-off matter for most of the nineteenth and early twentieth centuries. There was a brief use of the census for taxation in the early nineteenth century, but this quickly fell away. And the early twentieth century saw scattered instances of using census-based formulas to distribute federal funds. The task of resource allocation increased in importance in the 1950s and 1960s. As summarized by Margo Anderson:

Congress expanded federal aid to state and local governments through the grant-in-aid system. Congress enacted programs that used statistically based funding formulas to cover federal support to states and localities for school lunch programs, airport construction, and hospital construction in 1946, and for water pollution control in 1948. In the 1950s and 1960s, the grant-in-aid [system] was used to supply federal funds for the interstate highway system, housing assistance, anti-poverty programs, employment and training programs, urban redevelopment, and water and sewer projects.6

As we begin the twenty-first century, the annual dollar amount of federal funding that uses census data as one of the criteria for program design

Click to return to Table of Contents 6 T HE US DECENNIAL CENSUS

TABLE 1 Summary of annual funds allocated by federal agencies through funding formulas and eligibility requirements based on census data as of 1999 Federal agency Total amount (dollars)

Department of Housing and Urban Development 7.2 billion Department of Transportation 25.4 billion Department of Education 12.6 billion Department of Health and Human Services 133.5 billion Department of Justice 1.6 billion Environmental Protection Agency 2.3 billion Department of Agriculture 10.7 billion Department of Commerce 351 million Department of Labor 4.5 billion Total 198.2 billion

SOURCE: General Accounting Office, US Government. >http://www.gao.gov< and allocation approaches $200 billion (see Table 1). Barring radical change in how the federal government works, the allocation of more than $2 trillion will be affected by Census 2000 counts before the next decennial census in 2010. This already huge number does not include state and local spending and private-sector investments, which together would add many millions more. Given these stakes—a socially just election system and huge flows of federal funds—it is no surprise that there is a partisan edge to the focus on census numbers. But this in itself does not explain the particular battle that erupted over census methodology. For that explanation we retrace our steps, returning to the 1940s and the differential undercount.

The differential undercount

The significant expansion of the purposes to which census data are put occurs in the wake of the “discovery” of what has come to be known as the differential undercount. A census undercount is a measure of how many persons short of 100 percent are not recorded in a census. An undercount of some magnitude has occurred in every US decennial census since 1790, when George Washington complained that a combination of citizen resistance and flaws in the enumeration procedure produced an official count that fell short of the true number of residents in the new country. In no decennial census since has the enumeration process been able to account for every resident in the United States, nor does it do so in any other nation. Demographers everywhere assume that a census is an approximation of the true count, perhaps an overestimation but more likely an underestimation.

Click to return to Table of Contents K ENNETH PREWITT 7

TABLE 2 Percent net undercount for black and white males in decennial US censuses, 1940–90 1940 1950 1960 1970 1980 1990

Black 8.4 7.5 6.6 6.5 4.5 4.4 Nonblack 5.0 3.8 2.7 2.2 0.8 1.2 Overall 5.4 4.1 3.1 2.7 1.2 1.6

NOTE: The net undercount is the result of the total overcount minus the total undercount; the point estimates presented here are based on demographic analysis. SOURCE: U.S. Census Bureau, summarized in “The Census: A history,” Population Resource Center, 1999. Additional documentation available at >www.census.gov<

Insofar as the benefits of a census are allocated on a share basis—as is true if the benefit is a fixed number of congressional seats or a fixed amount of federal funds—an undercount distributed equally across geographic units and population groups does not result in inequitable outcomes. Inequity or injustice emerges only if some areas or groups are counted at lower rates than others: that is, if the undercount is differential. That we know to be the case for the US census. Although census professionals had long accepted that the undercount rate differed from one demographic group to another, circumstances related to World War II provided the first systematic measure. In the early 1940s the government initiated mandatory, universal selective service registration. Though obviously not its intention, this universal registration provided statisticians with two independent counts of males between ages 21 and 35: the count recorded in the 1940 census and the count of those registered for potential enlistment to military service. Comparison of these counts provided the first reliable measure of how many persons, at least in this demographic group, had been missed in the census. Save for one factor, this finding would have attracted little interest beyond the demographers and statisticians concerned with improving census practice. What attracted wider interest was that African-American males of draft age had been missed in the census at much higher rates than white males. Here was the first systematic evidence of a differential undercount.7 Using a variety of techniques, the Census Bureau began to study the differential undercount—which, we now know, has persisted across the last half-century, as illustrated in Table 2. The first important national examination of the implications of a differential undercount was a 1967 Conference on Social Statistics and the City. The summary report drew the by-then-obvious conclusion:

Where a group defined by racial or ethnic terms, and concentrated in special political jurisdictions is significantly undercounted in relation to other groups,

Click to return to Table of Contents 8 T HE US DECENNIAL CENSUS

then individual members of that group are thereby deprived of the constitutional right to equal representation in the House of Representatives and, by inference, in other legislative bodies. They are also deprived of their entitle- ment to partake in federal and other programs designed for areas and populations with their characteristics. In other words, miscounting the population could unconstitutionally deny minorities political representation or protection under the Voting Rights Act. It could also deny local jurisdictions grant funds from federal programs.8

These themes are now commonplace in public discussion about census design issues—repeated on editorial pages and in congressional testimony, scholarly studies, civil rights newsletters, and Census Bureau documents. As a result, no one with even a passing interest in the decennial census can have failed to grasp that fair and equitable outcomes are jeopardized if the counts are more complete for some groups than others. And when the undercounted are racial and ethnic minorities, it is self-evident that the politics of representation will become framed as a civil rights issue. With significant issues of fairness at stake, what was the Census Bu- reau to do about the persistent differential undercount?

Census methodology: Dual system estimation

There are two ways to measure the differential undercount. One method compares census counts with population estimations based on demographic analysis. These estimations take into account births and deaths from vital statistics and rates of immigration and emigration and calculate the number of persons resident in the United States. These data, at the national level, can be used to estimate the population of selected racial groups, which can then be compared to census counts of the same groups to calculate whether underenumeration in the census differs by race. The second method derives from sample survey data collected inde- pendently of the census, but for the same time frame, that are then matched with census records. This matching process is the basis for calculating the rate of under- and overenumeration of population groups in the census counts. This statistical method, known in wildlife studies as capture/recap- ture, has a major advantage over demographic analysis. It permits the calculation of undercount rates for many more population groups and in greater geographic detail than national or state estimates.9 Moreover, dual system estimation allows the initial census count to be adjusted to correct for any detected under- or overcounting in the original census. Nicholson, quoted earlier, had dual system estimation in mind when he wrote of “a controversial Executive decision to use a complex mathematical formula to estimate and ‘adjust’ the 2000 census.” In short, dual

Click to return to Table of Contents K ENNETH PREWITT 9 system estimation and the correction of the initial census count are at the center of the partisan dispute over methodology for the 2000 census, as they were also in the 1990 census and to a lesser extent in 1980. My task here is not to assess the strengths or weaknesses of dual system estimation when applied to a national census. But note should be taken that a number of distinguished statisticians have expressed reservations about its application for reasons that have nothing to do with prospective partisan outcomes. Citing operational and statistical considerations, these experts conclude that trying to correct for the under- and overenumeration can introduce new sources of error in the census.10 Other statisticians agree with the judgment of the Census Bureau that an adjusted census based on dual system estimation yields a closer approximation of the true population count than an unadjusted census.11 Note should also be taken that adjusted counts move legislative seats and federal funds only at the margin. That is, although all of the seats in the House of Representatives (435) and significant federal funds are distributed on the basis of census counts, the changes in shares produced by adjusted counts are relatively small. The political debate, however, often focuses on the large totals rather than the comparatively small changes in shares, and thus tends to attribute exaggerated consequences to undercount rates.12

Census 2000 and the differential undercount

Early in its planning for Census 2000, the Census Bureau concluded that the differential undercount, which has been measured over a half-century but probably has characterized every decennial census for 200 years, could not be eliminated or even substantially reduced through traditional enumeration methods. No US population census can reach and solicit a response from every resident in the country. There will be evasion, refusals, and fraudulent responses; there will be missed housing units and incomplete counts within houses; there will be undetected errors in census operations. In short, there will be an undercount. And some population groups—the linguistically isolated, the less socially integrated, recent immigrants, those without legal status, the highly mobile, the poor and uneducated, those indifferent to civic responsibility—are more likely to be missed. The undercount, in short, will be differential. The Census Bureau accordingly proposed to use a post-census sample and dual system estimation as part of its methodology. It further proposed to apply dual system estimation techniques to adjust the under- and over- counts. For reasons by now obvious, this proposal was strongly opposed by Republican Party leaders in Congress. The Speaker of the House sued the Department of Commerce to preclude the use of sampling for apportionment.13

Click to return to Table of Contents 10 T HE US DECENNIAL CENSUS

In January 1999, a closely divided Supreme Court ruled that the federal statute governing the Census Bureau prohibited sampling for reapportionment purposes. The ruling appears to require the use of sampling (if feasible) for purposes other than reapportionment. These other purposes include the use of census data for redistricting and for the formulas that distribute federal funds. The Census Bureau then designed census operations to conduct as full a basic enumeration as possible in order to deliver state-by-state apportionment counts by the statutory deadline of 31 December 2000. In addition, the Bureau plans to collect data from a large independent sample of approximately 300,000 households. Results from this sample will be matched to the census counts in those households, and estimates of under- and overcounting will be calculated. These estimates can then be used to correct the original enumeration-based census, and the corrected counts will then be provided for purposes of redistricting and federal funding formulas. It is likely that legal action will be taken by states and cities that believe they will be either harmed or helped by the use of the corrected census counts. In particular, there is likely to be a legal effort to prohibit the use of adjusted census data in the state-by-state redistricting process. It is also likely that legal action will be initiated by advocates for the use of corrected numbers. As of this writing, therefore, the only certain outcome is that apportioning congressional seats across the 50 states will not involve any correction for the differential undercount.

Additional Census 2000 design issues

Although the differential undercount has attracted the most political attention, other design issues have not gone unnoticed, and here I make brief mention of three topics: the cost of the decennial census, the questions asked in the census, and the response rate or level of public cooperation with the census. Following the 1990 census, widely described as “less accurate” than the one that preceded it (see Table 2), the Census Bureau was instructed to prepare for a census in 2000 that would be more accurate and less expensive. The only design available to meet this exacting charge is one that em- ploys sampling methods. Basic enumeration is labor intensive, therefore expensive; it cannot reach everyone, therefore inaccurate. The initial design for Census 2000 recommended sampling at two points. One use, as noted above, was for dual system estimation to correct for the differential undercount. Sampling was also recommended as part of the initial enumeration procedures. In this application, the census would enumerate 90 percent or more of every census tract and then estimate the characteristics of the remaining nonresponders on the basis of a sample survey.

Click to return to Table of Contents K ENNETH PREWITT 11

The Supreme Court ruling that prohibited sampling for apportionment required the Census Bureau to abandon sampling during nonresponse followup as well as for dual system estimation. In other words, the earlier instruction that Census 2000 be less expensive and more accurate was impossible to meet. In the tradeoff between cost and accuracy, the Census Bureau opted for the latter and found the US Congress to be in agreement. The Bureau’s strategy was to increase enumeration accuracy by increasing public cooperation, both in the mailback phase of the census and in the nonresponse followup phase. In this effort Census 2000 includes a number of new operations: three first-class mailings to every housing unit (an alert letter, the census form, a reminder card); a costly paid-advertising campaign; extensive promotion activities such as a census-in-schools effort and paid partnership specialists; questionnaire assistance centers staffed by 15,000 paid employees; assistance guides in 49 languages; and so forth. These operations have been designed in response to an assessment that the population is harder to count than it was in 1990—the result of a higher proportion of new immigrants and of linguistically isolated households, an increase in non- standard living arrangements in which no one feels responsible for others in the household, demanding work schedules and greater mobility, and declining levels of civic engagement. The several innovations in publicity/promotion/assistance operations substantially increased the cost of the census, well above the standard increase that can be attributed to inflation and to growth in population and housing units since 1990. Full cycle costs of Cen- sus 2000 will exceed $6.5 billion. A related effort to increase mailback response rates was to simplify the census form, making it more user-friendly and reducing respondent burden. The former was achieved by shifting from a Fosdic data scanning technology (which is more demanding on the respondent) to intelligent character recognition that permits a questionnaire design easier to use in a self-enumeration context. To reduce respondent burden, the census in 2000 eliminated a number of questions and shifted others from the short to the long form.14 No questions are asked unless they are required by law or are necessary for an existing federal program. Short-form data can be used to estimate individual and household characteristics at the block level, and thus for the most part are restricted to those items necessary for enforcement of the Voting Rights Act of 1965. Long-form information is used to estimate characteristics at higher levels of geography. Of course, what is or is not asked in a census or, in some instances, what is on the short or long form, attracts keen political attention. One such example in 2000 had to do with the shift of the marital status question from the short form to the long form, which raised concern in

Click to return to Table of Contents 12 T HE US DECENNIAL CENSUS

Congress that the Census Bureau was taking a stand on (and slighting) family values. The Census Bureau has no policy position on marriage or on any other of the dozens of items on the long form: education, veteran status, language, occupation, income. It has simply been determined that information on these items is not needed at the block level. The political cry over the placement of the marital item—which led to a “Sense of the Senate” resolution endorsing family values and urging that the census respect these values—is interesting only in what it reveals about the political symbolism of a census. Of far greater relevance to social policy is a provision in Census 2000 to allow respondents to self-identify as belonging to more than one racial group.15 It is doubtful whether more than a few percent of the population will describe themselves in Census 2000 as multiracial, but the proportion will increase as interracial marriage becomes more common. This expected change in self-identification has long-term and unpredictable consequences for race-based social policy. Laws prohibiting racial or ethnic discrimination in education, housing, employment, or how voting districts are drawn assume a small number of fixed racial or ethnic groups, primarily black, His- panic, and Asian, though sometimes also including American Indians, Alas- kan Natives, and Native Hawaiian and Pacific Islanders.16 Enforcement often requires a numerator against which to measure observed practice: for example, if 12 percent of the population is black, but only 6 percent of, say, medical professionals are black, might this indicate racial discrimination? Census data are frequently the source of the numerators used to enforce nondiscriminatory social policies. With the proliferation of different multiracial groups in society and the general blurring of racial boundaries, the future of such enforcement is unclear.17 An underlying problem facing Census 2000 is the decline in mailback response rate. Since 1970 the decennial census has depended on a mailout/ mailback design. The census starts with a comprehensive address file and then delivers the census form to every residence in the country.18 In 1970 approximately 78 percent of households returned the form by mail. This dropped slightly to 75 percent in 1980, and then more sharply to 65 percent in 1990. Early planning for Census 2000 suggested that the response rate would be approximately 61 percent, and on that estimate the Census Bureau has budgeted and designed its operational requirements. Reversing the decline in public cooperation has become a major challenge. Low response rates negatively affect the quality of census data and increase costs. They are also indicative of a lack of civic obligation among large segments of the population. In fact, some thoughtful observers question whether the use of sampling in the nonresponse phase might further weaken civic obligation as persons concluded that the experts, deploying advanced statistical methods, could count the population irrespective of their cooperation. Although this concern is now moot, because sampling for

Click to return to Table of Contents K ENNETH PREWITT 13 nonresponse will not be used in 2000, it helps raise the broader question of whether the decennial census can be presented as a “civic ceremony” based on the idea of civic responsibility. It remains to be seen whether the decline in response rate can be reversed, but Census 2000 will make the effort to do so. It will heavily publi- cize the unsatisfactory 1990 mailback response rate, and, in a campaign la- beled “90 Plus Five,” challenge communities across the nation to improve their 1990 response by five percentage points. The overall promotion/publicity effort in the 2000 census might be viewed as a very large applied social science project. To what extent can this effort “talk the nation” into a higher level of public cooperation than our models indicate to be the natural level? This section has briefly covered three issues that are central to Census 2000, albeit not of a character as to have invited the same kind of intense partisan battle associated with using dual system estimation to correct for the differential undercount. I return to that issue and offer concluding comments on the politics of Census 2000.

The political battle revisited

To summarize the argument advanced in this article, the decennial census was established in 1787 to serve the political purpose of allocating seats in the House of Representatives. In the course of two centuries the census has had added to it two other major political tasks: federal fund allocation and enforcement of nondiscriminatory elections. When a census carries this much political weight, it is not surprising that partisan differences have emerged from time to time. Historically, these differences generally have concerned the application of census counts. In recent years a significant shift has occurred, with a heated battle emerging over the methodology of census-taking. This battle was initiated by the charge that the decennial census could be designed to predetermine partisan outcomes in the reapportionment and redistricting processes. As Census 2000 was planned, the charge shifted from “could be” to “is.” The accusation that the Census Bureau has acted and would act on a narrow partisan agenda fails to take into account two counter forces. First, neither the traditions nor the competencies of the Census Bu- reau are consistent with advancing a partisan agenda. By tradition the Bu- reau is resolutely professional and apolitical.19 Its peer community is the professional statistical community worldwide. It earns respect in that community by providing accurate data. Only persons uninformed about the importance of peer approval to professionals could believe that the Census Bureau would exchange its high standing among statisticians and demographers for a short-term political purpose. Moreover, the Census Bureau

Click to return to Table of Contents 14 T HE US DECENNIAL CENSUS enjoys the confidence of the American public—a confidence that is indis- pensable for public cooperation with its surveys and studies (the decennial census being only one of many Census Bureau activities). Even if its traditions were to allow it, the Census Bureau does not have the competence to predetermine partisan outcomes. It has no statistical expertise in reapportionment or redistricting and no expertise on trends in voting behavior. To predetermine partisan outcomes, the Census Bureau would need to bring to bear such expertise when it selected data-collection methodologies several years in advance of when census counts are used for reapportionment or redistricting. A second consideration is that the decennial census is designed and executed under a spotlight. For Census 2000, for example, four congressional committees have oversight responsibility; continuous examinations are undertaken by the General Accounting Office, an investigatory arm of Congress, and also by the Inspector General, an investigatory arm of the Executive Branch; a specially appointed Census Monitoring Board (well- funded and well-staffed) is given the explicit charge to report to the Con- gress and the President as to whether there is any evidence of political bias in census operations; a half-dozen major standing advisory committees, including one appointed by the National Academy of Sciences, pay close attention to details of census design and operations; and several dozen jour- nalists track census issues while many other private-sector groups have taken a lively interest in the preparations for and implementation of Census 2000. None of these oversight processes, advisory groups, or public watchdog efforts has identified any instance of partisan intention in the Census 2000 design—not for lack of looking, but because none is to be found. Although Census 2000 was designed without partisan intent or consideration, the question remains whether statistical adjustment could be implemented in a manner that favors a particular outcome. Statistical adjustment involves a number of technical choices. Concern has been voiced that these choices might reflect partisan considerations, that is, might take into account how the adjustment could benefit particular regions, states, or cities. Recognizing this concern, the Census Bureau is taking extraordinary steps to protect the statistical integrity of decisions at every step in the adjustment process. Internal to the Bureau is a committee comprised of senior statisticians and survey managers that reviews and documents every technical decision made by those directly responsible for implementing dual system estimation. A careful record is kept of these deliberations. Census design decisions are widely shared—through documents, congressional hear- ings, and in open forums with the oversight panel of the National Academy of Sciences. This process of internal deliberation and external transparency will continue through every step of implementation and will be available for public inspection. In short, the Census Bureau will continue to take every step possible to demonstrate that, in its design and implementation, sta-

Click to return to Table of Contents K ENNETH PREWITT 15 tistical adjustment is based on the best technical judgment available and never involves partisan considerations.

Notes

This article is based on the W. Edwards or undercount of the population,” Federal Reg- Deming Lecture of the American Statistical ister (Vol. 56, no. 140, 22 July 1991), p. 33583. Association, 14 August 1999, Baltimore, MD. The passage cited presents only one of the rea- 1 The Constitutional clause, cited above, sons offered by the Secretary; other consider- mandating the decennial census also antici- ations were technical and operational. pated that states would be taxed on the ba- 5 Emphasis in original. This estimation sis of population size. This practice fell into of how many Republican seats would be disuse early in the nineteenth century and “lost” has not been documented in any in- the reference to taxes was deleted in the dependent analysis. Most students of reap- Fourteenth Amendment to the Constitution, portionment and redistricting have con- ratified in 1868. James Madison, ever alert cluded that it is nearly impossible a priori to to the need to balance differing political in- calculate partisan shifts in legislatures that terests, noted in Federalist Paper No. 54 the could follow a decennial census. As indicated benefit to an accurate census of attaching later in this article, the Census Bureau does both representation and taxation to its count: not collect information on political party af- “It is of great importance that the States filiation, does not conduct analysis of pos- should feel as little bias as possible to swell sible partisan outcomes, and has no position or reduce the amount of their numbers. . . . on the soundness of Nicholson’s forecast. By extending the rule to both [representa- 6 Anderson, The American Census, cited in tion and taxation], . . . the States will have note 2, p. 203. opposite interests which will control and balance each other and produce the requisite 7 Historians and demographers have es- impartiality.” In this and in so many other timated undercounts, including differential areas, Madison was prescient. Now that a undercounts, for earlier decennial census re- greater population size registered in the de- sults. See, for example, John W. Adams and cennial census conveys only benefits and no Alice Bee Kasakoff, “Estimates of census penalties, there is a strong urge across thou- underenumeration based on genealogies,” sands of jurisdictions to “swell” but never to Social Science History 15, no. 4 (winter 1991): “reduce the amount of their numbers.” 527–544. 2 For an excellent treatment of this event, 8 David Heer (ed.), Social Statistics and the as well as for a broad and instructive overview City (Cambridge, MA: Joint Center for Ur- of the census in historical perspective, see ban Studies of the Massachusetts Institute of Margo J. Anderson, The American Census: A So- Technology and Harvard University, 1968), cial History (New Haven: Yale University Press, p. 11. 1988). 9 An excellent nontechnical overview of 3 A complete account of this event in how the Census Bureau uses dual system census history, on which I have relied, ap- estimation can be found in Tommy Wright pears in Margo J. Anderson and Stephen E. and Howard Hogan, “Census 2000: Evolu- Fienberg, Who Counts? The Politics of Census- tion of the revised plan,” CHANCE: A Maga- Taking in Contemporary America (New York: zine of the American Statistical Association 12, Russell Sage Foundation, 1999). no. 4 (fall 1999): 11–19. For more technical 4 Robert A. Mosbacher, “Decision of the treatments see the references cited in Wright Secretary of Commerce on whether a statis- and Hogan or consult the web site of the tical adjustment of the 1990 Census of Popu- Census Bureau. lation and Housing should be made for cov- 10 See Lawrence D. Brown et al., “Sta- erage deficiencies resulting in an overcount tistical controversies in Census 2000,” Juri-

Click to return to Table of Contents

16 T HE US DECENNIAL CENSUS

metrics 39 (summer 1999): 347–375. Critics Voting Rights Act mentioned above. Redis- of dual-system estimation focus on “distri- tricting sometimes also refers to the draw- butional accuracy” or how well the census ing of legislative boundaries for election to data distribute population shares across vari- state legislatures, county governments, city ous geographical units. The Census Bureau councils, and school boards. believes it has a Constitutional responsibil- 14 For Census 2000 only seven ques- ity to improve “numerical accuracy,” that is, tions are asked of 100 percent of the house- to count everyone possible, even though its holds. Most households receive these ques- operations cannot guarantee distributional tions on the short form. A sample of one out accuracy. It also believes that a census can- of six households receives the long form, not be planned with the explicit goal of dis- which includes the basic seven items as well tributional accuracy without serious damage as approximately 35 more questions asking to numerical accuracy. for demographic and housing information. 11 See, for example, the report of the 15 This decision is pursuant to a 1997 National Research Council’s Panel on Alter- statistical policy directive of the Office of native Census Methodologies, Measuring a Management and Budget (OMB). Changing Nation: Modern Methods for the 2000 16 The OMB directive and Census Bu- Census (Washington, DC: National Academy reau practice treat Hispanic as an ethnicity, Press, 1999). For a more general treatment recognizing that persons of Hispanic origin see Anderson and Fienberg, Who Counts?, can be white, black, Asian, or Native Ameri- cited in note 3. can Indian. 12 Adjustment of a state’s population for 17 Census 2000 identifies five discrete underenumeration by x percent would in- racial groups: white; African-American, crease that state’s claim on federal funds by black, or Negro; Asian; Native Hawaiian or roughly x percent if the population counts other Pacific Islander; and American Indian of all other states remained unchanged. If the or Alaskan Native. It also allows respondents population of all other states, too, were ad- to select an Other category, making a total justed by the same x percentage, the alloca- of six. There are 63 possible combinations to tion of funds would remain unaffected. As how the race question can be answered. the undercount would tend to differ from state to state, some states (those with above- 18 This sentence summarizes and over- average adjustment for undercount) would simplifies a number of operations. In 2000, gain, and others (with below-average adjust- for example, about 1.5 percent of the ad- ment for undercount) would lose. The per- dresses will be visited and directly enumer- centage deviations from the average in prac- ated by a census-taker; in rural areas, which tice are likely to be modest for most states. account for about 20 percent of the ad- For additional analysis see Formula Grants: dresses, Census Bureau employees will de- Effects of Adjusted Population Counts on Federal liver the questionnaire. For the remaining Funding to States: Report to Congressional Re- addresses, primarily in cities and towns, the questers (General Accounting Office, US Gov- form will be mailed through the post office ernment, February 1999). system. 13 The term “apportionment” refers to 19 Of a professional staff that numbers the allocation of the 435 seats in the House in the thousands, three positions are politi- of Representatives to the 50 states on the cal appointees: the Director and one person basis of census counts. It is to be distin- each in public information and in legislative guished from “redistricting,” which refers to relations. Only the Director is presidentially the drawing of legislative boundaries within appointed and congressionally confirmed. each state, the geographic areas so con- For instructive discussion of Census Bureau structed being electoral districts from which professionalism see Harvey M. Choldin, Look- Representatives are chosen. Census data are ing for the Last Percent: The Controversy Over used in redistricting because states must con- Census Undercounts (New Brunswick, NJ: form with the “one person, one vote” ruling Rutgers University Press, 1994). and the nondiscriminatory provisions of the

Click to print article Click to return to Table of Contents

Challenging Neo-Malthusian Deforestation Analyses in West Africa’s Dynamic Forest Landscapes

MELISSA LEACH JAMES FAIRHEAD

AMID MOUNTING GLOBAL concern over forest loss, studies of forest cover change have proliferated over the last decade. Whether produced by international forestry and conservation organizations, by national departments, or by academic researchers, many of these have at the core of their analysis and predictions neo-Malthusian assumptions about population–forest relationships. West Africa is a prime case in point. Assessments abound of regional forest loss as both dramatic and recent, with the Food and Agriculture Organiza- tion (FAO 1993) estimating 30 million hectares to have disappeared during the twentieth century, and the World Conservation Monitoring Centre (WCMC) suggesting that only 13 percent of “original” forest cover remains (Sayer, Harcourt, and Collins 1992: 74). Such statistics are linked to images of rapid population growth from an initially low baseline, swelled by immigration into forest areas. Such narratives at regional and national scales are supported by local ones: accounts which assume that expanding farming populations progressively degrade forest land and resources. Together, these narratives frame programs and policies designed to combat or redress forest loss, whether for reasons of biodiversity, climate, or resources and livelihoods, and whether by internationally initiated protection efforts or “community-based” natural resource management and rehabilitation. In broader population–environment debates, neo-Malthusian scenarios are frequently countered by neo-Boserupian1 alternatives. More people, it is argued, can stimulate technological and other changes that intensify land use so that higher populations per unit area can be supported without overall resource decline. In the case of forests, attention to such possibilities has been limited by the strength of the dominant assumption that Malthusian

POPULATION AND DEVELOPMENT REVIEW 26(1):17–43 (MARCH 2000) 17

Click to print article Click to return to Table of Contents 18 C HALLENGING DEFORESTATION ANALYSES relationships prevail, and by the preoccupation with people’s impact on “pristine” or barely disturbed forest environments, leaving little place for concern with precise population–resource relationships. In this article, we argue that neo-Malthusian deforestation narratives misrepresent the relationships between people and forests in West Africa and misstate the ways these relationships have developed during the last century. In so doing, these narratives obscure widespread processes by which people have enriched landscapes with trees, and in which the peopling of a landscape has sometimes meant an expansion of tree and forest cover where the opposite was assumed. We present this argument and a range of supportive evidence in the first half of the article. Nevertheless a neo-Boserupian explanation, while more appropriate, proves inadequate to account entirely for these processes and for the differences between trajectories of change that have occurred from place to place. By framing the issue primarily in terms of relationships between aggregate populations, an aggregate “environment” or resource set, and technology, both neo-Malthusian and neo-Boserupian perspectives exclude crucial questions relating to social and ecological specificity and history. The second half of the article presents a “landscape structuration” perspective grounded in fuller appreciation of social differences in environmental and resource priorities; of the diverse institutions that shape resource access and control; and of ecological variability and the legacies and pathways through which landscapes respond to use. Through examples from Ghana and Guinea, we show why such a perspective is useful for comprehending locale-specific trajectories of change and the factors that can tip the balance between “positive” and “negative” population–forest relationships.

Deforestation analyses in West Africa

There is a strong conviction in international conservation and forestry circles that West Africa has experienced marked forest loss during the twentieth century, only accelerating in the last few decades. The World Resources Institute, for instance, recently argued that “In West Africa, nearly 90% of the original moist forest is gone, and what remains is heavily fragmented and degraded” (Bryant, Nielsen, and Tangley 1997: 26). Such arguments place West Africa at the negative extreme of a devastating global picture recently summarized by Gardner-Outlaw and Engelman:

Half of the world’s original forest cover is gone—the cumulative effect of humanity’s use of forests over hundreds of generations. . . . In less than four decades, per capita forest area—the forest-to-people ratio—has fallen by more than 50%, from a global average of 1.2 hectares in 1960 to 0.6 hectares in 1995. (1999: 25–26)

Click to return to Table of Contents M ELISSA LEACH / JAMES FAIRHEAD 19

Ample statistics appear to bear out such a story, with West Africa– wide figures apparently supported by those for particular countries. Table 1 presents recent FAO figures concerning the nature, extent, and rate of forest loss over the decade timescale (1980–90), showing the relentless demise of remaining forest area with more than 10 percent lost over the decade on average. Table 2 presents data from an influential survey of anthropogenic vegetation change in West Africa during the twentieth century (Gornitz and NASA 1985), showing countries to have lost between 69 percent and 96 percent of the forest area that they had at the turn of the century. Table 3 presents the World Conservation Monitoring Centre’s analysis of the extent of present forest in relation to “original forest” (Sayer, Harcourt, and Collins 1992), showing that except in Liberia less than 13 percent of the original forest cover remains. Such figures concerning deforestation circulate widely in international scientific and policy arenas, where they are put to a variety of uses—whether

TABLE 1 Forest cover change 1980–90 in West African countries according to FAO Forest area 1990 (tropical rainforest and Forest loss/year moist deciduous forest) 1981–90 Percent of total Country (thousand hectares) (thousand hectares) forest lost/year

Benin 4,183 56.7 1.4 Ghana 9,151 134 1.5 Guinea 6,565 86.6 1.2 Ivory Coast 10,831 119.4 1.1 Liberia 4,634 25.4 0.54 Sierra Leone 1,889 12.3 0.65 Togo 1,318 21.8 1.6

SOURCE: FAO 1993: Annex 1, Tables 7a and 8a.

TABLE 2 Estimates of anthropogenic deforestation during the twentieth century in West African countries according to Gornitz and NASA Forest area c. 1900 Forest area c. 1985 Percent loss Country (hectares) (hectares) 1900–85

Benin 1,120,000 47,000 96 Ghana 9,871,000 1,718,000 83 Ivory Coast 14,500,000 3,993,000 72 Liberia 6,475,000 2,000,000 69

NOTE: Estimates for Guinea, Sierra Leone, and Togo are not available. SOURCE: Gornitz and NASA 1985: 291–294, 316–324.

Click to return to Table of Contents 20 C HALLENGING DEFORESTATION ANALYSES

TABLE 3 Forest cover c. 1985 in relation to “original” forest cover according to Sayer, Harcourt, and Collins “Original” Forest area c. 1985 forest cover Forest area c. 1985 as percent of Country (hectares) (hectares) original forest cover

Benin 1,680,000 42,400 2.5 Ghana 14,500,000 1,584,200 10.9 Guinea 18,580,000 765,500 4.1 Ivory Coast 22,940,000 2,746,400 12.0 Liberia 9,600,000 4,123,800 43.0 Sierra Leone 7,170,000 506,400 7.1 Togo 1,800,000 136,000 7.6 Total 76,270,000 9,904,700 13.0

SOURCE: Sayer, Harcourt, and Collins 1992: 74.

in formulating international agreements or in justifying conservation policies. Equally, they are reiterated in national contexts, becoming established in the literature because they are seen as authoritative (Grainger 1996). Analyses carried out in relation to specific countries frequently por- tray a remarkably similar picture of rapid forest loss from an “original” baseline. For example in Ghana, numerous assessments suggest a forest cover figure of 8–10 million hectares around 1900, now reduced to less than 2 million hectares.2 These are the figures that appear in the Seminar Pro- ceedings (1989) of the Forest Inventory Project sponsored by the then Brit- ish Overseas Development Administration, for instance, where Frimpong- Mensah (1989) asserts that:

[At] the turn of the century, Ghana had over 8,800,000 ha of forests. . . . Only 4,200,000 ha of this remained by about 1950. The estimate for 1980 puts the forest area at about 1,900,000 ha. This means that Ghana has lost over 75 percent of its tropical forest within this century, due to inefficient agricultural practices (shifting cultivation) and over-exploitation. (p. 72)

Ghartey, also working on this major forest inventory, asserts that around 1900 Ghana had 8.1 million hectares of forest, compared with 2.1 million hectares today (Ghartey 1989 in Parren and de Graaf 1995). Fair (1992) states that Ghana’s rainforests “have been reduced from 8.2 million hectares to 2 million hectares since 1900.” Ebregt (1995) notes that “at the turn of the century, it was estimated that Ghana had 8.8 million hectares of forest. By 1950, this had fallen to 4.2 million and by 1980 it was estimated at 1.9 million.” In keeping with these figures, the World Bank (1988) has

Click to return to Table of Contents M ELISSA LEACH / JAMES FAIRHEAD 21 estimated that closed forest has been lost at an annual rate of 75,000 hectares since the beginning of the twentieth century. If, as most authors point out, deforestation has recently slackened, this is presumably because little forest remains, and what remains is reserved. The statistics for Ivory Coast are even more ominous. Parren and de Graaf wrote recently that: “Côte d’Ivoire has experienced the most rapid deforestation taking place in any country in the world since the mid-1950s. The average annual deforestation rate, as a percent of the remaining forest, rose from 2.4 percent in 1956–65 to 7.3 percent in 1981–85, over ten times the pan tropical average rate of 0.6 percent” (Parren and de Graaf 1995: 29). Gornitz and NASA (1985: 290, 293) complement these deforestation rates with figures for forest cover. By 1980, they write, “around 70 percent of the forest present in 1900 had been cleared. . . . The area of closed forest in Ivory Coast decreased from 15 million ha in 1900 to 4.46 million ha in 1980, representing a decrease of 70.3 percent.” “The forest was not exploited until 1880,” write Arnaud and Sournia (1979: 290). “Until 1951, exploitation was very limited. From 1951–57 it developed slowly. . . . Extending over 12 million ha in 1956, the Ivorian dense forest now counts less than 4 million ha.”3 While many publications examine Ivorian deforestation, most of these rely on a very limited number of sources. Indeed most can be reduced to two key publications. Lanly (1969) attempted to identify changes in forest area between 1955 and 1965. FAO (1981) used these figures, add- ing an estimate for 1980 based on them, updated using agricultural land use statistics and an unsourced figure for 1900. Studies drawing on these works published between 1960 and 1990 suggest that Ivory Coast had between 14.5 and 16 million hectares of forest in 1900, began to lose it dramatically around 1955, and by 1990 had only about 2.7 million hectares remaining. Thus, estimates of forest loss for the interval 1950–90 stand at around 330,000 hectares per year. Across all the countries of the West African region, the major proximate causes of deforestation are claimed to be conversion of land for agriculture, with logging playing a major role in some areas. Of the root causes and systemic forces underlying these, population growth assumes a prominent place among other interrelated causes such as economic growth, short- sighted economic policies, poverty, and tenure insecurity (Gardner-Outlaw and Engelman 1999: 31). As the World Resources Institute argues from a global perspective:

Just since 1950, the world’s population has more than doubled. As a result, in many regions, forests have been cleared to grow food and to make way for new settlements. Population growth also drives up demand for timber, paper, fuelwood and other products from an ever-shrinking resource base. (5 Oc- tober 1998) >http://www.wri.org<

Click to return to Table of Contents 22 C HALLENGING DEFORESTATION ANALYSES

In the West African humid forest zone, the major process of concern is the conversion of dense forest cover to various forms of perennial crop plantation, fields, and bush fallow. While selective logging does not “deforest” in the same sense, it is held to stimulate the influx of farming populations into forest areas along logging tracks and to stimulate the agricultural conversion of forest land. In these processes, local population growth is seen to interact with population growth elsewhere—such as in urban areas, where it drives up the demand for forest zone commodities. On the northern margins of the zone, these same processes are held to bring about savannization: the progressive conversion of humid forest forms to vegetation dominated by grasses and subject to annual fires. Many social science analyses of the causes of deforestation are inter- linked with a view of deforestation and savannization as recent and rapid. Socioeconomic and political changes over the past century, which may have compromised once-effective “traditional” systems of resource use and management, are seen as contributing to deforestation, with effects accentuated by population growth. The significance of these issues to analysts’ views of modern rural society is clearly magnified the more widespread the deforestation with which they are linked. In this way, forestry and social science analyses can be seen to frame each other. This broader linking of assumptions about forest cover change and assumptions about the effects of population growth is exemplified in a World Bank overview of deforestation in West and Central Africa, which argues that:

[T]raditional farming and livestock husbandry practices, traditional dependency on wood for energy and for building material, traditional land tenure arrangements and traditional burdens on rural women worked well when population densities were low and population grew slowly. With the shock of extremely rapid population growth . . . these practices could not evolve fast enough. Thus they became the major source of forest destruction and degradation of the rural environment. (Cleaver 1992: 67)

In West Africa, arguments about the effects of population growth have been related not just to natural demographic increase, but also to immigration into the forest zone. Whether the southward immigration of Mande- speaking peoples into the forest zone since 1500, or more recent immigration of “environmental refugees” from drier, more northerly regions or refugees from the region’s war-torn countries, immigrants are held both to swell forest zone populations and to destabilize sustainable resource use practices. In Liberia, for example, a number of social, anthropological, and historical works propose seemingly powerful arguments concerning population movements and growth, and agricultural transformation, to account for

Click to return to Table of Contents M ELISSA LEACH / JAMES FAIRHEAD 23 the gradual loss of the country’s “original” high forest cover from around 1500, albeit accelerating in more recent times. Hasselman sums up this argument:

Numerous data show that before the historical shifts of the West African population in the sixteenth century, Liberia was almost totally covered in high forest. With the arrival of the coastal peoples, then the migrations from the Sahel towards the north of Liberia, large areas have been transformed into anthropic savannas. (1991: 51)

In Ivory Coast, Lanly (1969) summarizes an oft-cited history: “Until the seventeenth century the area of forest was not modified by the inhabitants who lived in close symbiosis with the natural environment. From the seventeenth century, and more particularly since the start of the colonial period, the introduction of food crops, export crops and population increase due to immigration” were responsible for a major reduction of forest area (1969: 46). However, it is not simply that population growth is called upon as a causal variable to explain deforestation estimated from “independent” sources. On the contrary, many estimates of forest cover change actually depend on (and tautologically, in turn support) assumptions about population change and its effects. This interdependency operates in a variety of ways. First, estimates of “original” forest cover, such as those forming the basis of the WCMC statistics above, are generally based on the assumption that a country’s forest zone—the zone deemed capable of supporting humid forest under today’s bioclimatic conditions—was at origin fully forested. It is assumed that the subsequent loss was caused principally by people’s land use, and hence that “at origin” the zone was either uninhabited or only sparsely inhabited by populations whose agricultural technologies were not harmful to forest cover. Frequently, this “origin” baseline is linked, either implicitly or explicitly, to the “precolonial period.” Thus, in Ghana in 1880–1900, the forest zone is assumed to have been fully forested—the basis for the figures cited above. In Ivory Coast, the widely cited FAO analysis presumes that the whole area of the bioclimatic “forest zone” was more or less intact forest around 1900. Thulet (1981) is explicit on this point. Work- ing for the forestry service on the National Environmental Commission, and author of “The state of the environment in Côte d’Ivoire,” he writes that “In 1900, the area effectively covered by the forest corresponded almost to that of the forest zone: 15,600,000 m ha” (see Parren and de Graaf 1995; Myers 1980, both of whom also make this assumption). In other cases, the dating of the baseline is left vague or is pushed back in time. For instance, in the figures in Table 3 and the recent statements of the World Resources Institute, “original” forest was calculated according to a map produced by the WCMC (Billington et al. 1996) of estimated forest

Click to return to Table of Contents 24 C HALLENGING DEFORESTATION ANALYSES cover “prior to the impact of modern man,” which they date about 8,000 years ago. The methodology used to produce this map again relied on the assumption that bioclimatic zones currently capable of supporting humid forest did, at this time, carry little-disturbed forest cover. The map thus depicts where forest might be expected to occur today in the absence of hu- mans (Bryant, Nielson, and Tangley 1997). It is admitted to be an indicator, not a direct measure of original cover, and in particular that the picture might look somewhat different were long-term climatic change to be taken into account. Nevertheless the analysis is used as the basis for constructing a country-by-country “forest frontier index” and for suggesting where forest protection efforts need to be targeted to protect these remaining undisturbed frontiers. Second, a number of influential analyses of forest cover change have explicitly used population growth as a proxy for vegetation change in the absence of other data. This is the case, for instance, with the FAO’s 1990 assessment and hence for the rates of change given in Table 1. For the majority of African countries, figures for the 1990 assessment were derived from single surveys, forward- and back-dated to 1980/90 using a model that linked population increase (from census estimates) to agricultural land clearance. Neo-Malthusian assumptions about local population–forest relationships are thus embedded in the forest statistics themselves. That such deductions are made even where good satellite data from the 1970s onward are available only testifies to the entrenched power of neo-Malthu- sian narratives in science-policy circles. Indeed, agencies such as the FAO still depend for deforestation data less on recent satellite imagery than on national forest surveys, which are frequently outdated and sometimes influenced by political considerations, as Gardner-Outlaw and Engelman (1999) and Grainger (1996) point out. Third, cross-country statistics concerning deforestation have been linked with population variables in the analysis of causes of forest loss. By corre- lating country-by-country deforestation rates with such other national variables as poverty, income growth, external indebtedness, interest rates, land market values, or structural adjustment, analysts attempt to identify the causal significance of each of these (e.g., Brown and Pearce 1994; Hyde, Amacher, and Magrath 1996). Indeed, as the discipline of environmental economics has gained strength, so international deforestation figures have increasingly been relied on to explore these economic and other variables as underlying causes. Similar methods are used to explore the impact of population growth, densities, and agricultural land clearance (Allen and Barnes 1985; Burgess 1992; Palo 1994). Fourth, assumptions about the causes of deforestation have often been formalized in models designed to predict future deforestation rates and trends. IDIOM, for example, a model developed in the framework of the

Click to return to Table of Contents M ELISSA LEACH / JAMES FAIRHEAD 25

Tropenbos program (Jepma 1995), was devised to simulate the effect of various policy options on deforestation rates, based on changes in the global economy, timber supply and demand, and changes in agricultural land use. A second model, GEOMOD (Hall et al. 1995), is also designed to simulate rates and patterns of land use change, using spatial preferences in land colonization (along rivers, up rivers, and up slope) and preferences for adja- cency and dispersion, considered on a region-by-region basis, to predict rates and patterns of tropical land use change. Supposed linkages between population growth and land cover change have formed the basis of numerous models predicting deforestation rates and associated carbon emissions (e.g., Houghton et al. 1983). Whether linked formally in models or in broader framings of problems, then, assumptions about forest cover change are linked to assumptions about population growth and vice versa.

Challenging deforestation analyses

This mutually supportive edifice of deforestation–population analysis invites cross-checking from independent sources—either from independent data on demographic change or from sources on forest cover change that are not dependent on population estimates. It is the latter that we now undertake, drawing briefly on work reported at length elsewhere (Fairhead and Leach 1998). The analyses of forest cover change since 1900 described above, based on deductions linked to assumptions about population growth, have not made use of historical sources or have done so uncritically. Historical sources that describe vegetation and landscapes as they were early in the twentieth century provide a useful means to falsify—or validate—the image of West Africa’s contemporary forest zone as virtually intact, little-disturbed forest. At the same time, they call into question rates of forest loss calculated according to these deductions. To pursue the two country cases introduced earlier, a consultation of early sources on Ghana suggests that the determination of past forest cover used in today’s estimates is highly questionable. Many authors have simply assumed that the entire “forest zone” (around 8 million hectares) was forest. Yet Table 4 assembles a number of estimates of the area of closed forest made by foresters early in the twentieth century. It indicates that the closed forest zone was then considered smaller than it is today (c. 6.6–7.2 million hectares), and that some 1.8 million hectares of land within the closed forest zone was even then considered to be farmland. A straightforward sub- traction suggests that the area of “intact forest” around 1922 was perhaps as much as 5 million hectares but not more. Furthermore, the forester Meniaud (1933) suggested that even these early statistics exaggerated the extent of forest cover. Criticizing earlier ana-

Click to return to Table of Contents 26 C HALLENGING DEFORESTATION ANALYSES

TABLE 4 Early land cover estimates for Ghana (hectares), from various sources Gold Coast Gold Coast Gold Coast Forestry Land cover Chipp Handbook Handbook Handbook Department category 1922 1925 1928 1937 1938

Closed forest zone 7,257,600 7,224,000 6,601,644 Merchantable forest 2,954,880 2,588,880 2,329,992 Unprofitable or inaccessible forest 1,451,520 1,294,440 1,268,551 Forest set aside for fuel 1,036,800 Total foresta 5,443,200 5,160,000 3,883,320 3,598,543 4,789,000 Agriculture, fallow, plantation 1,814,000 2,064,000 3,003,100 of which 390,000 cocoa aRows 2, 3, and 4 combined.

lysts for failing to distinguish between areas of the forest zone and areas within it that actually then carried forest, he suggested that in Ghana the area of “grande forêt” (closed forest) was only 4.5 million hectares or even less (1930: 537). These early estimates, generally based on field-level observations, do of course contain their own biases. Nevertheless, even allowing for inaccuracies, they confirm that large areas of the forest zone were certainly not “under forest.” If, as is suggested in most modern sources, significant deforestation began in the late nineteenth century, one must suppose the forest zone was indeed a zone of forest in the nineteenth century. Yet descriptions of land use at the time strongly contradict such images. For example, of the Ashanti area, Wilks draws on the contemporary estimates of Bowditch and Free- man to suggest that the “Metropolitan Ashanti” zone of about 1.7 million hectares centered on Kumasi had a population in the early nineteenth century of 500,000–725,000. Farming by these populations, combined with the frequent movement of villagers, meant, as Thompson noted: “The result is that large tracts of forest-covered country, quite out of proportion to the inhabitants they have to support become involved in the general purpose of destruction, and this is brought about by quite a small population” (Th- ompson 1910: 33). Thus when Huppenbauer visited Kumasi in 1881, he noted that “The actual land of Asante is not forest, as Akyem for instance, but mostly cultivated” (cited in Wilks 1978–79: 52). Accounts from other areas of Ghana similarly confirm the presence of large farming populations in the nineteenth and eighteenth centuries. In western Ghana, for instance, the seventeenth-century state of Aowin had similar effects on landscape prior to its decline in the eighteenth century, and such effects were enduring. Indeed Thompson, visiting the area during his 1908 survey of forests, remarked that:

Click to return to Table of Contents M ELISSA LEACH / JAMES FAIRHEAD 27

I was rather disappointed in these forests as we were led to understand by the guides that they were extensive and practically virgin in character. This we found to be very far from the case, and the whole tract of country showed unmistakeable signs of villages, having been once pretty well inhabited. Large tracts of forest were found to be of secondary origin, and signs of villages having once existed here were also not wanting. (1910: 46)

In Ivory Coast, a similar picture emerges. Table 5 summarizes estimates of forest cover at the beginning of the twentieth century by colonial foresters, again based on ground observation. By taking into account farmed and savanna areas in the forest zone, Meniaud calculated that “the primary forest, or that exploited only by the export timber industry” then totaled only 8 million hectares (Meniaud 1933). A little later, in 1937 (but still well before the midcentury onset of devastation as pictured today) Aubréville noted that “in travelling outside the major routes, one can no longer hold any illusion. Entire regions are covered only in secondary forest.” He was precise about the location of “primary” forest: “There exist still some large blocks of primary forest in uninhabited or uneasily settled regions, in middle Comoe and between Sassandra and Cavally for example, but in all the inhabited country, the secondary forest dominates” (Aubréville 1932: 239). Certain large savanna inliers within the forest zone were clearly marked on early maps. Chevalier’s 1912 map, for example, includes a vast savanna southeast of Dimbokro of around 100,000 hectares. And Aubréville noted the presence of open grasslands 34 km south of Daloa, far within the forest zone (1932: 241). Meniaud’s estimate was, in fact, larger than those of his contemporaries. Indeed Chevalier (1909) put the figure of unused forest at only around 6 million hectares. While such figures are of course problematic to interpret, they certainly cannot be taken to support the orthodox view that there were 14.5 million hectares of intact forest around 1900, gradually declining to 11 or even 9 million hectares by 1955. As in Ghana, locally specific descriptions of landscape also contradict the image of an uninhabited, pristine forest zone. For example, the region

TABLE 5 Early estimates of the area of the forest zone and area of high forest in Ivory Coast (million hectares) Chevalier Gros Sargos Mangin Meniaud Meniaud Aubréville 1909 1910 1928 1924 1922 1933 1937

Forest zone 12 12 11–12 N.A. 12 11 11 High forest 6 7.2–7.8 7.5–8 7.5 N.A. 8 N.A.

NOTE: Most of these scientists are known to have made independent assessments. Chevalier made two visits to Ivory Coast, spending 1.5 years to detail its botany. Gros and later Mangin were sent to make formal evaluations of forest resources. Meniaud was head of the Ivorian forestry service, while Sargos reported on the Ivorian wood industry.

Click to return to Table of Contents 28 C HALLENGING DEFORESTATION ANALYSES of Vavoua in western Ivory Coast is today considered to lie well within the forest zone. Several analyses today frame farming and other changes in this region in terms of massive deforestation, assuming that the Vavoua region contained forest in the 1950s that has subsequently been destroyed. Yet from all early maps (Chevalier 1912; Mangin 1924; Meniaud 1922), it is clear that Vavoua was classified outside the forest zone in early years of the twentieth century. An indication of the area’s vegetation can be gleaned from the account of Thomann, the first French officer to arrive there in 1903. On the road from Balogué to Bologué some 25 km south of Vavoua he described the first clearings. “One crosses a first very large clearing, with immense fan palms standing here and there at long intervals, as in Baoule. At last the misson has left the dense forest” (Thomann 1902–03: 632). Be- tween Bologué and Gaibi (5–10 km south of Vavoua) lay thicket and “brousse épaisse” (bushland without clearings) but whatever this was— whether he means tall grass savannas or relatively wooded land—it was not forest. Thence he proceeded to Vavoua, where the vegetation began to “open up seriously” (presumably into grasslands). If the region around Vavoua was in the forest around 1950, it was certainly not so in 1900. Table 6 summarizes a plausible reconsideration of deforestation in Ivory Coast, according to historical sources such as those noted above (see also Fairhead and Leach 1998). In short, 7–8 million hectares—less than half of the orthodox view—would seem to be a generous figure for Ivory Coast’s forest cover around 1900. The country may have had similar forest cover in 1955. Subsequent rates of forest loss remain quite high, at around 130,000 hectares per year on average, but this is only about 40 percent of generally accepted loss rates. We have elsewhere carried out systematic and detailed comparisons between the views of forest loss in international circulation and the evidence from historical sources, for Sierra Leone, Liberia, Togo, and Benin as well as Ghana and Ivory Coast (Fairhead and Leach 1998) and the findings are brought together in Table 7. The table suggests that twentieth-century

TABLE 6 Forest decline in Ivory Coast reconsidered (millions of hectares) Date Forest area (FAO 1981) Reconsideration c. 1900 14.5 c. 7–8 1955 11.8 c. 7–8.8 1965 9.0 c. 6.3 1973 6.2 c. 5.5 1980 4.0 — 1990 2.7 2.7 Total 11.8 c. 5.3–7.3

Click to return to Table of Contents M ELISSA LEACH / JAMES FAIRHEAD 29

TABLE 7 Suggested revisions to estimates of deforestation since 1900 (million hectares) Forest area lost “Orthodox” according to estimate of World Conservation Country forest area lost Monitoring Centre Suggested revision

Benin 0.7 1.6 0 Ghana 7 12.9 3.9 Ivory Coast 13 20.2 4.3–5.3 Liberia 4–4.5 5.5 1.3 Sierra Leone 0.8–5 6.7 0 Togo 0 1.7 0 Total 25.5–30.2 48.6 9.5–10.5

NOTE: The WCMC estimates are larger than other “orthodox” estimates since WCMC’s analysis equates forest cover in 1900 with “original” forest cover.

deforestation in these countries is probably only one-third of that suggested by the estimates in international circulation.

Challenging neo-Malthusian images of people–forest relationships

Reconsideration of forest cover change challenges two key elements in the population–deforestation thesis described earlier. First, it renders invalid the notion that an equilibrium of low (or no) population and undisturbed forest existed throughout the West African forest zone at the turn of the twentieth century, before “modern” rates of population growth took hold. Sec- ond, it revises downward the assumed deforestation rates over the time scale of the last 100–150 years, thus forcing modifications to cross-country analyses and models that correlate deforestation rates with measured population growth and make predictions from them. In itself, however, such a reconsideration offers no challenge to the assumed process by which growing populations convert forest to crops, fallow, and savanna, and no challenge to the image of this process as progressive and inevitable. One could argue that it alters one’s view of the extent and speed, but not the fact, of the neo-Malthusian population–deforestation relationship. More broadly, though, belief in rapid, recent loss of pristine forest has pushed commentators to make assumptions about the effects of population growth—growth that, in many parts of the region, has been rapid. It has deterred consideration of other possible processes and relationships, including those by which people may be enhancing tree and forest cover. In analyses such as those cited in the first section of this ar-

Click to return to Table of Contents 30 C HALLENGING DEFORESTATION ANALYSES ticle, evidence for these processes is either ignored or reinterpreted to fit the dominant neo-Malthusian vision. When the extent and rate of deforestation are reconsidered, however, such processes come more clearly into view. There is, in fact, evidence from localities throughout the West African forest zone of processes by which rural inhabitants through their farming and other everyday activities increase the density of woody vegetation and encourage the formation of forest in fallow and savanna. In these circumstances, population increase can be associated with forest cover increase, rather than with the opposite. Elsewhere we have assembled and discussed this evidence, drawn from a variety of historical, ecological, and ethnographic works (Fairhead and Leach 1996, 1998). In summary, widespread evidence exists for three key processes, frequently linked in their operation and effects. First, certain farming practices modify soils and related ecological processes in ways that promote afforestation. This effect is particularly significant in the forest–savanna transition zone, where gardening, mounding, and other techniques, linked to fire protection, contribute to the establishment of woody vegetation in one-time savannas. In contemporary times, this is happening, for example, in the Baoulé V of Ivory Coast where farmers say that “where one cultivates, the forest advances” (Spichiger and Blanc- Pamard 1973); in the Kissidougou region of Guinea (Fairhead and Leach 1996); and in Togo (Guelly, Roussel, and Guyot 1993) among other places. Historically, there is evidence that forest formations—including many current forest reserves—are “new forests” overlying once-cultivated savanna land, where the effects of farming contributed to forest establishment once farming populations declined (e.g., Fairhead and Leach 1994). Even in the heart of the forest zone, earlier cultivation can influence the type of vegetation in “new forests” that emerge as a result of population dynamics. Second, the creation of “islands” of forest in savanna, or of denser, high forest in bush fallow, appears to be a common aspect of settlement throughout the West African forest and transition zones. Interpreted within neo-Malthusian deforestation narratives as spared relics of former forest cover, evidence from Benin (Guelly, Roussel, and Guyot 1993), Guinea (Fairhead and Leach 1996), Nigeria (Lamb 1942; Jones 1963), and elsewhere confirms that these “new forests” are frequently the products of human action. Villagers variously appreciate such islands for the convenient forest products they afford, for the seclusion they offer for personal and communal activities, and for defense, whether against attack or fire. The islands are formed over time through various activities, including soil enrichment through gardening (as noted above) and the deposition of household waste, and the active planting or transplanting of tree seedlings. If settlements are linked to the formation of forest patches, it follows that population growth, where this implies a multiplication of settlements, can also multiply forest patches and hence forest area.

Click to return to Table of Contents M ELISSA LEACH / JAMES FAIRHEAD 31

Third, the planting and transplanting of trees in fields and fallows appears to be common throughout the forest and transition zones. Where tall trees of indigenous forest species stand amid crops, grass, or bush fallow, they have frequently—through neo-Malthusian deforestation lenses—been seen as surviving relics of farmer-led deforestation. Yet in many cases, farmers have precipitated or actively encouraged their growth, whether for improved agricultural productivity, for their valued timber and non-timber products, or as loci for ancestral or bush spirits. Do these processes suggest that the relationship between population and forest/tree cover change should be restated in neo-Boserupian terms? They do seem to bear out the contention, originally championed by Boserup for agricultural intensification and extended by others into the environmental field, that the expansion of human populations and settlement can result in positive changes in environmental resources. Yet several provisos must be introduced. The first relates to the word “can.” For while there is plenty of locale-specific evidence that these processes are taking place where populations are growing, there are other locales in which they patently are not. What accounts for these differences? The second relates to the word “result.” To what extent does population growth drive these (technological) changes, and how does it interact with other variables—including institutional and ecological dynamics—that are absent or accorded less prominence in the neo-Boserupian perspective? In particular, Boserup’s was a density- dependent thesis in which resource destruction could occur at low population densities but would be expected to reverse with population growth and intensification of resource use. How do other factors influence whether and how such density-dependent responses actually occur? And the third relates to the word “positive.” Changes look different depending on the particular vegetation form taken as the baseline—whether indeed “pristine” forest or savanna, bush fallow or “new forest.” It is also important to ask to what extent increases in tree cover and density of the type described here are valued positively by the relevant actors, whether by villagers, adminis- trators, or those in broader scientific and policy arenas. The answers to these questions are closely related, but they need to be addressed through a different conceptual framework from that underlying both neo-Malthusian and neo-Boserupian perspectives.

The institutional and ecological dynamics of population–forest relationships

Although they differ in viewing the relationships in question in negative or positive terms, both neo-Malthusian and neo-Boserupian perspectives generally characterize them as between an aggregate “population” and an aggregate “environment.” This applies whatever the scale of analysis; in local

Click to return to Table of Contents 32 C HALLENGING DEFORESTATION ANALYSES terms, the “population” may be considered as “the farmers” or “the community” of a given place. Yet in any particular place, we now suggest, the trajectory of change that occurs—and the reasons for it—need to be understood in different terms. First, a large social science literature underscores the need to disaggregate the “population” or “the community” so as to differentiate among women and men, youth and elders, people with different backgrounds, engaged in social relationships with each other, as well as with actors in the state, scientific, and policy communities. Different people may have different concerns about environmental resources, and in this context they may value trees, types of forest, or even forest versus savanna in different terms. Sec- ond, attention should be paid to the means through which people access, control, and direct the management of resources and features of the environment, and perhaps contest their claims and priorities with others. These means are influenced by a variety of institutions, both formal and informal, frequently overlapping, and operating at a variety of levels from norms of intrahousehold behavior to the laws and policies of state and international agencies (see Leach, Mearns, and Scoones 1999). Third, the ecological zones with which these institutionally shaped practices engage are themselves dynamic. Scientific debates have firmly challenged earlier views of ecosystems as characterized by balance, stability, and equilibrium (e.g., Botkin 1990), showing instead great spatial and temporal variability in ecological states and highlighting the importance of contin- gency and path dependency in ecological change. In the forestry sphere, ecological science now rejects many of its earlier models, such as the notion of climatic climax vegetation and hence the concept of a stable, “natural” forest climax predating human disturbance (Sprugel 1991; Hawthorne 1996). As such, concepts that, as we have seen, underpinned many analyses of forest cover change and neo-Malthusian narratives are no longer supported by many scientists. In particular, these concepts are challenged by evidence of long-term climatic fluctuations in West Africa, including—it now seems— a deep drought phase from 3,000 years ago, ending perhaps as recently as 600 years ago (Vincens et al. forthcoming). Over the long term, the contention of the World Conservation Monitoring Centre and others that “original” vegetation some 8,000 years ago consisted of forest in all areas now bioclimatically able to support it is untenable in this light. In more recent times, this rethinking means that ongoing climatic change and its shorter- term legacies must be seen as factors in people–forest relationships. Indeed, the perspective we have in mind conceives of ecological agency and human agency as interacting in the structuration of environments (Leach, Mearns, and Scoones 1999; Scoones 1999). Structuration (following Giddens 1984) refers to the interplay of agency and structure in processes of change. Environments are constantly being transformed, emerg-

Click to return to Table of Contents M ELISSA LEACH / JAMES FAIRHEAD 33 ing as the outcome of dynamic and variable ecological processes and distur- bances, in constant interaction with human use—use shaped by socially differentiated perceptions, management of environmental resources, and institutional dynamics. Ecological and social history intertwine not in linear fashion, but in contingent and path-dependent ways; practices at any given time may build on the legacy of earlier ones. Landscapes, a hybrid of the ecological and social spheres, may come to embody layer upon layer of the legacies of former institutional arrangements, almost as an archive. We now consider two case studies in these terms. Both are locales toward the northern margin of the forest zone, in areas where dominant neo- Malthusian narratives, prevalent in policy discussions, have assumed ongoing forest loss and savannization. In both cases, detailed social science research has revealed the occurrence of contrary processes of tree cover increase, enhanced by land use; yet these processes are seen only in some parts of the landscape and not in others, with precise trajectories of change depending on institutionally shaped processes of landscape structuration.

Wenchi district, Ghana

The first case, from the northern forest–savanna transition zone of Ghana, draws on field research by Afikorah-Danquah (1997) and Amanor (1993). The northern part of Wenchi district is characterized by a mosaic of gallery forest and semi-deciduous forest patches in savanna. While many analysts have assumed that this area, like other parts of Ghana’s forest–savanna transition zone, is undergoing progressive conversion to savanna, ecological research has suggested that the mosaic has been quite stable over the last few decades, with the distribution of forest and savanna shaped by underlying soil and topography (Asare 1962; Markham and Babbedge 1979). Never- theless, within this broad stability are areas where forest is ceding to savanna and others where tree densities are increasing and savannas ceding to bush and forest vegetation. Population density, at approximately 31 persons per square kilometer, is not high, yet the density of population nearly doubled between 1970 and 1984 (Amanor 1993). As we summarize here, the landscape effects of this moderately growing cultivation pressure have been strongly differentiated by the institutionally mediated practices of diverse social groups. In particular, the district is strikingly heterogeneous in terms of language and migration history, with the long-established, majority Akan-speaking Brongs joined by Hausa, Dagomba, Grunshi, Lobi-Dagarti, and the Sissala—all of whom have migrated from the northern savanna regions. Farming makes use of both forest and savanna sites in the forest–savanna mosaic. Cocoa was an important crop even in these marginal northern areas until the early 1980s, and forest sites were sought after and carefully protected for this purpose, modified by thinning the forest canopy and

Click to return to Table of Contents 34 C HALLENGING DEFORESTATION ANALYSES by repeated weeding of ground cover. However, in the dry season of 1982– 83 Ghana suffered a major drought during which devastating bush fires swept the area. Damage to cocoa crops, coinciding with a fall in world market prices, diminished the product’s viability and virtually all farmers abandoned cocoa. Forest sites, like savanna sites, were thenceforth used mainly for annual crops, including maize (increasingly important as a cash crop), yams, plantain, cocoyam, groundnuts, chili peppers, and tobacco. Farming practices for these crops vary considerably. Most farmers of long-established Brong-speaking families prefer to make use of fertile forest sites and are able to gain secure access to them, whether through inheritance, links to local chieftaincy, or marriage relations. These days most men have individualized rights over their landholdings; wives usually farm a part of their husbands’ field. On such forest sites, the farming technologies they employ center on the use of the cutlass in a “minimum tillage” cultivation regime that leaves the root mat of forest species little disturbed (Amanor 1993). Seven-to-ten-year fallows, which regenerate from seed or coppice, are maintained where possible. Although high forest and former cocoa forest sites are thus converted to rotationally managed bush fallow, many trees of forest species survive. Furthermore farmers preserve and encourage a variety of forest tree species within their fields, whether for their soil-enhancing properties or their economic value, commonly including Triplochiton scleroxyllon, Ceiba pentandra, and Anogeissus leiocarpus. Under such conditions forest sites regenerate as forest fallow rich in forest tree species. In some localities, farmers have begun to feel pressure to shorten fallows on these forest sites, as a result of the combined effects of population increase and commercial change. In these circumstances they recognize the dangers of conversion to savanna. A number of farmers have responded by enriching their fallows, integrating high densities of numerous varieties of more rapidly growing trees, many of which are preserved during cropping in new agroforestry patterns. Popular species include Margaritaria discoidea, Milicia excelsa, and Ficus spp. (Amanor 1993: 71). Such techniques are being used and experimented with actively by farmers who have secure landholdings, both to maintain the productivity of forest sites and to upgrade sites where the vegetation contains grasses. This can, therefore, be a means by which savanna is transformed into woody vegetation. As a further means by which savanna areas are being converted to woody vegetation, those farmers with land to spare and securely tenured holdings are now estab- lishing teak plantations, with the support of private corporations (especially the Pioneer Tobacco Company, which requires its outgrower-farmers to re- plant ex-tobacco areas with teak as a condition of their contracts). It is rare for farmers who are recent immigrants to undertake such practices. They generally farm under conditions of highly insecure tenure, having obtained land from landholding families on short-term borrowing

Click to return to Table of Contents M ELISSA LEACH / JAMES FAIRHEAD 35 arrangements, including sharecropping or in exchange for services. They thus lack incentives to plant and protect trees, and such activities would be discouraged by landowners as an attempt to stake long-term claims over land. The fields they are lent are frequently in savanna or semi-savannized forest, less preferred by landowner-farmers. Nevertheless, immigrants from northern savanna regions frequently prefer to farm savanna, using practices and technologies—including hoe cultivation and the use of ridges and mounds—to which they are accustomed. Their use of such techniques in savanna thus maintains the sites as savanna. Where used in short-fallow forest sites and without the tree enrichment techniques practiced by landowners, conversion to savanna can be hastened. To the extent that forest is being savannized in the area, then, this is largely attributable to the institutional arrangements under which immigrant farmers work (Afikorah-Dan- quah 1997). Notably though, to the extent that immigrant farmers themselves value savanna sites, such changes do not, in their perspective, constitute “degradation.” The long-established practice among landowners of protecting and encouraging forest trees in their fields intersects with another important set of institutional imperatives: those of the state Forest Department. Since the early colonial period, the state has claimed ownership of all forest trees in Ghana, a legal claim justified in part by the conviction that farmers only degraded this “natural” vegetation. As these trees have acquired greater value for timber and as the activities of loggers backed with Forest Department permits have increased, farmers fear that their field trees will be felled. This has reduced their incentive to protect and plant them, limiting the extent of landscape enrichment that might otherwise occur. Current moves to change Ghana’s forest laws in favor of farmers’ control over field trees may reverse this situation and encourage further tree protection.

Kissidougou prefecture, Guinea

Observers of the forest–savanna mosaic region of Guinea have also been convinced that a progressive savannization was underway as a result of the actions of burgeoning farming populations. Yet as we have documented in detail elsewhere (Fairhead and Leach 1996), historical analysis shows on the contrary that forest is generally advancing into savanna, with forest and forest fallow cover across Kissidougou prefecture having increased by around 40 percent since aerial photographic surveillance began in 1952. Crucial steps in this advancement have been the activities of the prefecture’s Kissi, Kuranko, and Lele populations in forming forest islands around their settlements and engaging in gardening-like farming practices that upgrade savanna soils and vegetation to woody forms. These steps seem to have taken place in a supportive climatic context, with the effects of rehumidification

Click to return to Table of Contents 36 C HALLENGING DEFORESTATION ANALYSES in the past 600–700 years complementing the forest-enhancing effects of local land use practices. Nevertheless, these effects have been far from uniform across the prefecture. As institutional conditions have interacted with ecological processes, savannization has occurred in some times and sites, albeit compensated by forest advance in others. Here we exemplify some of these differences by focusing first on trajectories of change in the most densely populated, eastern part of the prefecture, and then by considering how broad variations in patterns of land tenure and use have differently interacted with state policy to shape forest–savanna dynamics. Oral accounts suggest that the area around the village of Fondambadou in eastern Kissidougou was dominated by sparsely wooded savanna during the nineteenth century. Farmers, both Kissi- and Malinke-speaking, cultivated upland rice and other food crops in savanna fields. As in other areas of the prefecture, villagers encouraged islands of dense forest vegetation to form around their settlements through a combination of tree-seedling trans- plantation and various fire protection techniques. During the nineteenth century, fortification was important and rings of fast-growing Ceiba pentandra trees were established to assist defense. By the late nineteenth century, the forest islands of Fondambadou and three neighboring villages had enlarged so far that they merged, excluding fire from and thus afforesting the inter- vening savannas. The Kissi people, who had become politically dominant in the area by this time, actively encouraged this “uniting in one forest” as a symbolic and practical means of alliance and defense. This large single forest persisted into the twentieth century, with the forest near village pe- ripheries increasingly adapted and expanded to house kola and coffee, suggesting a positive relationship between forest growth and economic policies promoting certain cash crops. Population densities were increasing rapidly in this area, however: the village population of Fondambadou grew by about 140 percent between 1917 and 1993. By the 1940s, the need for upland farming for relatively large populations—combined with production pressures wrought by state- mandated requisitioning related to World War II—was prompting farmers to fell the areas of forest between each village forest island, subsequently managing them as forest fallow. Farmers commonly link this change to the arrival of fossa, a rice variety particularly suited to such upland forest sites. They cultivated upland rice in short cultivation sequences (rice, fonio, cas- sava) followed by seven to eight years fallow. With time and increasing population, however, many farmers encountered land shortages under this system. As fallow times shortened, they found it increasingly difficult to maintain rice yields and to prevent savannization of forest fallows. Instead, many farmers abandoned rice and used their land in intensive gardening- like cultivation of root and tuber crops, especially taro. In effect, they incor-

Click to return to Table of Contents M ELISSA LEACH / JAMES FAIRHEAD 37 porated more of the uplands into practices that had once been dominated by women in their kitchen gardens. Corresponding shifts occurred in patterns of labor use, with these practices particularly suiting farmers with limited access to the labor of others. The effect on the landscape was to transform certain forest fallow sites into sites of long (around five-year) cultivation sequences, interspersed with short, grass-dominated fallows. Rice farming became the prerogative of certain land-privileged social groups—principally older men heading landholding families. With much food cropping—including by junior family members—concentrated on gardened savanna sites, these groups have been able to retain upland rice sites under longer forest fallows. They invest in carefully timed early-burning and fire-break creation around fallows to protect them from potentially savannizing dry-season fires. Hence village territories in the area have, over the last 150 years and under continuing population increase, shifted from savanna-dominated to high- forest-dominated sites, to a mosaic of forest island, forest fallow, and savanna—with today’s mosaic reflecting institutionally shaped differences in farming practices. At least three major patterns of territorial organization in farming in the prefecture intersect with population and other factors in shaping local change. In the eastern areas just described, where current population densities are typically above 30 persons per square kilometer, the majority of land is cultivated and landholding families organize their fields and fallows within their own segment of divided village territory (Figure 1 panel a). In contrast, throughout many of the southern, Kissi-speaking areas of the prefecture, with similar current population densities, a different pattern pre- vails (Figure 1 panel b). Here many villages farm their entire territory, but do so by grouping their fields in a single block, moved annually in a coordinated village rotation around the territory. Also common in the area are large men’s and women’s labor groups that perform tasks in turn on each member’s field. These village institutions, whose broad forms have remained similar during the last century, have shaped different opportunities for fallow and fire management. Around the village of Toly, for example, the entire upland territory was largely savanna until the 1950s. Villagers farmed upland rice and—in a coordinated group—grazed about 200 head of cattle on unfarmed portions of village territory. Fire management was carried out by coordinated village institutions, which would mobilize to stop running fires and attempt to create fire-breaks around the entire fallow area, facili- tated by the relatively low boundary-to-area ratio of this land use pattern. While late-burning was valued for the maintenance of grazing land, there seems to have been little incentive to enforce fire protection. In the 1960s, however, village cattle numbers declined sharply. There was thus no need to maintain pasture, but the incentive to protect fallows from fire became greater. Villagers thus brought village-wide institutions into full play. In this,

Click to return to Table of Contents 38 C HALLENGING DEFORESTATION ANALYSES

FIGURE 1 Schematic diagram of contrasting patterns of land tenure and use in Kissidougou prefecture, Guinea

Panel A

Panel B

Panel C

Forest island Land area held by one landholding family Indicates fallow rotation their practices and their effects were reinforced in interaction with state fire policy. The latter—although justified on the faulty conviction that farmers had recently savannized the land—was at the time based on the use of fire-

Click to return to Table of Contents M ELISSA LEACH / JAMES FAIRHEAD 39 breaks and the prevention of all running fire, and close surveillance by forest guards actually strengthened village efforts. In this context, and in con- junction with climatic rehumidification and the legacy of earlier cattle grazing, the conversion of savannas to forest fallows was rapid and, by the early 1990s, nearly total. Areas of the Kuranko-dominated north of the prefecture follow a third, broad land use pattern. Here current population densities are typically less than 20, and in places less than 10, persons per square kilometer. The landscape in this drier region remains dominated by savanna. Cultivation is concentrated in certain portions of village territory (Figure 1 panel c) controlled by landholding families. The rest of the territory remains under village-level control and is used for pasture, hunting, and product collection. On many of the family-controlled sites, fallow vegetation is forest or forest thicket. Some of these are long-existing gallery forests; others are former-savanna sites where families have, over time, invested in gardening practices that have enriched soils and vegetation; and some are ancient village sites where settlers once created forest islands, and descendants have continued to make use of their enriched soils and forest fallow vegetation. Around these sites, farmers seek to protect their crops and fallows from fire by using fire-breaks and early-burning. On the savannas in between, grassland is valued for many activities, while, given the area’s prevailing dry conditions and low population density, villagers consider total fire prevention impossible. Instead, they have developed techniques for living with and managing fire, directing it to their advantage. In stark contrast with the southern areas considered above, these village institutions have clashed strongly with the state’s uniform fire prevention policy of the 1950s–70s; as it became more difficult to use established management practices openly, villagers experienced more dam- aging fires and greater losses of crops and property.

Conclusions

These two examples, from Ghana and Guinea, emphasize the extent to which population–forest relationships may be variable, nonlinear, and unpredictable. The trajectories described here—of shifts between savanna, forest fallow, and high forest of various types—have involved transitions in the direction of both more and less forest, sometimes several times in the same site during the last 100–150 years. Precise pathways of change have been shaped by the interactions of population change with diverse institutional and policy arrangements, ways of valuing vegetation at certain times, and dynamic ecologies. Neo-Malthusian perspectives fail to account for these processes. While such processes can theoretically be accommodated within a neo-Boserupian perspective, understanding them requires a further degree of specification of how social and ecological dynamics intersect with

Click to return to Table of Contents 40 C HALLENGING DEFORESTATION ANALYSES population change than is usual in analyses based on Boserup’s conceptual framework. Even if one were to focus on the effects of population growth, the two West African examples suggest that in the forest–savanna transition zone more people can lead to more forest, both by multiplying numbers of forest islands and places in the landscape that people wish to protect, and by increasing the labor available for managing fire and improving soils. Yet threshold effects clearly operate, with a point reached at which forest fallow needs to be reconverted for farming and other more intensive land uses. Fondamba- dou, in Guinea, is perhaps at such a point. Beyond this point, farmers may move into intensive tree planting for field and fallow enrichment, as in the Ghana case. But the population densities at which such practices are adopted—and whether, and by whom, they are implemented—are unpredictable because they depend on interactions with numerous other economic, institutional, and ecological changes. Furthermore, and as the cases illustrate, these diverse trajectories of change intersect in variable ways with the institutions of the state and its macroeconomic and environmental policies. Macroeconomic policies that enhance commercialization and commoditization of forest resources do not necessarily bring about more forest degradation: cash crop production, for instance, has had both positive effects (such as where tree crops have helped expand Guinea’s forest islands) and negative ones (such as where tenant farmers commercialize maize in Ghana) depending on its precise ecological and social setting. Environmental policies such as Forest Department tree tenure laws or fire protection initiatives have frequently been founded on neo-Malthusian images: of rapid, recent forest loss under growing population pressure. At a broad-brush level, one may assert that such policies— frequently born of colonial-era scientific and administrative influence, but persisting to the present—have been ill-conceived in relation to the realities of West African land use and forest cover change and have not only failed to achieve their goals but have impoverished land users (e.g., Fairhead and Leach 1998). Nevertheless, the task is not just to refute narratives of one-way deforestation under population pressure in order to reveal “accurate” realities beneath. Rather, attention should also be paid to how these narratives arise and become entrenched, and how they serve the institutions and individuals who deploy them. This requires a rather different project, focused on the interactions between science and policy; such a project is one that we and others are pursuing elsewhere.

Click to return to Table of Contents M ELISSA LEACH / JAMES FAIRHEAD 41

Notes

1 We preface these terms with “neo” to suggests a figure of 2.1 million hectares of re- emphasize that the perspectives discussed in maining forest, whereas at the lower end EPC this article follow broadly the spirit of (1991) and Paivinen and Witt (1988) estimate Malthus’s and Boserup’s original works, but 1.7 million hectares. Such discrepancies seem not necessarily the precise details and stric- to hinge on variable estimates of the forest tures of their assumptions and arguments. area outside the reserves. 2 Estimates of present forest cover in 3 All quotations originally in French have Ghana generally fall between 1.5 and 2.0 mil- been translated by the authors. lion hectares. At the upper end, EPC (1991)

References

Allen, J. C. and D. F. Barnes. 1985. “The causes of deforestation in developing countries,” Annales of the Association of American Geographers 75(2): 163–184. Afikorah-Danquah, S. 1997. “Local resource management in the forest–savanna transition zone: The case of Wenchi District, Ghana,” IDS Bulletin 28(4): 36–46. Amanor, K. S. 1993. “Wenchi Farmer Training Project: Social/environmental baseline study.” Accra: unpublished report to Overseas Development Administration. Arnaud, J.-C. and G. Sournia. 1979. “Les forêts de Côte d’Ivoire: une richesse naturelle en voie de disparition,” Cahiers d’Outre-mer 127: 281–301. Asare, E. O. 1962. “A note on the vegetation of the transition zone of the Tain Basin in Ghana,” Ghana Journal of Science 2: 60–373. Aubréville, A. 1932. “La forêt de la Côte d’Ivoire: essai de géobotanique forestière,” Bulletin commercial et historique du science, AOF 15(2–3): 205–249. ———. 1937. “Les forêts du Dahomey et du Togo,” Bulletin du comité d’études historiques et scientifiques de l’Afrique occidentale française 20(1–2): 1–221. Billington, C. et al. 1996. Estimated Original Forest Cover Map: A First Attempt. Cambridge: World Conservation Monitoring Centre. Botkin, D. B. 1990. Discordant Harmonies: A New Ecology for the Twenty-first Century. New York: Oxford University Press. Brown, K. and D. W. Pearce. 1994. The Causes of Tropical Deforestation: The Economic and Sta- tistical Analysis of Factors Giving Rise to the Loss of the Tropical Forests. London: UCL Press. Bryant, D., D. Nielsen, and L. Tangley. 1997. The Last Frontier Forests: Ecosystems and Econo- mies on the Edge. Washington, DC: World Resources Institute. Burgess, J. C. 1992. “Economic analysis of the causes of tropical deforestation,” Discussion Paper 92-03, London Environmental Economics Centre. Chevalier, A. 1909. “L’extension et la regression de la forêt vierge de l’Afrique tropicale,” Comptes rendus des séances de l’Académie des Sciences, Séance 30 Août 1909: 458–461. ———. 1912. Rapport sur une mission scientifique dans l’ouest Africain (1908–1910), Paris, 12 Janvier 1912. Missions Scientifiques. Chipp, T. 1922. The Forest Officer’s Handbook of the Gold Coast, Ashanti and Northern Territories. London: Crown Agents. Cleaver, K. 1992. “Deforestation in the western and central African forest: The agricultural and demographic causes, and some solutions,” in K. Cleaver et al. (eds.), Conservation of West and Central African Rainforests, World Bank Environment Paper No. 1. Washing- ton, DC: World Bank, pp. 65–78. Ebregt, A. 1995. Report in Tropical Rainforest and Biodiversity Conservation in Ghana. The Neth- erlands: Dutch Government Ministry of Foreign Affairs Directorate General for Inter- national Cooperation.

Click to return to Table of Contents 42 C HALLENGING DEFORESTATION ANALYSES

EPC (Environment Protection Council). 1991. Environment and Development in Ghana: Report to the United Nations Conference on Environment and Development (UNCED). Accra: Envi- ronmental Protection Council. FAO. 1981. Forest Resources of Tropical Africa, Part I and II (Country Briefs). Tropical Forest Re- sources Assessment Project. Rome: Food and Agriculture Organization. ———. 1993. “Forest Resources Assessment 1990: Tropical countries,” FAO Forestry Paper 112. Fair, D. 1992. “Africa’s rain forests—retreat and hold,” Africa Insight 22(1): 23–28. Fairhead, J. and M. Leach. 1994. “Contested forests: Modern conservation and historical land use in Guinea’s Ziama reserve,” African Affairs 93: 481–512. ———. 1996. Misreading the African Landscape: Society and Ecology in a Forest–Savanna Mosaic. Cambridge and New York: Cambridge University Press. ———. 1998. Reframing Deforestation: Global Analyses and Local Realities: Studies in West Africa. London: Routledge. Forestry Department. 1938. Annual Report 1937/8. Accra: Forestry Department of the Gov- ernment of the Gold Coast. Frimpong-Mensah, K. 1989. “Requirement of the timber industry,” in Ghana Forest Inven- tory Proceedings. Accra: Ghana Forest Department/ODA. Gardner-Outlaw, T. and R. Engelman. 1999. Forest Futures: Population, Consumption and Wood Resources. Washington, DC: Population Action International. Ghartey, K. K. F. 1989. “Results of the inventory,” in J. L. G. Wong (ed.), Ghana Forest Inventory Project. Seminar Proceedings, 29–30 March, Accra, pp. 32–46. Giddens, A. 1984. The Constitution of Society: Outline of the Theory of Structuration. Cambridge: Polity Press. Gold Coast Handbook. 1925. Accra: Government Press. ———. 1928. London: Crown Agents for the Government of the Gold Coast. ———. 1937. London: Crown Agents for the Government of the Gold Coast. Gornitz, V. and NASA. 1985. “A survey of anthropogenic vegetation changes in West Af- rica during the last century—climatic implications,” Climatic Change 7: 285–325. Grainger, A. 1996. “An evaluation of FAO Tropical Forest Resource Assessment 1990,” The Geographical Journal 162(1): 73–79. Gros, Cne. 1910. “Mission forestière à la Côte d’Ivoire (1908–1909),” Bulletin de la Société de géographie commerciale de Paris, 1910: 289–308. Guelly, K. A., B. Roussel, and M. Guyot. 1993. “Initiation of forest succession in savanna fallows in SW Togo,” Bois et Forêts des Tropiques 235: 37–48. Hall, C. A. S. et al. 1995. “Spatially explicit models of land use change and their application to the tropics.” CDIAC (Oak Ridge National Laboratory) DOE Research Summary 31. Hasselman, K. H. 1991. “Problèmes économiques et écologiques des forêts denses libériennes,” Cahiers d’Outre mer 44(1): 49–60. Hawthorne, W. D. 1996. “Holes and the sums of parts in Ghanaian forest: Regeneration, scale and sustainable use,” Proceedings of the Royal Society of Edinburgh 104B: 75–176. Houghton, R. A. et al. 1983. “Changes in the carbon content of terrestrial biota and soils

between 1860 and 1980: A net release of CO2 to the atmosphere,” Ecological Mono- graphs, 53(3): 235–262. Hyde, W. F., G. S. Amacher, and W. Magrath. 1996. “Deforestation and forest land use: Theory, evidence and policy implications,” The World Bank Research Observer 11(2): 223–248. Jepma, C. J. 1995. Tropical Deforestation: A Socio-Economic Approach. London: Earthscan Pub- lications. Jones, E. W. 1963. “The forest outliers in the guinea zone of northern Nigeria,” Journal of Ecology 51: 415–434. Lamb, A. F. 1942. “The Kurmis of Northern Nigeria,” Farm and Forest 3: 187–192. Lanly, J. P. 1969. “La regression de la forêt dense en Côte d’Ivoire,” BFT 127: 45–59. Leach, M., R. Mearns, and I. Scoones. 1999. “Environmental entitlements: Dynamics and institutions in community-based natural resource management,” World Development 27(2): 225–247.

Click to return to Table of Contents

M ELISSA LEACH / JAMES FAIRHEAD 43

Mangin, M. 1924. “Une mission forestière en Afrique occidentale française,” La géographie 42: 449–483; 629–650. Markham, R. H. and A. J. Babbedge. 1979. “Soil and vegetation catenas on the forest– savanna boundary in Ghana,” Biotropica 11(3): 224–234. Meniaud, J. 1922. La forêt de la Côte d’Ivoire et son exploitation. Paris: L. Dufay. ———. 1933. “L’arbre et la forêt en Afrique noire,” in Académie des Sciences Coloniales, Comptes rendus mensuels des séances de l’Académie des sciences coloniales: Communications, vol. 14, 1929–30. Myers, N. 1980. Conversion of Tropical Moist Forests. Washington, DC: National Academy of Sciences. Paivinen, R. T. M. and R. Witt. 1988. “Application of NOAA/AVHRR data for tropical forest cover mapping in Ghana,” paper presented at IUFRO meeting, 29 August–2 September. Palo, M. 1994. “Population and deforestation,” in K. Brown and D. W. Pearce, The Causes of Tropical Deforestation: The Economic and Statistical Analysis of Factors Giving Rise to the Loss of the Tropical Forests. London: UCL Press. Parren, M. P. E. and N. R. de Graaf. 1995. The Quest for Natural Forest Management in Ghana, Côte d’Ivoire and Liberia. Wageningen: Wageningen Agricultural University. Sargos, R. 1928. “Rapport général sur les bois coloniaux,” Actes et comptes rendus de l’Association colonies-sciences 34: 91–99. Sayer, J., C. S. Harcourt, and N. M. Collins. 1992. Conservation Atlas of Tropical Forests: Africa. Cambridge: World Conservation Monitoring Centre and IUCN. Scoones, I. 1999. “New ecology and the social sciences: What prospects for a fruitful engagement?” Annual Review of Anthropology 28: 479–507. Spichiger, R. and C. Blanc-Pamard. 1973. “Recherches sur le contact forêt-savane en Côte d’Ivoire: étude du recru forestier sur des parcelles cultivées en lisière d’un ilôt forestier dans le sud du pays baoulé,” Candollea 28: 21–37. Sprugel, D. G. 1991. “Disturbance, equilibrium, and environmental variability: What is ‘natural’ vegetation in a changing environment?” Biological Conservation 58: 1–18. Thomann, G. 1902–03. “Mission de Sassandra à Séguela,” Revue coloniale 1902/3: 621–653. Thompson, H. 1910. Gold Coast: Report on Forests. Colonial Reports—Miscellaneous No. 66. London: HMSO. Thulet, J.-Ch. 1981. “La disparition de la forêt ivoirienne: pertes et profits pour une société,” L’Information Géographique 45: 153–160. Vincens, A. et al. Forthcoming. “Forest’s responses to climate changes in Atlantic Equato- rial Africa during the last 4000 years BP., and inheritance on modern landscapes,” Journal of Biogeography. Wilks, I. 1978–79. “Huppenbauer’s account of Kumasi in 1881,” Asantesem 8: 50–52; 9: 58– 62; 10: 59–63. World Bank. 1988. Ghana Forest Resource Management Project. Working Papers 1–6. Un- published.

Click to print article Click to return to Table of Contents 44 C HALLENGING DEFORESTATION ANALYSES

Click to return to Table of Contents

Reproduction, Compositional Demography, and Economic Growth: Family Planning in England Long Before the Fertility Decline

SIMON SZRETER EILIDH GARRETT

IN RECENT YEARS, historical demographers have increasingly challenged the previous orthodoxy that the route to “modern” levels of fertility is bounded by married couples’ willingness or ability to adopt conscious, parity-specific limitation of births. Research on the initial stages of fertility declines occurring in America, in France, in England and Wales, and in Africa, for example, has provided compelling evidence of the widespread and prolonged spacing of births in addition to, and in some cases instead of “stopping” behavior.1 In view of current debates over the best measures by which to monitor fertility behavior, visual comparison of age-specific marital fertility rate (ASMFR) curves is perhaps the least controversial means by which to illustrate this point for England and Wales.2 Figure 1 depicts the “standard” ASMFR curve devised by Coale and Trussell to represent the fertility profile of a population with “natural fertility,” where no attempt is made to curtail fertility in a parity-specific way. Under this schedule a woman marrying on her twentieth birthday and surviving, still married, to celebrate her fiftieth birthday would, on average, bear a total of 8.75 children. National ASMFRs for England and Wales are not available until 1938 when age of mother at birth of child was first noted on all birth certificates. The lack of national data has made it difficult to pursue the path of marital fertility for the Eng- lish and Welsh population before the outbreak of World War II. Using their reconstitution of the historical parish registers of England and Wales, Wrigley et al. were able to estimate marital fertility schedules for quarter-centuries between 1600 and 1824. They found marital fertility to be highest in the first quarter of the nineteenth century when, on average, a married woman

POPULATION AND DEVELOPMENT REVIEW 26(1):45–80 (MARCH 2000) 45

Click to print article Click to return to Table of Contents 46 F AMILY PLANNING IN ENGLAND BEFORE THE FERTILITY DECLINE

FIGURE 1 Coale–Trussell age-specific marital fertility schedule, and selected nineteenth- and twentieth-century schedules for England and Wales

Coale–Trussell standard England and Wales 1800–24 0.5 OPCS staple industry 1891 OPCS white collar 1891 OPCS white collar 1911 England and Wales 1941–45 0.4 England and Wales 1951–55 England and Wales 1961–65 England and Wales 1971–75

0.3

0.2

0.1 Age-specific marital fertility rate

0.0 20–24 25–29 30–34 35–39 40–44 45–49 Age group

SOURCES: Coale–Trussell standard: Coale and Trussell, “Technical note” (cited in note 10), Table 1, p. 205; England and Wales 1800–24: Wrigley et al. (cited in note 3); OPCS staple industry and white collar: Garrett et al. (cited in note 5); England and Wales 1941–75: Office of Population Censuses and Surveys, Birth Statistics: Historical Series of Statistics from Registrations of Births in England and Wales 1837–1983 (London: HMSO, 1987), Table 3.2, p. 55. could expect to bear 7.67 children.3 Comparison of the schedules in Figure 1 indicates that the shortfall of marital fertility in England and Wales, compared with the Coale–Trussell standard, occurred mainly among women aged 25–40, suggesting that women in England were perhaps breastfeeding their children for longer periods or having less frequent sexual intercourse than the populations represented in the standard ASMFR schedule. By 1941– 45, as seen in the figure, the national fertility curve had taken on the classi- cally concave profile of a population exerting intensive parity-specific control on their fertility. In the post–World War II years, marital fertility rose among women in their 20s, creating the baby boom, but by the early 1970s marital fertility rates had fallen among women of all age groups: a synthetic cohort of married women experiencing these rates would, on average, produce 2.35 children. The familiar interpretation of the fertility decline in England and Wales posits that here, as elsewhere in most of Europe and where European peoples had settled in other continents, a new era of human control over reproduction commenced at some point during the last quarter of the nineteenth century. A taboo of sorts was broken and something previously believed to be in the hands of divine providence was brought under widespread human control: couples dared to tamper consciously and instrumentally with their own childbearing capacities. Individuals in many places and for a va-

Click to return to Table of Contents S IMON SZRETER/ EILIDH GARRETT 47 riety of reasons began to perceive that the relative costs and benefits of having large numbers of children were changing and that to exercise restraint on the number of children conceived was both morally legitimate and practically attainable. In demographers’ terms these populations abandoned natural fertility and became modern, rationally planning their family size. In the case of England and Wales the onset of fertility transition is often dated from 1876–77, the point at which the crude birth rate reached a peak value. Certainly from around this date couples from all levels of society began to reduce their family size, as the results from the 1911 fertility census attest.4 The manner in which this was achieved, however, was far from clear, as only the number of children born and not the timing of their births was elicited by the census questions seeking to monitor changes in “the fertility of marriage.” Recent work on an anonymized selection of individual returns from the 1891, 1901, and 1911 censuses of England and Wales has, however, allowed marital fertility schedules to be estimated for the populations of districts with particular social profiles in these years.5 Three of the curves can be seen in Figure 1. The curves for 1891 indicate marked differences in the level of fertility between districts mainly inhabited by the middle classes and their servants (the white-collar districts) and those situated in the industrial heartlands and peopled predominantly by members of the working classes (the staple industry areas). Any difference between the two 1891 curves in terms of shape is, however, much less marked; the middle classes were having fewer children because they were bearing them at a slower rate than the working classes, not necessarily because they were actively curtailing their fertility in a parity-specific manner. This impression is enhanced when the 1891 curve for the white-collar districts is compared with its counterpart for 1911; the decline in fertility observed has occurred largely among those in the youngest age groups. While it is possible that a proportion of these couples were using parity-specific stopping to produce exceedingly small families, as demonstrated recently by Anderson, these age-specific marital fertility curves strongly support the growing body of literature that argues that the role of spacing behavior in the fertility transitions of the industrial nations has been greatly underesti- mated.6 It also suggests not only that greater spaces were injected between births, but that the gap between marriage and first birth was also being extended: couples were also employing “starting” behavior. To what degree such behavior patterns were consciously devised remains open to question. Of relevance to this issue is that Kate Fisher’s recent oral history investiga- tions have uncovered an intriguing absence of consciously calculated family planning, or even acknowledgment of explicit communication between marriage partners on matters of birth control and family size, in her sample of respondents drawn from the 1920s and 1930s marriage cohorts in Eng- land and Wales. According to the demographic orthodoxy this should have

Click to return to Table of Contents 48 F AMILY PLANNING IN ENGLAND BEFORE THE FERTILITY DECLINE been a European generation thoroughly accustomed to parity-specific conscious planning, but Fisher found almost no testimony from either sex supporting this. Avoidance of conceptions seems to have been accomplished through implicit “understandings” and tacit concurrence, rather than explicitly agreed targets or reasoned discourse.7 Thus, the sharp historical divide around 1870 between “unplanned” premodern natural fertility and conscious parity-specific calculation is beginning to look distinctly untenable in Britain’s case. This article argues that if we abandon the conventional thesis of historical demography—that the emergence of parity-specific birth control is the single most important aspect of the fertility transition—then a new history of reproductive change under conditions of economic growth can be constructed. In this reconstruction we argue that the date 1816 should be accorded at least as much prominence as 1876. Indeed, the relationship between rapid economic growth and fertility should be seen as a protracted and socially diversified sequence of gradual adaptations, and not as a singu- lar, revolutionary event, or a one-time transition from a steady state, that of tradition, to another, that of modernity. We further argue that such a history requires probing beneath aggregated national trends of demographic change in order to observe and evaluate the diverse demographic responses to economic change of the different social groups and communities involved.

Late marriage as family planning in early modern England and Wales

During the 1950s and 1960s the demographic study of fertility became increasingly concerned with the policy aim of reducing fertility in the less developed world. All the advanced economies of the world were held to have passed through a demographic transition following industrialization, which included a shift to much lower birth rates. Certain leading demographers had come to believe that it was also possible to reverse the causal path and to engender economic growth with policies aimed at reducing birth rates in less developed countries.8 Of the diverse ways the anthropological and historical record documented that human groups restrict their fertility, only the specifically “modern” parity-specific birth control within marriage through contraception was promoted as the correct policy solution to overpopulation in poor countries.9 This form of control was, of course, the predominant practice in the West in the 1950s and 1960s. It was considered to have the policy virtue that it constituted a permanent behavioral change in the populace, guaranteeing irreversible low fertility. It was presumed that such behavior had emerged more or less spontaneously among the populations of the developed countries of the West during the previous 100 years or so.

Click to return to Table of Contents S IMON SZRETER/ EILIDH GARRETT 49

Thus, the policy perspective was conducive to a dichotomous view of world fertility history, which divided it into two simple categories, traditional and modern. It was acknowledged that the former could embrace many levels of natural fertility, but this was considered to be a relatively unimportant curiosity, since the focus of attention was on the emergence from many different cultures of the single modern form of controlled fertility: parity-specific rational family planning. Within the confines of this te- leological perspective, it was believed that the falling birth rates of all the Western nations during the nineteenth century reflected the spread of this modern and enlightened form of behavior. Provided an appropriate method for identifying and measuring the emergence of the modern form of birth control could be devised, history could assist in identifying how to trigger this desired form of behavior in the developing world. The demographic historian Louis Henry devised a metric equivalent to the concepts of traditional, unregulated natural fertility and modern, parity-specific controlled fertility. He apparently demonstrated that the presence or absence of parity-specific fertility control could be unambiguously identified and its incidence measured on an interval scale. The set of measurement techniques developed initially by Henry was elaborated subsequently at the Princeton Office of Population Research by Ansley Coale and James Trussell.10 A complementary methodology, the Princeton fertility indexes, had been developed at the same time by Coale, so that measures of marital fertility (Ig), independent of the confounding effects of either ille- gitimate fertility (Ih) or varying nuptiality patterns (Im), could be calculated from the rudimentary demographic data available in most European countries from official statistics in the nineteenth century.11 Demographers were thus able to interrogate European historical evidence for signs of the desirable parity-specific form of fertility reduction and to track the decline of marital fertility among national and provincial populations across Europe. It was generally assumed that these two distinct types of studies were providing measures of the same generic phenomenon. In fact, the Princeton fertility indexes had no power to discern whether marital fertility was falling because of parity-specific control (so-called stopping behavior) or any of a range of other possibilities. But until the late 1970s and early 1980s the belief that this modern kind of birth control was the only kind that could have been increasingly practiced among nineteenth-century and early-twentieth-century Europeans and North Americans went virtually unchallenged among demographic historians, so sure were they that it represented the early stages of a single social movement toward practices common among their own generation, namely contraceptive parity-specific family limitation. In focusing on the supposed emergence of this one modern form of parity-specific stopping behavior, without appreciating how recent an innovation this may have been, postwar demographic history lost sight of

Click to return to Table of Contents 50 F AMILY PLANNING IN ENGLAND BEFORE THE FERTILITY DECLINE many other possibilities. In dismissing as variants of natural fertility the manifest evidence of diversity in the fertility practices of earlier generations in the West, demographers and others overlooked the way in which these generations approached the management of reproduction and consequently failed to explore the relationship between these long-established forms of regulation and the further changes in the management of reproduction occurring during the more recent era of rapid economic change and fertility decline. The reduction of family size requires, in Coale’s well-known terms, the realization that such a thing is possible and permissible, the motivation to pursue it, and knowledge of and access to the means whereby it may be achieved.12 By setting chronologies of the emergence of widespread parity- specific fertility decline, demographic historians have sought to identify the features of modernization that generate such sweeping changes in behavior. This, however, has proven no easy task. The French as a nation abandoned natural fertility and began systematically to curb the size of their families in the prerevolutionary eighteenth century, long before the arrival of the “-izations” heralding modernity: industrialization, urbanization, secu- larization, and commercialization. Britain’s precocious industrial revolution meant that it experienced all the “-izations” generations before any continental society;13 yet its population was apparently following a natural fertility regime through the late 1870s. The country’s birth rate thereafter fell dramatically, at more or less the same time as many other much less economically developed and much less urban societies commenced their secular birth rate falls. In fact motivation to limit fertility was far from lacking in England, even in the eighteenth century. That there was awareness of the possibility that too many children could pose a predicament is attested to in a nursery rhyme, dated to 1797:

There was a little old woman, and she liv’d in a shoe, She had so many children, she didn’t know what to do.14

That most strata in contemporary society recognized the old woman’s plight and that many were already taking conscious steps to avoid her predicament are indicated by Malthus, in the first edition of his Essay published in 1798:

The preventive check appears to operate in some degree through all the ranks of society in England. There are some men, even in the highest rank, who are prevented from marrying by the idea of the expenses that they must re- trench, and the fancied pleasures that they must deprive themselves of, on the supposition of having a family.

Click to return to Table of Contents S IMON SZRETER/ EILIDH GARRETT 51

. . . A man of liberal education, but with an income only just sufficient to enable him to associate in the rank of gentlemen, must feel absolutely certain, that if he marries and has a family, he shall be obliged, if he mixes at all in society, to rank himself with moderate farmers, and the lower class of tradesmen. . . . The sons of tradesmen and farmers are exhorted not to marry, and generally find it necessary to pursue this advice, till they are settled in some business, or farm, that may enable them to support a family. . . . The labourer who earns eighteen pence a day, and lives with some degree of comfort as a single man, will hesitate a little before he divides that pittance among four or five, which seems to be but just sufficient for one. . . . The servants who live in gentlemen’s families, have restraints that are yet stronger to break through, in venturing upon marriage. They possess the necessaries, and even the comforts of life, almost in as great plenty as their masters. . . . If this sketch of the state of society in England be near the truth, and I do not conceive that it is exaggerated, it will be allowed, that the preventive check to population in this country operates, though with varied force, through all the classes of the community.15

According, then, both to Malthus and to the popular culture, reproduction and the burden of too many children were matters of conscious consideration in late-eighteenth-century English society; fertility was well within the sphere of conscious calculation in England in the 1790s. Indeed, Malthus himself proceeded, in subsequent editions of the Essay, to develop this theme of prudential marriage, in a burgeoning and very public discussion of the subject. Steve Hindle has found evidence of a carefully calculating approach to reproduction throughout the early modern period, in that representatives of the propertied governing class, such as Poor Law admin- istrators, consciously manipulated and influenced the marriage opportunities of the very poor in their communities. Not surprisingly the Poor Law itself, as this society’s welfare system, provided the focus for a widespread debate over the prolific poor and what to do about their lack of prudence.16 The discussion was not confined to a few practitioners of the dismal science and their radical protagonists, men such as Francis Place and Richard Carlile. Contemporary popular literature was every bit as preoccupied with the question of the appropriate financial and emotional circumstances for marriage, as Jane Austen’s novels clearly attest.17

Click to return to Table of Contents 52 F AMILY PLANNING IN ENGLAND BEFORE THE FERTILITY DECLINE

Research by demographic historians over the last three decades has confirmed Malthus’s detailed observations on prudential marriage as the primary means to regulate reproduction in early modern Britain.18 But this mechanism has been viewed as almost irrelevant to birth control, from the policy-oriented perspective of the second half of the twentieth century (although China, for instance, offers one vivid example of regulating age at marriage as part of a fertility control program).19 This is probably partly because postponed marriage is considered a method that is not obviously and easily transferable from one culture to another. Since John Hajnal’s an- nouncement in 1965 of a “North-West European marriage pattern” dating back at least to the seventeenth century, the late marriage regime has come to be understood as having existed only courtesy of a complex and unusual set of socioeconomic arrangements, such as young adult service and apprenticeship and collective provision for old age through the Poor Law.20 Thus, this early modern “late marriage” regime, apparently already centuries old in England by the time Malthus described it, has been conceptual- ized as having operated as an institutionalized, normative constraint on marriage, rather than as reflecting a process of conscious and rational weighing of the pros and cons by individuals in the past.21 Its history is therefore considered to be of no consequence in determining when the modern form of conscious parity-specific family planning first emerged. But if Malthus was accurately describing the conscious considerations, current sentiments, and motives of his contemporaries in all walks of life, then this suggests to us a very different chronology and history of the emergence of family planning among the population of England and Wales, one that would treat changes in nuptiality and fertility as part of an integrated subject, that of reproduction, and not as separate, distinct entities.

Reproduction and economic growth in England: The turning point of 1816

In Figure 2 estimates of marital fertility (depicted by lines and symbols) are graphed alongside the long-term movements of the quinquennial gross reproduction rate in terms of daughters per woman (solid and pecked lines) across the period c.1625–1935 in England and Wales. While marital fertility remained virtually constant across the seventeenth and eighteenth centuries before rising gradually across the nineteenth century to peak in the 1870s, reproduction rates underwent a sustained upward movement over a century and a half from the 1670s, followed, from 1816 onward, by a marked secular downturn. This was temporarily interrupted for a generation, 1846–76, before the gross reproduction rate continued downward to the mid-twentieth century. This figure therefore invites us to consider the possible significance of a much earlier reproductive revolution than the 1870s

Click to return to Table of Contents S IMON SZRETER/ EILIDH GARRETT 53

FIGURE 2 Trends in gross reproduction and marital fertility, England and Wales 1631–1935 3.2 3.0 2.8 2.6 2.4 2.2 2.0 1.8 1.6 1.4 1.2 a 1.0 Quinquennial gross reproduction rate (daughters per woman) 1631–1871 Alternate quinquennial gross reproduction rate 1841–1935b Rates (see legend) 0.8 0.33 x decadal total marital fertility rate (number of children per married woman): 0.6 50-year moving averages 1645–1805c 0.4 0.01 x quinquennial legitimate fertility rate (number of legitimate births per 1000 d 0.2 married women) 1851–1930 0.0 1636 1656 1676 1696 1716 1736 1756 1776 1796 1816 1836 1856 1876 1896 1916 1936 Quinquennia

SOURCES: aWrigley and Schofield (cited in note 18), Table 7.15, p. 230. bOffice of Population Censuses and Surveys, Birth Statistics (cited in note to Figure 1), Table 1.4, p. 26. cCalculated from Wrigley et al. (cited in note 3), Table 7.37, p. 450. Note that figures given in this source are 50-year averages. Symbols in Figure 2 are placed at mid-points of each five-decade period. dCalculated from Office of Population Censuses and Surveys, ibid. decline in marital fertility. Was there an epochal move away from relatively unrestrained nuptiality and fertility occurring soon after the beginning of the nineteenth century? This would be much closer in time to the precocious French fertility decline and more closely connected temporally to Britain’s early economic and urban growth. The upturn in reproduction that the figure shows for a generation from the late 1840s renders 1816 a much less obvious turning point than 1876, after which no such rally occurred. The skeptic might look at Figure 2 and see a temporary interruption, 1816–46, in a high fertility trend peaking in 1876. Indeed, Wrigley and Schofield interpret the period of falling overall fertility, 1816–46, as evidence of the continuing dominance of the homeostatic, preindustrial demographic regime of prudential marriage up to the mid-nineteenth century.22 According to this viewpoint the decline in nuptiality from the late 1810s to the 1840s should be seen as a normal early modern reaction to the economic difficulties faced by the populace at this time. Marriage age rose for a generation, just as it had always done during the early modern period in times of difficulty, as indexed by the price of wheat. According to this interpretation the homeostatic Malthusian preventive check, with its rising and falling nuptiality valve, was still in operation at this time, preserving the thesis that fundamental change in England’s reproductive regime came only in the 1870s when fertility within marriage began to fall, rising average real wages notwithstanding.

Click to return to Table of Contents 54 F AMILY PLANNING IN ENGLAND BEFORE THE FERTILITY DECLINE

In order to answer satisfactorily this skeptical viewpoint and to establish the prima facie plausibility of the interpretation presented here, a number of related issues need to be addressed. First, why should there have been a permanent change in society’s reproductive system in 1816, around the end of the quarter-century of wars with postrevolutionary France? Second, if this change was, indeed, a manifestation of conscious family planning, why did it take the form of delayed marriage, when apparently the French chose to restrict fertility after marriage, more in keeping with the parity-specific modern behavior identified by demographers? Third, why did overall fertility rise again for a generation after 1846 if such an important change in the population’s reproductive goals and practices had taken place? First, then, why should permanent change have occurred in the marriage market of young adults during the two or three decades after Napoleon’s final defeat in 1815 at Waterloo? Historians of the early modern demographic system generally agree that the quasi-universal institution of service or apprenticeship for older children and young adults was an integral part of the homeostatic demographic system’s successful functioning, providing, in Ann Kussmaul’s apt phrase, “ex post facto family planning” for the laboring poor.23 The many cottagers who, despite marrying relatively late, still found that they had produced too large a brood for their own household’s modest means and its equally modest labor requirements knew very well that the children could be sent off to self-supporting live-in service in others’ establishments from the age of ten or twelve years. One of the most substantial changes in the British agrarian economy during this period was the wholesale disappearance of living-in service arrangements, especially within the most densely populated and increasingly “enclosed” southern, eastern, and midland agricultural counties. This change was cumulative throughout the second half of the eighteenth and first quarter of the nineteenth century, as hiring of adult males by the month, week, or even the day became the norm, with no presumption of bed-and-board.24 Although living-in lingered in the north and the west, the nature of labor relations on the land was being completely transformed for most of the population. With the erosion of an essential element of those institutional arrangements whereby the homeostatic regime of the early modern period could function in the manner depicted by Wrigley and Schofield, that regime’s days were numbered. But why should the implications of this long-term trend have become salient only after 1816? One reason is that the artificial labor shortages and increased reliance on the land for all forms of food and raw materials during the two decades of the French Wars, 1793–1815, had put the longer-term effect into abeyance. Only with demobilization did the full implications of this profound secular change in the conditions of employment in the rural labor market become plainly visible to prospective

Click to return to Table of Contents S IMON SZRETER/ EILIDH GARRETT 55 parents. There was no longer a reliable and safe berth in the local economy for the surplus older children and young adults.25 Furthermore, during the same quarter-century, 1815–40, two other, independent sources of economic support for early marriage, both of which had probably sustained the increases in nuptiality and the rising fertility during the previous three-quarters of a century, were also coming under threat. First was the decline of cottage-based proto-industry. The mecha- nized, steam-driven factory, which began to transform the cotton spinning industry in the last decade of the previous century, was now spreading to the weaving side of the industry. This jeopardized the livelihoods of a quarter-million handloom weaving families in Lancashire alone. In due course the factories were to provide a more than equivalent number of livelihoods to replace those lost in the cottages, as overseas markets for the cheap Brit- ish manufactures expanded rapidly during the long mid-Victorian boom. The immediate and direct effect of expanding factory production, however, was to cause large-scale redundancies among cottage workers as one branch of industry after another succumbed to mechanization. That this threat was widely understood and keenly perceived by contemporaries from the second decade of the new century is demonstrated by the Luddite machine- breaking outbreaks of the time.26 Second were the ever-louder grumbles from landholders paying into the nation’s escalating Poor Law funds. They asserted that its “generosity” must be curtailed, particularly toward parents with a large number of dependents—the principle of outdoor relief enshrined in the so-called Speen- hamland system.27 Many parishes were already becoming meaner in their policies once the patriotic wars were over, in advance of the official regime of deterrent “less eligibility,” formally enacted in the New Poor Law of 1834, which cut the nation’s outlay on social security by 50 percent at a stroke.28 Thus, we have numerous well-documented explanations for a permanent change in England’s reproductive regime during the period 1816–46, operating primarily through a strong upward shift in the minimum age considered safe, sensible, or reasonable to enter into marriage from about 22.5 years to over 25 years for women.29 A key point in arguing that this was a historical departure is that this upward shift occurred in the context of gradually rising average real wages, contrary to the prevailing logic of the homeostatic system of the previous three centuries, as documented in Wrigley and Schofield’s time series. There are, therefore, significant reasons in the historical evidence to doubt the continuation of the early modern homeostatic demographic system into the nineteenth century and to see the year 1816 as marking a genuine reproductive disjunction. The difficulties experienced by the laboring poor in 1816–46 were conspicuously not the same kinds of periodic difficulties tracked by Wrigley and Schofield during the early modern pe-

Click to return to Table of Contents 56 F AMILY PLANNING IN ENGLAND BEFORE THE FERTILITY DECLINE riod. Their account emphasized lagged real wages as the most reliable nuptiality regulator of the preceding two and a half centuries. Yet there was no marked deterioration in the real wage index from the beginning of the nineteenth century. If anything, average real wages experienced a modest rise during this period of falling nuptiality, whereas they are believed to have been relatively static across the last half of the previous century when nuptiality fell so much.30 Also, why was this sharp rise in age at marriage not subsequently reversed during the mid-Victorian economic boom, c.1846– 76, when economic historians are agreed that real wages rose substantially for the great majority of the population?31 The homeostatic regime, had it been in place at that time, should have led to a substantial fall in marriage ages; yet female age at marriage fell by only a little over half a year.32 Thus, given the decline in farm service and much traditional apprenticeship along with significant change in the administration of the Poor Law over the period 1815–40, it does not seem credible to envisage the early modern demographic system continuing in operation when so much of its institutional context had changed. But why did the English not do what the French did at this time? Hav- ing apparently decided that smaller families were desirable, French couples became sufficiently adept at limiting the growth of their families that the historical demographic statistics demonstrate that from the late eighteenth century onward those provinces with the most restricted marital fertility were also those in which marriage was entered into at the earliest ages.33 We remain unsure of the precise ways in which French couples achieved their relatively low family sizes, with withdrawal, noncoital sex between married partners, and commercial sex for men all likely to have been important, along with various other possibilities such as postcoital douching, spermicidal plugs, and abortifacients.34 The available historical evidence indicates that modern barrier and rhythm methods were not utilized on a systematic scale in France (or in Britain).35 Nevertheless young French adults appear to have been sufficiently knowledgeable and confident of the efficacy of the techniques they used to reduce conceptions that they were able to marry relatively young yet avoid having large families. The difference between the two countries’ reproductive histories at this time has to be set in their respective political economic and demographic contexts. On the one hand, in France even more than in Britain, mortality was falling in the period 1780–1820, creating pressure and competition for economic resources through increased survivorship.36 However, unlike the case of Britain, the economy’s performance in France was indifferent and failed to create the same abundance of new jobs in manufacturing. Fur- thermore, the peasantry—much of it impoverished and increasingly land- less—along with the urban bourgeoisie had already been suffering penury under the corrupt and inequitable taxation of the Ancien Régime (which

Click to return to Table of Contents S IMON SZRETER/ EILIDH GARRETT 57 exempted the nobility from payment). Although outright celibacy in this Catholic country was generally lower than in England, the combined economic, fiscal, and demographic pressure was manifesting itself in ages at first marriage throughout the second half of the eighteenth century that were high (and were perceived to be high) by French historical standards.37 Thus, from a position in which marriage was already much postponed by their own culture’s standards, those sections of the French populace wishing to further restrict the size of their families in the late eighteenth and early nineteenth century apparently began to restrain fertility within marriage; and, once having perfected the means to achieve this, they were able to revert to earlier marriage. The historical experience of the British populace was quite different. From the mid-eighteenth century, sources of employment were multiplying rapidly in manufacturing, artisanal trades, and commerce, facilitating ever-younger marriages in each successive generation, without abrogating the prudential principle that marriage was not justified unless a couple had access to income sufficient to keep a family.38 Hence average age at marriage had steadily fallen to historically low levels by the opening decades of the nineteenth century (see above, note 29). Therefore, when prospects for employment and security of income seemed more doubtful during the decades following the French Wars, the immediate response of the populace was to revert to the delayed marriages that their parents and grandparents had practiced, though now in a changing institutional and cultural context. Precisely because marriage ages had fallen so low by the second decade of the nineteenth century, there was scope for a substantial fertility-controlling effect through delaying marriage once again, without need to further restrain fertility after marriage. The British were able to achieve a 30 percent reduction in fertility—not dissimilar to that achieved in France—through deferred marriage alone, whereas the French relied primarily on reducing the rate of childbearing within marriage. It is germane to the new story being proposed here to note that in England this was not simply an autonomic, homeostatic response but an ideologically and publicly negotiated and brokered change in the country’s reproductive practices. Considerable literary evidence demonstrates that a more radical departure into contraception within marriage was on the agenda in British society in the early decades of the nineteenth century. As pio- neering research by Norman Himes, David Glass, and Joseph Banks long ago demonstrated, Francis Godwin, Francis Place, and Richard Carlile were only the most well-known among a number of freethinkers and radicals who actively championed birth control after the French Wars.39 Indeed, anxiety over such a position being actively propounded by “godless” subversives engendered the vehement denigration of contraception as “vice” by conservative pillars of the establishment and advocacy of the draconian reform

Click to return to Table of Contents 58 F AMILY PLANNING IN ENGLAND BEFORE THE FERTILITY DECLINE of the Poor Law. The New Poor Law introduced incarceration as the pen- alty for indigence and unemployment, separating husbands and wives; and the infamous bastardy clauses dealt harshly with young single mothers. The law was enacted after a generation of turbulent debate and in the face of much hostility in a society in which the initiation of sexual relations and pregnancy in anticipation of marriage had been customary.40 In effect, an ideological battle played out in British society over the first third of the nineteenth century (in which the successive editions of Malthus’s Essay partici- pated) concerning the morality and propriety of contraception. It was a battle “won” hands down by the proponents of the conservative alternative: prudential marriage and “self-restraint.”41 This was quite unlike the situation at the time in France, where no such polarized debate occurred. Indeed, the current authority on the reception of Malthusian doctrine in France, Yves Charbit, has pointed out that Malthus’s principal popularizers departed so far from the original message as in some cases to invoke Malthusian sanction for the advice to the indi- gent poor to practice “contrainte prudente” inside marriage, in other words contraception; and avowing, quite to the contrary of the cardinal Malthu- sian position, that “le contrôle de la fécondité au sein du mariage n’est pas immoral.”42 In France the moral distinctions mobilized ferociously in Eng- land around fertility and sexual restraint outside, before, and inside marriage were lost in translation and rendered opaque. Whereas, on the other side of the English Channel, it became unthinkable for any respectable man of the upper, middle, or working class to let it be known by anyone that he offered his wife the use of a condom or any other obvious form of contraceptive. This conservative moral position was affirmed with great publicity at the notorious Bradlaugh–Besant trial of 1877–79, and the subject of contraceptives remained one of public opprobrium. As late as the 1920s a professional man could lose his job just for letting it be publicly known that he used them.43 The extent to which the British populace was constrained to revert to late marriage after the Napoleonic Wars by the outcome of this ideological battle over the public morality of contraception remains a moot question. The conflict was certainly significant, but it has also been argued that the English had a strong investment in a long-term culture of sexual abstinence.44 Before the late eighteenth century when illegitimacy temporarily increased, the early modern pattern of universal, late marriage accompanied by a low illegitimacy rate indicates a population of young adults systematically ab- staining from sexual intercourse for long periods before marriage. When- ever premarital intercourse led to pregnancy, marriage almost invariably followed. Although by no means monolithic, indeed always subject to periods of negotiation and phases of relaxation, this long-standing culture of sexual restraint was manifest in the well-documented waves of Puritan “ref-

Click to return to Table of Contents S IMON SZRETER/ EILIDH GARRETT 59 ormation of manners”: a sequence of episodic popular movements to promote sexual propriety c.1570–1620; 1648–60; 1688–1730; c.1780s onward. The original movement was a product of the Reformation aimed at the total moral regeneration of society. The second wave was part of the official ideology of the Puritan Interregnum. The third wave, at its most vigorous in the 1690s, was launched under the new Protestant dynasty of William and Mary and was more exclusively focused on sexual morality than the earlier movements, with the repression of prostitution, venereal disease, and illegitimacy as principal targets. In imposing the core of its ideological prescriptions on most of British society, in the form of the sexual codes of “Victorian” respectability, the final, post-1780s Evangelical movement came to exert a dominant, if never universal or uncontested, influence on the nation’s popular culture and practices right through to the 1960s.45 The essential point here, in any event, is that the population of early-nineteenth- century Britain was historically accustomed to a long-term culture and ideology of restraint and deeply familiar with the concepts and practices required for “controlled fertility,” without evident resort to barrier methods of “birth control.”46 Figure 2 may be read as being broadly consistent with the chronology of these early modern social movements for the reform of manners. It suggests that sexually abstemious social values had been in abeyance among the majority of the population, the laboring poor and artisans, during the second half of the eighteenth century. However, starting in the 1780s, and gaining momentum by the first decades of the next century, the ideology of self-restraint and anti-sensualism became generally influential once again as a result of the success of the final episode of the Evangelical reform of manners. Thus, the reversion to late marriage and the growing cultural emphasis on sexually abstinent male self-restraint as an aspect of self-improvement among the early Victorians of all classes may reflect long-standing cultural priorities in English society, as well as the specific impact of the Malthusian debate in the early nineteenth century.47 Is it the case, then, that at least two major reproductive changes took place in the course of the nineteenth century, and not just a single revolution toward its end? It is certain that from the late nineteenth century onward the control of fertility within marriage began to occur on a much wider social scale than ever before. But the historical evidence provides many reasons for evaluating the period 1816–46 as marking a climacteric in Britain’s reproductive regime, and not as a mere interruption in a century-long plateau of high fertility, c.1776–1876. This relates to a number of issues reviewed here, including what we know of the changed and fraught climate of discourse on marriage, sexuality, self-control, and abstinence at this time, in both genteel and plebeian circles, both among conservatives and pro- gressives of that generation. The New Poor Law of 1834 represented an

Click to return to Table of Contents 60 F AMILY PLANNING IN ENGLAND BEFORE THE FERTILITY DECLINE epochal change in the nation’s social security and welfare system, altering the rules of the game so as to change the culture of marriage and the balance of incentives.

Compositional demography: A new perspective on the historical demography of England and Wales

How, then, does the temporary rise in overall fertility for a generation after 1846 fit into the interpretation offered here? The perspective of compositional demography and some understanding of the principal socio-political and cultural changes occurring within British society in this period are required to answer this. In the simplest terms, British society can be envisaged as being pulled into two opposed camps over the century after 1750, in consequence of the intrinsically disruptive forces of rapid, market-oriented economic growth.48 A social polarization between the propertied middle class and the property- less working class had become apparent by the second quarter of the nineteenth century. Karl Marx, this era’s most acute and rigorous observer, correctly located the sources of this polarization in the exploitative logic of the capitalist system of market-oriented production. (What Marx did not an- ticipate, however, was that the century from the 1850s would witness a gradual process of negotiated socio-political rapprochement between capital and labor, which arguably only finally came to full fruition with Britain’s commitment to a welfare state following World War II.) In terms of compositional demography and geography, the effects of this class polarization, at its most pronounced during the second and third quarters of the nineteenth century, were reinforced by a sudden surge of population from the mostly rural small localities to the towns and cities, as shown by Table 1, Panel A for England and Wales. Panel B of the table shows that whereas the average increase in the percentage share of provincial towns and cities within the total population had been just over 1 percent per decade during the eighteenth century, by the first decade of the nineteenth century that increase had reached 1.65 percent. Over the second decade of the nineteenth century, however, the gain in the percentage share almost doubled to 3.01 percent, and the gain continued to rise substantially over each of the following three decades to reach 4.72 percent by the 1840s, which was then sustained through the rest of the century before falling back to a gain of 2.38 percent in the first decade of the twentieth century. The first line of the table shows that, in contrast, although expanding massively in absolute terms over the century, the percentage share of the population of London within the total population grew only gradually. The bottom line of the upper panel (the “rural” residual of places with fewer than 10,000 inhabitants) shows a corresponding progressive shrinking in

Click to return to Table of Contents S IMON SZRETER/ EILIDH GARRETT 61

TABLE 1 Panel A. Percent distribution of the population of England and Wales by size of place of residence, 1701 and 1801–1911 Category of place 1701 1801 1811 1821 1831 1841 1851 1861 1871 1881 1891 1901 1911

London 11 11.0 11.2 11.5 11.9 12.2 13.2 14.0 14.5 14.7 14.6 14.0 12.5 Over 100,000 0 0.0 2.5 4.1 6.7 8.5 11.7 14.8 18.1 21.5 24.8 29.7 31.3 10,000–100,000 2 12.9 12.1 13.5 15.1 17.6 19.2 20.2 21.7 23.3 24.8 25.5 26.3 Fewer than 10,000 87 76.0 74.2 70.9 66.3 61.7 55.9 51.0 45.7 40.5 35.8 30.9 29.9 Total 100 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Panel B. Decadal increase in the percentage share of provincial towns and cities with more than 10,000 inhabitants within the total population of England and Wales 1701– 1801– 1811– 1821– 1831– 1841– 1851– 1861– 1871– 1881– 1891– 1901– 1801 11 21 31 41 51 61 71 81 91 1901 11

1.09* 1.65 3.01 4.18 4.32 4.72 4.16 4.89 4.92 4.86 5.54 2.38

*Average per decade during the century NOTE: Percentage figures have been rounded to the nearest whole number; columns in the upper panel may not sum to exactly 100. SOURCE: Panel A is taken from R. Woods, “The effects of population redistribution on the level of mortality in nineteenth- century England and Wales,” Journal of Economic History 45 (1985): 645–651, Table 2. Panel B is calculated from Panel A.

percentage share, amounting to some 4.5 percent per decade during the nineteenth century. This compositional shift by place of residence was accompanied by a substantial south-to-north change in the balance of the population. Most of the preindustrial population had lived in the south of the country, where the agricultural land was best, whereas most of the new, expanding industrial centers were in the previously much less densely settled north and midlands, where the coalfields were best. Thus, the eight counties comprising most of the industrial north and midlands, which contained 22 percent of the population of England and Wales in 1701, had increased their share to 29 percent by 1801, to 34 percent by 1851, and to 39 percent by 1901 (falling back slightly to 38 percent by 1951).49 Furthermore, this south-to-north demographic shift accompanying industrialization had implications for class relations, imparting a pronounced geographical dimension to them. The heartlands of the established upper and middle classes lay in the south, which possessed the country’s best agricultural land and where London-based mercantile and manufacturing wealth had always been concentrated. The rapid expansion of new, coal-driven industry in the north and the midlands resulted in the growth of numerous large and small communities there, com- posed primarily of in-migrant proletarians admixed with petty bourgeois small masters and shopkeepers, but characterized, above all, by a dearth of established wealth-holders and professionals by comparison with the south and London.

Click to return to Table of Contents 62 F AMILY PLANNING IN ENGLAND BEFORE THE FERTILITY DECLINE

The resultant geographic imbalance in the distribution of the middle and working classes, after a century and a half of industrialization, can be illustrated through an analysis of data in the 1921 census. The population of each of 1,827 administrative units of local government in England and Wales (county boroughs, municipal boroughs, metropolitan boroughs, urban districts, and rural districts) was categorized as living in one of four types of local environment, “white collar,” “staple industry,” “light industry,” and “agricultural,” according to the proportions of resident adult males occupied in different kinds of employment. Twenty-one percent of the population lived in middle-class, “white-collar” environments, but over four-fifths of these lived in the southern half of the country (although only 42 percent of the total population lived in the southern half). On the other hand, 28 percent lived in strongly working-class, “staple industry” localities, and fully 98 percent of these lived either in the northern half of England or in Wales (comprising together the other 58 percent of the total population in 1921).50 In other words, industrial proletarian communities were almost unknown in southern England, while upper- and middle-class districts were a rarity outside the south. Thus, to approach the nineteenth century in terms of compositional demography would counsel against reading the performance of aggregate demographic indexes as necessarily indicative of uniform trends throughout British society. The urban proletarian north was becoming as important demographically as—but quite distinct in its economic fortunes and its social class characteristics from—the more patrician rural and metropolitan south. National trends in nuptiality and fertility during the period 1816–76 are therefore unlikely to be trustworthy as a guide to developments in either the north or the south. There is a clear parallel to this circumstance in the need for compositional interpretation of the mortality trends of this period. The national average indexes appear to show stable mortality conditions from the 1810s until the 1870s. In fact, this does not accurately describe the experience of any significant part of the population. The period actually witnessed marked deterioration in the health of the urban industrial proletariat in the north, especially during the economically turbulent 1830s and 1840s when the early trade cycle was highly volatile, contrasted with gradual improvement for the majority of those living in the south.51 Similarly, while the slight fall in national average age at marriage during the period 1846–76 may, then, reflect an increasing popularity of marriage among the growing mass of northern and midland urban workers, there could nevertheless have been different, even opposite, tendencies in the contracting of marriage (in proportionate terms) in the rural south and among the southern-based upper and middle classes.52 The compositional perspective, sensitive to the demographic divergence between north and south and between middle and working classes, raises

Click to return to Table of Contents S IMON SZRETER/ EILIDH GARRETT 63 the possibility that large sections of British society, the property-owning southern bourgeoisie or the rural poor in the south, or both, may have become committed to a regime of tight and conscious control over childrearing much earlier than previously thought. Although the middle class was growing in absolute numbers at this time, their fertility and nuptiality behavior has remained hidden by the quite different reproductive behavior of the urban laboring poor, particularly the disproportionately growing numbers filling the industrial towns and cities of the north and the midlands. Thus, when in the more stable and prosperous decades after 1846, the industrial proletariat once again embraced slightly earlier marriage and higher marital fertility for a generation, their behavior may have obscured a permanent shift to very late marriage by the bourgeoisie and possibly some sections of the rural poor in the south. From the late 1840s until the mid-1870s the British economy experienced its celebrated mid-Victorian boom: sustained economic prosperity associated with expanding global markets and rising real incomes and wages for virtually all sections of the populace. Even the severe disruption to the cotton trade in the early and mid-1860s occasioned by the American Civil War did not apparently dent the national economy’s overall performance during this period. It would have been extraordinary, with widespread prosperity increasing year after year, had there not been some relaxation of the postponement of marriage seen during the quarter-century of difficulties following the end of the Napoleonic Wars. Although the scale of resump- tion of earlier marriage was sufficient to break the single trend decline in national levels of reproduction from 1816 onward, this in no way represented a popular return to the early marriages of the late eighteenth and early nineteenth century. It is notable that though the general level of earnings was significantly higher after 1846 than it had been before 1816, and also continuously rising for the majority of the population, the average age at marriage declined only modestly. In short, the thesis that there had been a fundamental change in behavior with respect to marriage during the generation after 1816 is not necessarily contradicted by nuptiality behavior during the long mid-Victorian boom. Thus far we have argued that if one examines the course of reproduction as a whole, and not merely marital fertility, modern British history shows as much of a major, overall change of direction around 1816 as in 1876. We have also argued that these turning points in the national aggregate demographic indexes may not be accurate guides to the chronology of the underlying forces of change. This is because they are the net product of complex compositional shifts. A compositional perspective on demographic change is important because major shifts occurred during the nineteenth century in the proportion of the national population living close to the land in small communities or living in the towns, living in the south or living in

Click to return to Table of Contents 64 F AMILY PLANNING IN ENGLAND BEFORE THE FERTILITY DECLINE the north, living in a plebeian or in a genteel local environment. Further- more, these compositional shifts, in terms of settlement size, geographic region, and social class, were interrelated, thereby either magnifying or artifi- cially minimizing their effects on the time trends in national aggregate indicators. The significance of the continuity of the greatly delayed marriage of the propertied classes of the southeast during 1846–76 is largely obscured in national aggregate demographic trends because upper-class behavior is submerged in the counter-trend toward slightly earlier marriage among a larger group, the expanding proletarian majority of the towns. It is perfectly possible that this masking effect applies much earlier in time, too. Does 1816 mark the beginning of an entirely new era of widespread control over reproduction, just because that is when the national aggregate measure registers such a change? Malthus’s testimony in the late eighteenth century suggests a more complex picture of a still earlier history of family planning and a more protracted process of reproductive adaptation to the conditions of economic growth. Malthus claimed to be describing the current state of affairs in the 1790s: “the preventive check to population in this country operates, though with varied force, through all the classes of the community.” We also know from his other writings that he maintained that the prudential habit was in danger of falling into disuse among the laboring classes of his own generation, and he believed that an overly generous Poor Law was the principal reason for this. Thus, if we take Malthus seriously as an eyewitness guide to the social distribution and the motivation involved in the prudential control of reproduction, he is telling us that it has long been the habit for prospective marriage partners to weigh carefully their present and future economic position before committing themselves to the institution of marriage and the associated burdens and joys of childrearing. However, he believes that in his own times, in the closing years of the eighteenth century, prudential behavior is practiced more among the middle classes, with whom he himself identifies, than among the laboring poor. Had there indeed been such a differential change in the nuptiality behavior of the prosperous and the poor during the half-century or so before Malthus set down his observations? Hard evidence on social differentials in marriage practices and marriage ages for any place in the early modern period is not abundant; it is equally difficult to find reliable data tracking any possible change across the eighteenth and early nineteenth centuries. The substantial studies by Cham- bers of seventeenth-century Gloucestershire and early-eighteenth-century Nottinghamshire have long provided the only robust measures for rural and agrarian communities. In both cases they indicated that the wealthy married the youngest wives, the poor married the eldest.53 Further evidence tells a similar, but more complex story. Findings relating to the peerage, to the Quakers, and to seventeenth-century Londoners indicate a consider-

Click to return to Table of Contents S IMON SZRETER/ EILIDH GARRETT 65 able departure from common practices in the courtship and marriage behavior of both the upper class and these sections of the middle classes during the seventeenth and the eighteenth centuries.54 The very highest ech- elons of the peerage were married off relatively young as a rule throughout this period. The rest of the peerage tended to marry off their daughters considerably younger than average but their sons slightly older than average, making for a characteristically larger age gap between spouses among the elite than obtained for the rest of the population. This pattern also prevailed among the more substantial and well-established tradesmen of seventeenth- century London. However, both sexes in the seventeenth-century Quaker urban community (chiefly resident in London, Bristol, and Norwich) married very late; and nonurban Quakers also tended to marry a little later than the national average. The Quaker study is also significant in showing an ever-increasing divergence from the rest of the population over the en- suing century and a half of industrialization, as both sexes among nonurban Quakers delayed marriage more and more, exhibiting quite the contrary trend to the rest of the population. Urban Quakers, starting from a position of very late marriage, displayed only a modest fall in marriage ages for both sexes during the eighteenth century, but thereafter a slight rise again in the first half of the nineteenth century.55 As a result all Quaker men and women, from all samples, were marrying remarkably late during the first half of the nineteenth century, judged by national standards.56 Thus, different social groups varied widely in their approach to marriage throughout the early modern period. The late marriage regime, originally identified by Hajnal, has usually been thought of as a single uniform institution, but it may well have functioned in at least two (and very likely more than two) distinct ways for different social groups in early modern England. This point is further illustrated by the recent evidence supporting the “Sharlin thesis”: that throughout the early modern period, migrants to the towns tended to have distinctly lower marriage rates than established residents, interpreted as reflecting their more marginal position in the economy due to poorer skills and social networks.57 In Banbury, the single reconstituted English parish where it has been possible to analyze marriage behavior in detail from 1730 to 1820 according to seven socioeconomic categories (defined by occupation of husband at marriage), a number of distinct trajectories of change have been found. The social elite, for instance, exhibited much later female marriage than any other social group in the mid-eighteenth century, and much earlier marriage than other groups from 1790 through 1815, followed by a sharp rise in marriage age thereafter.58 As noted above, it has usually been argued that the institution of late marriage in early modern rural England was a form of societal control on access to marriage that functioned as a fertility regulator independent of individuals’ conscious decisions.59 The evidence presented here of the di-

Click to return to Table of Contents 66 F AMILY PLANNING IN ENGLAND BEFORE THE FERTILITY DECLINE vergent and changing behavior of propertied social groups indicates a quite different experience. Although they married relatively late, their behavior suggests a conscious appreciation of the importance of marriage timing and a preparedness to vary this significantly if and when circumstances demanded it, presumably to avoid prejudicing social status aspirations and material living standards by undertaking marriage without the means to discharge its responsibilities. A similar deliberative attitude toward the timing of marriage on the part of those administering the early modern Poor Law has also been confirmed by Hindle.60 Thus, an important consequence of the secular fall in plebeian marriage ages during the second half of the eighteenth century documented by Wrigley and Schofield was a historic crossover in the social differentials char- acterizing nuptiality behavior. The available evidence indicates that throughout the seventeenth century most of the propertied class, certainly the landed gentry, married younger than the property-less. However, by the beginning of the nineteenth century the common people, including even those of relatively modest means, were generally marrying much younger than all elements in the propertied classes. Although plebeian marriage age subsequently rose rapidly in the 1830s and 1840s to its pre-1750 levels, Table 2 shows that Ansell’s data record males of the propertied classes by then marrying even later still.61 Never again in modern British history would the middle and upper classes generally marry younger than the working classes. This novel reversal in the demography of social class was, no doubt, an important part of the context in which Malthus wrote his Essay and in which the heated and formative debate was conducted over the prolific poor and the Poor Law. The self-righteous Malthusian position of the wealthier classes, in accusing the poor of recklessness in countenancing early marriage, reflected a recent class reversal of roles in this regard; and the indignation may have been all the more vehement since it expressed the fervor of the recently converted.

TABLE 2 Bachelor average ages at marriage in England and Wales and among the professions c.1800–1906 1800–37 1851 1871 1884–85 1901–06

National average 25.3 26.9 26.4 26.8 27.4 Professions — 30.0 30.0 30.7 33.5 Difference (years) — 3.1 3.5 3.9 5.9

SOURCES: National averages: 1800–37: Wrigley et al. (cited in note 3), p. 159, Table 5.10; 1851–1906: Wrigley and Schofield (cited in note 18), values interpolated from figures given on p. 437, Table 10.3. Professions: 1851 and 1871: Ansell (cited in note 61), p. 45; 1884–85: 48th Annual Report of the Registrar- General, 1885, pp. ix–x; 1901–06: Census of England and Wales, 1911, Vol. 13, Fertility of Marriage, Part 2, Table VII.

Click to return to Table of Contents S IMON SZRETER/ EILIDH GARRETT 67

The whole population of England and Wales increasingly delayed marriage from the 1870s, an insufficiently appreciated yet highly significant aspect of the nation’s fertility decline.62 However, the male members of the upper and middle classes did so to such a disproportionate extent that by 1911 among women who married in their 20s almost a quarter were marrying men over the age of 35, producing a distinctly nonmodern, highly patriarchal age gap between spouses.63 And these couples were supposed to be in the vanguard of the modernizing process! It is certainly significant that men in the social groups with the lowest marital fertility in 1911, the professional and successful commercial classes, apparently began their strategy for family limitation by deferring marriage to a historically unprecedented extent. This analysis provokes a critical question. Why was it that, from the mid-eighteenth century onward in the economically fastest-growing and most prosperous society in the world, the most privileged strata, rather than their less fortunate countrymen, became increasingly conscious of the need to defer marriage? This question cannot be comprehensively addressed here but some general factors can be suggested. First, no doubt it has much to do with the sociology of affluence and with socio-demographic competition for scarce, status-enhancing goods in a socially mobile society. This at any rate was the view taken, in their historical sociological approach to the subject, by Joseph and Olive Banks, the only scholars to date to have analyzed middle-class marriage, and especially male reluctance to marry in the mid- nineteenth century.64 Second, life expectancy probably rose substantially and for the first time differentially (relative to the average) among the upper and middle classes during the later eighteenth century. This may have expressed itself in a longer typical wait for the inheritance of the means to start a comfort- able family life, the expectation that less would be available per inheritor because it was being shared among more survivors, and the need to sacri- fice more for the relatively fixed number of high-status positions available.65 Razzell’s study of social recruitment trends into the higher officer class of the Army, a classic, relatively scarce elite position carrying high honorific value, across this period, c.1760–1835, supports this hypothesis.66 Third, this era witnessed the emergence throughout the middle classes of the ideology and practice of separate spheres, whereby breadwinning became identified as an exclusively adult male role, socially and physically separating the husband from his wife, children, and home. Men of the propertied and aspirant classes increasingly came to perceive their families’ fortunes as resting solely on the success of their own exertions in the market- place for commercial and professional services. That the burden of this responsibility weighed heavily on them is attested to by the extent of the “flight from marriage” that provoked public comment at midcentury on the scandal of “redundant” women. Such surplus unmarried daughters were

Click to return to Table of Contents 68 F AMILY PLANNING IN ENGLAND BEFORE THE FERTILITY DECLINE frequently living with their middle-class fathers or brothers, since, according to the code of separate spheres, they were debarred from supporting themselves by taking a paying job.67 In an age in which credit and mortgages were scarce, postponing marriage until one’s 30s or 40s was the only respectable mechanism available to many Victorian would-be gentlemen to defray the inevitable expenses of marriage. It was an effective tactic to alter the timing between one’s income-earning trajectory and familial outlays. With later male marriage, children arrived after savings had been amassed by the husband and his income level could be expected to rise. Indeed, significant middle-class sources of employment, such as banks, had formal rules at this time prohibiting their young employees from marrying before reaching a certain stage of seniority and associated income level, institutionalizing the calculation that it was reckless to take on familial responsibilities too early in one’s career.68 The point, therefore, is that careful and conscious family planning was being systematically practiced, and behavior adjusted accordingly, by this segment of British society long before they adopted the systematic use of appliance methods of contraception and long before notions of ideal or target family sizes or parity-specific birth control began to circulate in common parlance. Whether middle-class males were delaying marriage only because they wished to postpone the point at which the costs of maintaining a household with children would begin or whether they also realized that by waiting and marrying an older mate they would produce a smaller completed family size remains an issue for further research, though it would seem probable that both motives were involved. One way or the other, reproduction was well within the framework of conscious choice, an attitude of mind that, in Coale’s formula, has tended to be associated only with the “modern” parity-specific contraceptive regulation of births within marriage. Despite the Bankses’ work, the significance of this historic change in the reproductive behavior of the bourgeoisie in Britain has been largely ignored by demographic historians. Perhaps demographers have been slow to see the significance of early- and mid-nineteenth-century middle-class nuptiality restraint because of their preoccupation with the search for the origins of parity-specific fertility control in the last quarter of the century. Thus, this patriarchal form of tailoring one’s fertility to one’s pocket has been too rapidly dismissed as a continuation of traditional prudential marriage practices, supposedly distinct from the subsequent modern developments whereby couples began systematically to restrict the number of births within marriage. Additionally, demographers appear to have assumed that because it is only later female marriage ages that can directly reduce fertility, the issue of male marriage ages is of secondary importance. In fact this may well be doubly wrong. It is likely that, in addition to the socioeconomic significance of extreme male prudence, demographers

Click to return to Table of Contents S IMON SZRETER/ EILIDH GARRETT 69 may have missed a range of other socio-biological effects. Marriages by men in their late 30s and 40s likely resulted in a statistically significant reduction of marital fertility independent of the well-known biological effects of female age. This would have been due to a combination of both sterility attributable to sexually transmitted diseases and, especially later in the marriage, the effect of age-related declines in virility and sexual activity.69

Conclusions: A revisionist history of reproduction and family planning in England and Wales

This analysis of compositional demography suggests a more evolutionary, less dichotomous, more socially variegated, and geographically complex picture of the path of reproductive change in Britain under conditions of economic growth during the later eighteenth and nineteenth centuries.70 Re- productive practices have been continually changing since the eighteenth century in response to the challenges and opportunities of the economic and social transpositions of industrialization. Since industrial and urban growth altered the fortunes of the propertied and the working classes in different, and initially divergent, ways it is far from surprising if their reproductive behavior moved in opposite directions. The marked shift toward very late marriage by the bourgeoisie during the period c.1750–1850 should therefore be seen as just one among several contingent stages of complex change in the reproductive practices of industrializing and postindustrial Britain. For a brief period, during the immediate post–World War II decades, virtually all sections of British society were opting for the same pattern of early marriage, early childbearing, and parity-specific birth control to achieve small families of two or three children.71 The almost uniform demography of this generation of Britain’s citizens, along with those of many other developed countries, was achieved within the ideological and political economy context of a relatively egalitarian, universalist welfare state, reflecting a genuine, if temporary, international consensus on the wisdom of this social ar- rangement. But this period of early and universal marriage that accompanied low marital fertility in the 1950s and 1960s was an extremely recent phenomenon in Britain, dated by Hajnal only to the late 1930s, quite unlike the case in France where, as noted above, such early marriage in the context of fertility restraint had been endorsed a century to a century and half earlier.72 The communities of Britain had arrived at this point through an entirely different trajectory from that observed in France: the earnest puritan campaigns for the reform of manners—personal and sexual morality—seem to have been a much less pronounced feature of the cultural history of Britain’s continental neighbor.

Click to return to Table of Contents 70 F AMILY PLANNING IN ENGLAND BEFORE THE FERTILITY DECLINE

The available historical evidence suggests that the late marriage regime of early modern Britain entailed systematic sexual restraint among young adults up to their mid-20s, the point at which marriage could be realistically anticipated. A sexual culture of this sort might lend itself to restraint after marriage, as well, if and when the need arose. If married couples rarely had recourse to abstinence in the early modern period, this was because there was little need, given the institution of late marriage itself, relatively low survivorship of children, and, especially, the institution of service and apprenticeship for the children of the laboring poor.73 Thus, England’s late-nineteenth and early-twentieth-century regime of extremely late age at marriage should be seen within the context of a culture that had long tended toward self-control in sexual matters.74 As industrialization deepened in Britain in the nineteenth century, it not only created large-scale divergent north–south, regional, class environments, but also fostered a number of distinctive “communication communities,” each reflecting the various means of livelihood open to working-class families because of the great diversity of large-scale industries that characterized different parts of the country.75 The demographic change accompanying industrialization was therefore a protracted, uneven, and nonlinear historical process in which a number of distinct social and cultural groups and communities were altering their reproductive regimes in different ways. Not until the last quarter of the nineteenth century did couples living in diverse environments tend toward smaller family sizes in sufficient numbers that the aggregate, national measures of marital fertility began to register a downward trend. Thinking in terms of compositional demography and in terms of “reproduction,” rather than merely “fertility,” indicates that the 1816 turning point should be given at least equal significance to that of 1876. Instead of focusing on a comparatively short fertility decline, we should be analyzing long phases of reproductive adaptation, ideally decomposing national aggregate trends into the distinctive reproductive regimes of communication communities. According to this viewpoint the downturn of national aggregate fertility trends at the end of the 1870s does not necessarily signal the emergence of novel ideas and attitudes toward family planning or even the widespread uptake of new methods of birth control. Indeed, the first author has previously argued that there is no plausible evidence for Britain of widespread adoption of appliance methods of birth control before the inter-war decades of the twentieth century. He has pointed out that careful scrutiny of the available quantitative, qualitative, and survey evidence indicates that the traditional methods of abstinence and withdrawal, with some recourse to abortion where these failed, were probably the main techniques of birth control used within marriage before World War I.76 There was much more continuity, in terms of the methods used for control over reproduction, be-

Click to return to Table of Contents S IMON SZRETER/ EILIDH GARRETT 71 tween the seventeenth and eighteenth centuries and the late nineteenth and even early twentieth centuries, than has previously been considered. Change in fertility-regulating methods and practices between the late nineteenth and the late twentieth centuries was gradual and evolutionary for most segments of British society, with traditional methods persisting in use for many decades alongside the gradual adoption of appliance and, ultimately, pharmaceutical methods.77 Parity-specific birth control had become the norm in the West by the 1950s and 1960s when demographers began to construct their methodology for measuring this form of behavior. But it had probably only very recently emerged as such in Britain, at least. Demographers’ efforts to make clear analytical distinctions between starting, spacing, and stopping may, ultimately, be less helpful than a recognition that the actors themselves did not make these distinctions. It may be more productive to study the expe- dient and locally variable nature of the methods used and the manner in which they represented organic adaptations and extensions of preexisting practices and understandings of reproductive management. All communities practice some form of control over their reproduction (see references in endnote 9). Dramatic changes in the extent and mode of control reflect and require significant alterations in the socioeconomic and cultural environment that alter perceptions of the relative costs and benefits of childrearing. On the mistaken assumption that parity-specific birth control is the sole key to effective family planning, demography in the years since World War II may be seen as having taken a long social science detour. Over half a century has passed in looking for and promoting the emergence of this one form of behavior, as a cross-cultural uniform key to demographic transition. Meanwhile the most valuable and genuinely transferable policy lesson from the European past, or at least Britain’s part of it, has been steadily ignored. To reverse the principal emphasis of a much-quoted article by John Knodel and Etienne van de Walle published in this journal two decades ago, which aimed to cull the lessons of European history for policymakers, we claim here that Britain’s history, at least, gives no sanction to the long- dominant policy notion that it was the relatively sudden spreading or “diffusing” of parity-specific birth control through a population in the space of a generation that has been the most important fact about the relationship between reproduction and economic change during the last two centuries.78 Instead, indigenous understandings of reproduction entailed for centuries the important fertility-reducing practices of prudential marriage and sexual restraint. These methods evolved so as to deliver an increasingly tight regulation of the reproduction of the upper and middle classes during the late eighteenth and nineteenth centuries. This occurred gradually and episodi- cally, unevenly geographically and socially, accompanied by keen public

Click to return to Table of Contents 72 F AMILY PLANNING IN ENGLAND BEFORE THE FERTILITY DECLINE and private debate as agents negotiated their changing predicaments and forged new practices, always drawing from and adapting an extant matrix of attitudes and practices. Initially—and, indeed, well into the twentieth century—this included ever-more extreme extensions of the established methods of prudential marriage and sexual restraint. In arguing that an approach to fertility change which seeks to inte- grate its study within the more encompassing concept of reproduction can result in a significantly different understanding of the chronology and the relationship between economic growth and family planning, we have adopted a broader definition of family planning than is conventional. This is necessary to appreciate the varied practices that different cultures and communities have contrived, in order to manage their reproduction. Par- ity-specific control of marital fertility, or stopping behavior using contraception within marriage, has become the practice of many peoples around the world during the second half of the twentieth century, but the English case examined here provides evidence of careful and rigorous family planning using other means for a century and a half before stopping, via contraception, became widespread in the 1940s and 1950s. If we persist in looking for the origins of family planning in the past using only the tools to detect parity-specific control, then we will remain blind to the significance of these other forms of reproductive management and we will continue to indulge in the modernist bias that only since 1870 in most countries have people achieved successful control over human fertility. This in turn imbues us with a false sense of distance from and superiority over earlier generations and over other contemporary peoples, who apparently lack this “supremacy over nature” and defining characteristic of what it is to be modern. Further, unless we attempt to delineate the nature of the varied subcultures and communities contributing to the nation, the unit for which historical demographic statistics tend to be most readily available, the resort to the national aggregate trend will continue to predomi- nate so that we will remain with the misleadingly simple certainties, such as the dating of a supposed national fertility revolution in England to 1876. The story is much more interesting than that.

Notes

The authors acknowledge with gratitude the ning ESRC-funded project that will shortly comments made by participants at the semi- be published: E. Garrett, A. Reid, K. Schurer, nar of the Cambridge Group for the History and S. Szreter, Population Change in Context: of Population in January 2000 and at the Place, Class and Demography in England and British Society for Population Studies Malthus Wales 1891–1911 (Cambridge: Cambridge Bicentennial Conference at Cambridge in University Press, forthcoming 2001). The 1998, where earlier versions of this article data drawn from this project have been re- were presented. Much of the work discussed produced with permission of the Controllers here has been completed within a long-run- of H. M. Stationery Office.

Click to return to Table of Contents S IMON SZRETER/ EILIDH GARRETT 73

1 D. L. Anderton and L. L. Bean, “Birth control in Britain 1925–50,” Past and Present spacing and fertility limitation: A behavioral (forthcoming 2000/2001). analysis of a nineteenth century frontier popu- 8 S. Szreter, “The idea of demographic lation,” Demography 22 (1985): 169–183; P. A. transition and the study of fertility change: A David and T. A. Mroz, “Evidence of fertility critical intellectual history,” Population and De- regulation among rural French villagers, 1749– velopment Review 19 (1993): 659–701, esp. pp. 1789,” European Journal of Population 5 (1989): 667–675; see also D. Hodgson, “Demography 1–26; E. M. Garrett, “The trials of labour: as social science and policy science,” Popula- Motherhood versus employment in a nine- tion and Development Review 9 (1983): 1–34; P. teenth-century textile centre,” Continuity and Demeny, “Social science and population Change 5 (1990): 121–154; S. Szreter, Fertility, policy,” Population and Development Review 14 Class and Gender in Britain, 1860–1940 (Cam- (1988): 451–479. The most important work in bridge: Cambridge University Press, 1996), pp. this area advocating population curbs in order 377–382; J. C. Caldwell, I. O. Orubuloye, and to achieve successful capital accumulation and P. Caldwell, “Fertility decline in Africa: A new industrialization was carried out at the type of transition?” Population and Development Princeton Office of Population Research un- Review 18 (1992): 211–242. der Frank W. Notestein’s directorship, culmi- 2 For criticism of the more conventional nating in the highly influential publication by measures, Ig , M, and m, see for example T. W. Ansley J. Coale and Edgar M. Hoover, Popula- Guinnane, B. S. Okun, and J. Trussell, “What tion Growth and Economic Development in Low-In- do we know about the timing of fertility tran- come Countries (Princeton: Princeton University sitions in Europe?” Demography 31 (1994): 1– Press, 1958). 20. 9 A. M. Carr-Saunders, The Population 3 E. A. Wrigley, R. S. Davies, J. E. Oep- Problem (Oxford: Oxford University Press, pen, and R. S. Schofield, English Population His- 1922) provided demographers with an exhaus- tory from Family Reconstitution, 1580–1837 (Cam- tive, widely read inventory of nineteenth-cen- bridge: Cambridge University Press, 1997), tury ethnographic and early anthropological Table 7.1, p. 355. material of relevance to their interests. Many 4 Census of England and Wales, 1911, other important compilations and collections Volume XIII, Fertility of Marriage, Part 2 (HMSO, have since followed, such as the survey by 1923). Moni Nag, Factors Affecting Human Fertility in Non-Industrial Societies: A Cross-Cultural Study 5 E. Garrett, A. Reid, K. Schurer, and S. (New Haven: Human Relations Area Files Szreter, Population Change in Context: Place, Class Press, 1976). Above all, the influential article and Demography in England and Wales 1891–1911 coauthored by Kingsley Davis, one of the lead- (Cambridge: Cambridge University Press, ing influences in postwar American demogra- forthcoming 2001); E. Garrett and A. Reid, phy, clearly demonstrated and attempted to “Thinking of England and taking care: Family classify demographers’ knowledge of a wide building strategies and infant mortality in Eng- range of fertility-reducing institutions and land and Wales 1891–1911,” International Jour- practices: K. Davis and J. Blake, “Social struc- nal of Population Geography 1 (1995): 69–102. ture and fertility: An analytic framework,” Eco- 6 M. Anderson, “Highly restricted fertil- nomic Development and Cultural Change 4 (1956): ity: Very small families in the British fertility 211–235. decline,” Population Studies 52 (1998): 177–199. 10 L. Henry, “Some data on natural fer- See Anderton and Bean, “Birth spacing,” cited tility,” Eugenics Quarterly 8 (1961): 81–91; see in note 1; Szreter, Fertility, Class and Gender, also his Anciennes familles génévoises: étude cited in note 1, pp. 377–382. démographique (Paris: Presses Universitaires de 7 K. Fisher, “Negotiating the fertility tran- France, 1956); A. J. Coale, “Age patterns of sition: Birth control practices and family plan- marriage,” Population Studies 25 (1971): 193– ning strategies in Britain, 1925–50,” Population 214; A. J. Coale and T. J. Trussell, “Model fer- and Development Review 26 (forthcoming June tility schedules: Variations in the age structure 2000); K. Fisher, “’She was quite satisfied with of childbearing in human populations,” Popu- the arrangements I made’: Gender and birth lation Index 40 (1974): 185–258, and “Erra-

Click to return to Table of Contents 74 F AMILY PLANNING IN ENGLAND BEFORE THE FERTILITY DECLINE tum,” Population Index 41 (1975): 572; A. J. of Atonement: The Influence of Evangelicalism on Coale and T. J. Trussell, “Technical note: Find- Social and Economic Thought 1785–1865 (Oxford: ing the two parameters that specify a model Clarendon Press, 1988); A. M. C. Waterman, schedule of marital fertility,” Population Index Revolution, Economics and Religion: Christian Po- 44 (1978): 203–213. litical Economy 1798–1833 (Cambridge: Cam- 11 A. J. Coale, “Factors associated with bridge University Press, 1991); and M. Mason, the development of low fertility: An historic The Making of Victorian Sexual Attitudes (Oxford: survey,” in Proceedings of the World Population Oxford University Press, 1994). Conference, Belgrade 1965, vol. II (New York: 18 P. Laslett and R. Wall (eds.), Household United Nations, 1967), pp. 205–209. and Family in Past Time (Cambridge: Cambridge 12 A. J. Coale, “The demographic transi- Univesity Press, 1972); E. A. Wrigley and R. S. tion reconsidered,” in International Population Schofield, The Population History of England Conference, Liège, vol. I (Liège: IUSSP, 1973), pp. 1541–1871: A Reconstruction (Cambridge, MA: 53–72. Harvard University Press, 1981); A. Kussmaul, Servants in Husbandry in Early Modern England 13 P. Langford, A Polite and Commercial (Cambridge: Cambridge University Press, People: England 1727–83 (Oxford: Oxford Uni- 1981); Wrigley et al., English Population History, versity Press, 1992). cited in note 3. 14 As cited in P. Kane, Victorian Families 19 P. Kane, The Second Billion: Population in Fact and Fiction (London: Macmillan, 1995), and Family Planning in China (New York: Pen- p. 134. Kane further explores an antipathy to guin Books, 1987), pp. 30–32. numerous offspring in this period. 20 J. Hajnal, “European marriage patterns 15 Thomas Robert Malthus, An Essay on in perspective,” in D. V. Glass and D. E. C. the Principle of Population, 1798, ch. IV, from The Eversley (eds.), Population in History: Essays in Works of Thomas Robert Malthus, Vol. 1, ed. E. Historical Demography (London: Edward Arnold, A. Wrigley and D. Souden (London: William 1965), pp. 101–143; the institutional context Pickering), pp. 26–28. was clearly laid out in R. M. Smith, “Fertility, 16 S. Hindle, “The problem of pauper mar- economy, and household formation in Eng- riage in seventeenth-century England,” Trans- land over three centuries,” Population and De- actions of the Royal Historical Society 6 (1998): 71– velopment Review 7 (1981): 595–622. 89. 21 E. A. Wrigley, “Fertility strategy for the 17 J. A. Banks originally drew attention individual and the group,” in C. Tilly (ed.), His- to the importance of this literature and its pre- torical Studies of Changing Fertility (Princeton: occupations with “The ‘proper’ time to marry,” Princeton University Press, 1978), pp. 133–154; as he titled chapter 3 of his classic account, Pros- Smith, “Fertility, economy, and household,” perity and Parenthood: A Study of Family Planning cited in note 20, pp. 618–619. Among the Victorian Middle Classes (London: 22 Wrigley and Schofield, Population His- Routledge & Kegan Paul, 1954). On the ex- tory, cited in note 18, pp. 236–248, 476; R. tensive literature of the period, see J. A. Banks Schofield, “Family structure, demographic and D. V. Glass, “A list of books, pamphlets behaviour, and economic growth,” in J. Walter and articles on the population question pub- and R. Schofield (eds.), Famine, Disease and the lished in Britain in the period 1793 to 1880,” Social Order in Early Modern Society (Cambridge: in D. V. Glass (ed.), Introduction to Malthus (New Cambridge University Press, 1989), pp. 279– York: Wiley, 1953). For further and more de- 304. tailed accounts of both clerics and the dismal scientists who took up the debate with 23 Kussmaul, Servants in Husbandry, cited Malthus, see in particular R. A. Soloway, Prel- in note 18, p. 26. ates and People: Ecclesiastical Social Thought in Eng- 24 K. D. M. Snell, Annals of the Labouring land 1783–1852 (London: Routledge & Kegan Poor: Social Change and Agrarian England, 1660– Paul, 1969); J. A. Banks, Victorian Values: Secu- 1900 (Cambridge: Cambridge University Press larism and the Size of Families (London: Rout- 1985), esp. chs. 1–2; Kussmaul, Servants in Hus- ledge & Kegan Paul, 1981); B. Hilton, The Age bandry, cited in note 18, Part III.

Click to return to Table of Contents S IMON SZRETER/ EILIDH GARRETT 75

25 We accept that the decline of live-in their homes) and, instead, insisting that the service may well have been one of the rea- unemployed and their families leave their sons for the fall in marriage age down to 1816, homes and reside in the prison-like, sex-seg- on the argument that young women elected regated workhouses, where they would re- for earlier marriage since it provided an alter- ceive “indoor relief” in return for compulsory native means to attain economic independence work. from the parental home. However, the oppo- 29 R. Schofield, “English marriage pat- site effect is being argued for here for the pe- terns revisited,” Journal of Family History 10 riod after 1816. This is not necessarily logically (1985): 2–20, Table 2, p. 10. Schofield’s best inconsistent, since it is perfectly plausible that estimates of average female age at marriage in- the decline in live-in service simultaneously dicate a value of 22.6 years for those marry- had the potential to affect marriage age in these ing in the 1810s (the birth cohort of 1791) and two diametrically opposed directions. Whether a value of 25.2 years for those marrying in the or not the first or the second effect was domi- 1840s (the birth cohort of 1816). This prescient nant would then depend on other significant article by Schofield anticipated many of the factors in the historically changing context. The themes taken up here, in particular an inter- two following text paragraphs discuss such im- est in the possible implications of compositional portant and relevant changes between the pre- differences in nuptiality (pp. 5 and 17) and the and post-1816 periods. point that the early modern institution of late 26 Hence, David Levine’s Family Formation marriage had a history and was not just a con- in an Age of Nascent Capitalism (New York: Aca- stant (p. 15). However, Schofield, like Levine demic Press, 1977), ch. 5, found signs of mari- (see note 75 below), remained impressed with tal fertility control among Shepshed cottage in- the “demographic transition” perspective, en- dustry workers during the economic difficulties visaging an epochal shift from traditional to that followed the Napoleonic Wars. modern forms of fertility control around the 27 In 1795 the Justices of the Peace in 1870s (pp. 15–16). Speenhamland parish in Hampshire published 30 C. H. Feinstein, “Pessimism perpetu- a scale of allowances in aid of wages for un- ated: Real wages and the standard of living in employed men, setting out the amounts they Britain during and after the industrial revolu- would receive to supplement their incomes in tion,” Journal of Economic History 58 (1998): order to maintain their families, calculated ac- 625–658. cording to the price of bread and the number 31 Ibid. of dependents in the household. The system was never formally adopted on a national ba- 32 Wrigley and Schofield, Population His- sis but was deemed, especially by its opponents, tory, cited in note 18, Table 10.3, p. 437 gives to be typical of an over-generous approach that figures for female singulate mean age at mar- characterized the closing decades of the eigh- riage calculated from the decennial census, teenth century and the period of the Napole- which show age at marriage falling from 25.77 onic Wars before reaction set in. For an acces- years in 1851 to 25.13 years in 1871 (subse- sible introduction to this complex and quently rising back to 26.27 years by 1901). much-debated era of British history, see J. D. 33 E. van de Walle, The Female Population Marshall, The Old Poor Law, 1795–1834 (Basing- of France in the Nineteenth Century: A Reconstruc- stoke: Macmillan, 2nd edition, 1985). tion of 82 Départements (Princeton: Princeton 28 P. Lindert, “Unequal living standards,” University Press, 1974), ch. 7, section IV; D. in R. Floud and D. N. McCloskey (eds.), The Weir, “New estimates of nuptiality and mari- Economic History of Britain since 1700, Vol. I, tal fertility in France, 1740–1911,” Population 1700–1860 (Cambridge: Cambridge University Studies 48 (1994): 307–331, esp. pp. 322–323. Press, 2nd edition, 1994), pp. 368–372. 34 H. Bergues et al., La prévention des Marshall, The Old Poor Law, cited in note 27, naissances dans la famille: ses origines dans les temps Table 1, pp. 28–29. “Less eligibility” refers to modernes (INED, Travaux et documents, cahier the deliberate policy of curtailing “outdoor re- no. 35, Paris: Presses Universitaires de France, lief” (cash payments to the unemployed in 1960), pp. 253–307, 317–321, 345–355, 369–

Click to return to Table of Contents 76 F AMILY PLANNING IN ENGLAND BEFORE THE FERTILITY DECLINE

382 (with thanks to Jacques Beauroy, for lend- population 1840–1870 (Paris: Presses Universi- ing his copy of this book); J-L. Flandrin, “Con- taires de France, 1981), pp. 63–64. See also traception, marriage and sexual relations in the next endnote. Christian West,” in R. Forster and O. Ranum 43 Szreter, Fertility, Class and Gender, (eds.), The Biology of Man in History (selections cited in note 1, p. 410. France at this time from Annales, trans. by E. Forster and P. witnessed substantial, conservative moral Ranum), pp. 23–47; J-L. Flandrin, Families in mobilization over issues of birth control and Former Times: Kinship, Household and Sexuality, sexuality, with the passing in July 1920 of a trans. by R. Southern (Cambridge: Cambridge law against contraception. However, this had University Press, 1979); G. Santow, “Coitus in- virtually no connection with French thought terruptus and the control of natural fertility,” or policy during the first half of the nine- Population Studies 49 (1995): 19–43. teenth century. Its origins lay in a distinc- 35 On Britain, see below, notes 76 and 77. tive, late-nineteenth-century pronatalist 36 E. A. Wrigley, “The fall of marital fer- anxiety at France’s low birth rate, a discourse tility in nineteenth-century France: Exemplar on “dépopulation et décadence,” that dated or exception?” in E. A. Wrigley, People, Cities from defeat in the Franco–Prussian War of and Wealth: The Transformation of Traditional So- 1870–71 and was, by 1920, exacerbated by ciety (Oxford: Blackwell, 1987), pp. 270–321, the massive losses sustained in World War I. esp. pp. 287–291; Bergues et al., La prévention See M. S. Teitelbaum and J. M. Winter, The des naissances, cited in note 34, p. 319. Fear of Population Decline (Orlando: Academic Press, 1985), pp. 18–30. 37 Weir, “New estimates,” cited in note 33, pp. 317–318. 44 The principal evidence supporting the 38 J. A. Goldstone, “The demographic inference that sexual restraint among young revolution in England: A re-examination,” adults was deeply institutionalized has been Population Studies 40 (1986): 5–33. the repeated findings from a number of historians using various sources that illegitimacy 39 N. E. Himes, “Editor’s introduction” to rates in England were comparatively low and Francis Place, Illustrations and Proofs of the Prin- that they fluctuated in sympathy with the mar- ciple of Population (Boston: Houghton, Mifflin, riage rate, rather than as an alternative to it: 1930; first edition 1822); N. E. Himes, A Medi- P. Laslett and K. Oosterveen, “Long-term cal History of Contraception (Baltimore: Williams trends in bastardy in England: A study of the & Wilkins, 1936), esp. p. 213; see also Banks illegitimacy figures in the parish registers and and Glass, “A list of books, pamphlets and ar- in the reports of the registrar general, 1561– ticles,” cited in note 17, esp. Section 8, “For 1960,” Population Studies 27 (1973): 255–286; and against family limitation.” P. Laslett, K. Oosterveen, and R. Smith (eds.), 40 From 1538 to 1750 prenuptial preg- Bastardy and Its Comparative History (London: nancies accounted for about 15–30 percent of Edward Arnold, 1980); R. Adair, Courtship, Il- first births in the Cambridge Group’s sample legitimacy and Marriage in Early Modern England of 26 reconstituted parishes, rising to just over (Manchester: Manchester University Press, 35 percent from 1775 to 1837: Wrigley et al., 1996). Also see note 40 above. The same per- English Population History, cited in note 3, Table haps cannot be said for the Welsh or the High- 7.27, p. 421. In addition to the relatively low land Scots (but probably pertains to the low- prevalence of prenuptial pregnancy in prein- land Scots). On Wales and Scotland, see S. dustrial times, about half of those who were Szreter, “Falling fertilities and changing sexu- prenuptially pregnant gave birth 6–9 months alities in Europe since c.1850: A comparative after marriage, suggesting that sex had only survey of national demographic patterns,” in been undertaken when marriage was already L. A. Hall, F. Eder, and G. Hekma (eds.), Sexual imminent. Cultures in Europe, Vol. II, Themes in Sexuality 41 Mason, Victorian Sexual Attitudes, cited (Manchester: Manchester University Press, in note 17. 1999), pp. 159–194, esp. pp. 163–164. 42 Y. Charbit, Du malthusianisme au 45 On the early modern era, see M. populationnisme: les économistes français et la Ingram, “The reformation of manners in early

Click to return to Table of Contents S IMON SZRETER/ EILIDH GARRETT 77 modern England,” in P. Griffiths, A. Fox, and Gender and the Making of the British Working Class S. Hindle (eds.), The Experience of Authority in (Berkeley: University of California Press, Early Modern England (Basingstoke: Macmillan, 1995); and D. Wahrman, “Percy’s prologue: 1996), pp. 47–88; K. Wrightson and D. Levine, From gender play to gender panic in eigh- Poverty and Piety in an English Village: Terling teenth-century England,” Past and Present 159 1525–1700 (Oxford: Clarendon Press, Revised (1998): 113–160. ed., 1995), “Postscript.” On the late-eigh- 47 On male self-restraint, see Mason, Vic- teenth-century and Victorian movements, see, torian Sexuality, cited in note 46. for instance: M. J. Quinlan, Victorian Prelude: A 48 S. Szreter, “Economic growth, disrup- History of English Manners 1700–1830 (New York: tion, deprivation, disease, and death: On the Columbia University Press, 1941); F. K. Brown, importance of the politics of public health for Fathers of the Victorians: The Age of Wilberforce development,” Population and Development Re- (Cambridge: Cambridge University Press, view 23 (1997): 693–728. 1961); Mason, Victorian Sexual Attitudes, cited in note 17, esp. ch. 2, section 1; and on its long- 49 The eight counties in question were lasting influence, see S. Szreter, “Victorian Brit- Cheshire, Derbyshire, Lancashire, Leicester- ain, 1837–1963: Towards a social history of shire, Nottinghamshire, Staffordshire, War- sexuality,” Journal of Victorian Culture 1 (1996): wickshire, and Yorkshire. Figures for the eigh- 136–149. teenth century from Langford, Polite and Commercial People, cited in note 13, p. 673, Table 46 Szreter, Fertility, Class and Gender, cited 7, with additional calculations for the nine- in note 1, pp. 392–393. For studies of the his- teenth century from B. R. Mitchell and P. tory of sexuality that broadly support the in- Deane, Abstract of British Historical Statistics terpretation of a long-standing culture of (Cambridge: Cambridge University Press, sexual restraint among young adults tempo- 1962), pp. 6, 20–22. The combined midland rarily losing its relevance for plebeians during and northern counties would have represented the second half of the eighteenth century and almost 45 percent of the total in 1901 had therefore eliciting the strong reassertions of the Northumberland and Durham also been in- need for self-control found among such late- cluded with the eight other counties. eighteenth-century texts as those of Hannah More, Thomas Malthus, and their ilk, see R. 50 These calculations are derived from fig- Porter, “Mixed feelings: The Enlightenment ures to be presented in Garrett et al., Popula- and sexuality in eighteenth-century Britain,” tion Change in Context, cited in note 5. The defi- in P.-G. Boucé (ed.), Sexuality in Eighteenth-Cen- nition of the south and the north in the text tury Britain (Manchester: Manchester Univer- here is in terms of the 11 regions used by the sity Press, 1982), pp. 1–27; J. R. Gillis, For Bet- Registrar-General in 1921 to divide England ter, for Worse: British Marriages, 1600 to the Present and Wales, with “the South” comprising 5 re- (Oxford: Oxford University Press, 1985); T. gions: London, South-East, South-West, East Laqueur, “Orgasm, generation and the politics Anglia, and South Midlands. “The north” (in- of reproductive biology,” in C. Gallagher and cluding the industrial midlands) is defined here T. Laqueur (eds.), The Making of the Modern as six regions: West Midlands, North Midlands, Body: Sexuality and Society in the Nineteenth Cen- Yorkshire, North-West, North, and Wales. tury (Berkeley: University of California Press, 51 S. Szreter and G. Mooney, “Urbanisa- 1987), pp. 1–41; I. McCalman, Radical Under- tion, mortality and the standard of living de- world: Prophets, Revolutionaries and Pornographers bate: New estimates of the expectation of life in London, 1795–1840 (Cambridge: Cambridge at birth in nineteenth-century British cities,” University Press, 1988); G. J. Barker-Benfield, Economic History Review 50 (1998): 84–112. The Culture of Sensibility: Sex and Society in Eigh- There would, of course, be many localized ex- teenth-Century Britain (Chicago: University of ceptions to these regional generalizations. Chicago Press, 1992); M. Mason, The Making Poorer districts in London, such as the East of Victorian Sexuality: Sexual Behaviour and Its Un- End, would inevitably have had very high derstanding (Oxford: Oxford University Press, mortality; and Reay’s study of the Blean dis- 1994); Mason, Victorian Sexual Attitudes, cited trict of rural Kent indicates a deteriorating in note 17; A. Clark, The Struggle for the Breeches: health record there from 1834 to 1880, though

Click to return to Table of Contents 78 F AMILY PLANNING IN ENGLAND BEFORE THE FERTILITY DECLINE the rates of infant mortality reported there Lynch, “The European marriage pattern in the were still much lower in 1880 than those typi- cities: Variations on a theme by Hajnal,” Jour- cal in northern towns: B. Reay, Microhistories: nal of Family History 16 (1991): 79–96; C. Gal- Demography, Society and Culture in Rural England, ley, “A model of early modern urban demog- 1800–1930 (Cambridge: Cambridge University raphy,” Economic History Review 48 (1995): Press, 1996), ch. 3. On differential mortality 448–469; S. King, “Dying with style: Infant in London and recent progress with its mea- death and its context in a rural industrial town- surement, see G. Mooney, B. Luckin, and A. ship 1650–1830,” Social History of Medicine 10 Tanner, “Patient pathways: Solving the prob- (1997): 3–24. lem of institutional mortality in London dur- 58 S. Lauricella, “Economic and social in- ing the later nineteenth century,” Social His- fluences on marriage in Banbury, 1730–1841” tory of Medicine 12 (1999): 227–269. (unpublished Ph.D. thesis, University of Cam- 52 This is confirmed by the only evidence bridge, 1997), pp. 63–91. for agricultural communities in the rural South 59 Wrigley, “Fertility strategy,” cited in for this period: Reay, Microhistories, cited in note note 21, pp. 133–154. 51, ch. 2, Table 2.13. In the Blean district of Kent, Reay found a marked disinclination to 60 Hindle, “The problem of pauper mar- marry starting around the time of the enact- riage,” cited in note 16. ment of the New Poor Law, 1834, and con- 61 C. Ansell, Jr., On the Rate of Mortality in tinuing across the subsequent period to 1880. Early Periods of Life, the Age at Marriage, the Num- He also found evidence of consistent curtail- ber of Children to a Marriage, the Length of Gen- ment of marital fertility in these rural commu- eration, and Other Statistics of Families in the Up- nities from the same date. per and Professional Classes (1874), p. 45. Ansell’s 53 J. D. Chambers “The course of popu- data referred to the period 1841–71. lation change,” in Glass and Eversley (eds.), 62 Szreter, Fertility, Class and Gender, cited Population in History, cited in note 20, p. 332, in note 1, pp. 382–398. Tables 1 and 2. 63 Census of England and Wales, 1911, 54 T. H. Hollingsworth, “The demography Volume 13, Fertility of Marriage, Part 2 (Lon- of the British peerage,” Supplement to Popu- don: HMSO, 1923), Table VII, p. xvii. lation Studies 18 (1964); R. T. Vann and D. E. C. 64 Banks, Prosperity and Parenthood, cited Eversley, Friends in Life and Death: The British in note 17, chs. 3–6; J. A. Banks and O. Banks, and Irish Quakers in the Demographic Transition Feminism and Family Planning in Victorian Eng- 1650–1900 (Cambridge: Cambridge University land: Studies in the Life of Women (Liverpool: Press, 1992), ch. 3; V. Brodsky Elliott, “Single Liverpool University Press, 1964), pp. 29–30 women in the London marriage market: Age, and ch. 3; Banks, Victorian Values, cited in note status and mobility, 1598–1619,” in R. B. 17. The Bankses’ thesis documented the spi- Outhwaite (ed.), Marriage and Society: Studies in raling consumption aspirations of the mid-Vic- the Social History of Marriage (New York: St. torian propertied classes, emphasizing the way Martin’s Press, 1982), pp. 81–100, esp. Table in which competition to enter the security of II, p. 85. the professional vocations drove up the costs 55 Vann and Eversley, Friends in Life and of education so that restraining family size be- Death, cited in note 54, ch. 3, esp. Table 3.8. came critical. Recently John Tosh has also ad- These urban Quakers were not resident in in- dressed this issue, though focusing primarily dustrializing northern cities but mainly drawn on the closing decades of the nineteenth cen- from the flourishing ancient port of Bristol. tury: J. Tosh, A Man’s Place: Masculinity and the 56 Ibid., Table 3.8. Middle-Class Home in Victorian England (New Haven, CT: Yale University Press, 1999), ch. 8. 57 A. Sharlin, “Natural decrease in early modern cities: A reconsideration,” Past and 65 Hollingsworth, “The demography of Present 79 (1978): 126–138; J. de Vries, Euro- the British peerage,” cited in note 54. pean Urbanization 1500–1800 (Cambridge, MA: 66 Razzell found, not surprisingly during Harvard University Press, 1984), p. 190; K. this period of population growth, warfare, and

Click to return to Table of Contents S IMON SZRETER/ EILIDH GARRETT 79 expanding imperial commitments, that be- Attitudes and Lifestyles (London: Penguin Books, tween the later eighteenth century and the 1994), pp. 136–140. Third, and interacting 1830s both the lower-status Indian Army and with the first two phenomena, the encompass- the higher-status British Home Army more ing “Victorian” culture of sexual inhibition and than doubled their officer complements at all anxiety, which would have contributed fur- levels. However, in this unreformed honorific ther to male functional impotence (see the dis- career, where rank was bought not merited, cussion in Szreter, Fertility, Class and Gender, the principal additional recruits were not de- cited in note 1, pp. 410–411, 561–562; and rived from the expanding middle classes but Bancroft, Human Sexuality, pp. 362–364, 392– disproportionately from among the sons of the 395, 655–657). Fourth, and again interacting, longer-living landed gentry and the aristocracy. the incidence of sexually transmitted diseases The latter were so keen to avoid actual service that might be presumed to be higher among abroad in unhealthy climates that there was a later-marrying males, since the principal cli- roaring black market in commissions in the ents of prostitutes were unmarried men (see Home Army and an unseemly scramble among Tosh, A Man’s Place, cited in note 64, pp. 130– the underemployed elite—those with titles to 131; Mason, Victorian Sexuality, cited in note their names—to acquire the highest ranks in 46, p. 102). The incidence of sexually trans- the most fashionable regiments. P. E. Razzell, mitted diseases was officially estimated in 1911 “Social origins of officers in the Indian and Brit- to have been higher among middle-class males ish Home Army: 1758–1962,” British Journal than among all but the poorest working-class of Sociology 14 (1963): 248–260, esp. Tables 1, men. Their prolonged periods of bachelorhood 7, 9, 10. We are grateful to one of the anony- may well have been the primary cause of this, mous reviewers for alerting us to the relevance effectively increasing their period at risk be- of this rarely cited but important and detailed fore marriage: see Szreter, Fertility, Class and study. Gender, cited in note 1, pp. 393–394. There is 67 The classic study of the genesis of the now a fifth hypothesis under scientific inves- ideology of separate spheres among the middle tigation: that male sperm quality deteriorates classes in Britain is L. Davidoff and C. Hall, significantly above the age of 39 (The Indepen- Family Fortunes: Men and Women of the English dent newspaper, 3 July 1999, reporting the re- Middle Class 1780–1850 (Chicago: University of search of Dr. Orhan Bukulmez of Hacettepe Chicago Press, 1987). On the “redundant University, Ankara, Turkey). woman” issue, see Banks and Banks, Feminisim 70 Garrett et al., Population Change in Con- and Family Planning, cited in note 64, pp. 27– text, cited in note 5. 34. 71 This was never, of course, a socially 68 See for instance, G. Jones, British Mul- uniform pattern even in the 1950s and 1960s tinational Banking 1830–1990 (Oxford: Claren- in Britain, Western Europe, or North America. don Press, 1993), p. 49. On important exceptions documented by urban sociologists at the time see, for instance, 69 The effect of age-related declines in vi- M. D. Young and P. Wilmott, Family and Kin- rility and sexual activity is attributable to at ship in East London (Penguin Books, 1957); and least four interrelated influences. First, male L. Rainwater, And the Poor Get Children: Sex, Con- erectile dysfunction, which Bancroft reports to traception, and Family Planning in the Working affect 10 percent of men by age 55 and 20 per- Class (Chicago: Quadrangle, 1960). cent by age 65: J. Bancroft, Human Sexuality and Its Problems (Edinburgh: Churchill Living- 72 J. Hajnal, “Age at marriage and pro- stone, 2nd edition, 1989), p. 288. Second, the portions marrying,” Population Studies 7 (1953): declining inclination to sexual activity in both 111–136. sexes with age. The most recent detailed evi- 73 Nevertheless, there has always been dence on this for Britain in the early 1990s in- evidence in the demographic record that in dicates a sharp fall, from six to three acts per some periods early modern couples did restrain month, among both married and cohabiting the fertility of their marriages, presumably by couples during their 40s: K. Wellings, Sexual abstinence or withdrawal. Wrigley emphasized Behaviour in Britain: The National Survey of Sexual precisely this finding when presenting the re-

Click to return to Table of Contents

80 F AMILY PLANNING IN ENGLAND BEFORE THE FERTILITY DECLINE

sults of his first-ever British family reconstitu- in nineteenth-century Britain quite at vari- tion study of Colyton: E. A. Wrigley, “Family ance with this insight by arguing that the fer- limitation in pre-industrial England,” Economic tility decline among all working-class com- History Review 18 (1966): 82–109; see also the munities could be subsumed within a single results found by Levine and by Reay, cited explanatory model of diffusing “respectabil- above in notes 26 and 51 respectively. ity.” On the contrary, we see the composi- 74 Szreter, Fertility, Class and Gender, cited tional diversity point as becoming even more in note 1, pp. 389–398; and see references in significant over the period 1815–1914 than notes 44–46 above. over the previous century (for an extensive critique of Levine’s interpretation, see 75 On the concept of communication Szreter, Fertility, Class and Gender, cited in communities and its application to the study note 1, pp. 55–60). of fertility change, see Szreter, Fertility, Class and Gender, cited in note 1, pp. 533–558 and 76 Szreter, Fertility, Class and Gender, cited Figure 7.1 (p. 312). A communication com- in note 1, pp. 335–350, 382–398, 398–439. munity is a collectivity of similarly socialized 77 Thus postwar surveys and oral history persons and families, sharing aspirations, inquiries, where they have asked the right manners, values, norms, and dialect. This is questions, have found substantial continuing a crucial difference of historical interpreta- levels of use of withdrawal and even absti- tion from the earlier work of David Levine. nence long after the latex condom, cap, dia- Levine’s Reproducing Families: The Political phragm, and even the pill have become the Economy of English Population History (Cam- more conventional methods: Szreter, Fertility, bridge: Cambridge University Press, 1987), Class and Gender, cited in note 1, pp. 402–407. ch. 3, esp. pp. 115–137, demonstrated, with K. Fisher, “An oral history of birth control prac- the example of the proto-industrial cottage tice c.1925–50: A study of Oxford and South industry framework knitters he had studied Wales,” (unpublished Oxford D. Phil., 1997), in Shepshed, Leicestershire, the important p. 277, summarizing findings on preferences compositional point, which we are laboring for withdrawal over the condom. See also M. here: that the behavior of national average G. Santow, “Coitus interruptus in the twenti- indexes of demographic change may fail to eth century,” Population and Development Review reflect the actual experience of the regions 19 (1993): 767–792. and social groups that it averages. However, 78 J. Knodel and E. van de Walle, “Les- Levine seemed to believe this to be true only sons from the past: Policy implications of his- of the early modern period, as he then pro- torical fertility studies,” Population and Develop- ceeded to offer in Reproducing Families, ch. 4, ment Review 5 (1979): 217–245. an overall interpretation of fertility change

Click to print article Click to return to Table of Contents

Absent and Problematic Men: Demographic Accounts of Male Reproductive Roles

MARGARET E. GREENE ANN E. BIDDLECOM

ALTHOUGH BOTH MEN and women make important contributions to bringing children into life, demographic studies of fertility and family planning have focused overwhelmingly on women. This article describes how the emphasis on women developed and how it has slowly shifted in recent years. Our aim is to show that the limitations of demography’s theoretical approaches to reproduction and its empirical neglect of men have been mutually rein- forcing. The article addresses the ways in which assumptions about gender roles, the history and motivations of demography, and its methodological limitations have influenced the field. The assumption of women’s primacy in fertility and contraceptive use has led to a downplaying and neglect of men’s roles in studies of fertility and family planning. While conception requires two participants, demographic studies of family planning, and especially fertility, historically focused solely on women because of an overarching interest in reproductive outcomes, that is, the actual number of babies born. This article seeks to stimulate debate within the population field about how men should be incorporated into fertility research and, most impor- tantly, what questions should be asked about them. The predominant approach assumes that men might be interesting to study but are not inher- ently important for understanding reproductive behavior. An explanatory model of fertility that emphasizes proximate determinants such as sexual intercourse, fecundity, and the use of contraceptives does not impel one to collect information from both men and women as a matter of course. Mod- eling decisionmaking, however, requires data from both women and men that can answer questions about couple communication, negotiation, and the degree of men’s influence on choices and outcomes related to fertility

POPULATION AND DEVELOPMENT REVIEW 26(1):81–115 (MARCH 2000) 81

Click to print article Click to return to Table of Contents 82 M ALE REPRODUCTIVE ROLES or contraceptive use (Blanc et al. 1996; Rosen and Benson 1982). Further, if fertility is to be understood in specific cultural contexts, then men must be included in demographic research to explain fertility and to make predictions about fertility change (Goldscheider and Kaufman 1996). Our discussion in this article is structured around three specific aims: 1) to describe why men have had a relatively low profile in demographic research on reproduction; 2) to evaluate and characterize the existing research on men, primarily in developing countries; and 3) to suggest directions for future research on men and reproduction.

Barriers to incorporating men in studies of reproductive behavior

Demography as a field arose as an accounting system, a way of enumerating births, deaths, migration, and other population events. The primary purpose of these calculations was to describe and compare populations. As Bogue (1969) pointed out many years ago, demography has few unique concepts and theories that would explain “why” a particular demographic situation exists at a particular moment or what forces underlie an observed change in demographic status. Most of the variables and theories that “explain” demographic events come from other social science disciplines, and “demographic theory” is an organized synthesis of inferences and principles extracted from economics, sociology, social psychology, psychiatry, political science, anthropology, and geography (1969: 5). Current quantitative models in demography draw self-consciously on theoretical insights from anthropology, sociology, and economics. Yet many of the assumptions that guide the descriptive measures and constitute the explanatory models remain unquestioned. The relative lack of knowledge about men’s roles in fertility and family planning is thus a feature of theoretical, methodological, and even ideological aspects of demographic research. We should note that although we refer generally to demographic research, we are particularly interested in research motivated by the international population movement, whose assumptions have shaped research even in areas not concerned with reducing fertility. Throughout the article the term “demographic research” refers to this influential strand of investigation. Demography’s assumptions (like those of other fields) were informed by the social norms in place at the time in the Western countries in which the field developed, and these norms in large part emphasized the exclusive involvement of women with childbearing. Demography’s role in promoting family planning worldwide further shaped the development of the field and its reliance on assumptions of Western family norms: programs were established with funding from developed countries, promoting these

Click to return to Table of Contents M ARGARET E. GREENE / ANN E. BIDDLECOM 83 assumptions on a global scale, and the nature of demographic research was such that dissonances between these assumptions and local social realities were not always evident. The objective of most research was to point to ways in which women’s contraceptive use could be increased. Men were problematic from the outset because they maintained sexual relationships with women that often departed from the Western norm (notably, when their relationships were polygynous, serial, or simultaneous), and because it was understood that their distance from childrearing made them resist women’s family planning use.

The social context in which demography arose

The most important barrier to the inclusion of men in demographic research was normative and reflected the socialization of influential demographers and the research course they set. As Whiting and others point out, the white middle class of the United States “provides the implicit cultural context for demographic theorizing about fertility” (cited in Townsend 1997: 105). Fam- ily structures differing from Western norms may have challenged the way research was formulated, but these structural differences in families were largely ignored. Demography has regarded men as important economically but as typically uninvolved in fertility except to impregnate women and to stand in the way of their contraceptive use. As Watkins argues,

modernization [theory] and the New Household Economics share similar understandings of men and women. . . . Men work outside the home, whereas women are responsible for activities associated with the production of children and domestic services. (1993: 561)

The acceptance of women’s close accountability for children has been reflected in the collection of fertility data from women only. Another aspect of the Western family norm that has contributed to the neglect of men in research is the assumption of consonance between men’s and women’s interests within marriage. If the workings of marriage and childbearing are seen as the resolution of potentially conflicting interests on the part of husband and wife, an interest in men’s as well as women’s fertility behavior follows naturally. Since marriage has enormous psychological and romantic significance, it has been idealized as the joining together of two into one, and notions of conflict and negotiation have been glossed over. Two may likewise become one in contexts where a husband dominates his wife or where wives’ interests are subsumed into those of their husbands through social customs and legal structures. Both notions of unity are problematic, particularly where the links between marriage and

Click to return to Table of Contents 84 M ALE REPRODUCTIVE ROLES fertility are weaker or where husbands and wives have greater economic independence from each other (Lesthaeghe 1989). In recent years, demographic research has begun to address issues of power and negotiation, making use of theories from economics and anthropology. By routinely excluding men from demographic analyses regardless of cultural context, research has treated husbands and wives as entirely analogous individuals in a dyad. The neglect of power relations both inside and outside the relationship has made it difficult to make sense of reproductive decisions in different contexts. Again, however, it is natural that out of the assumption of coincident or even identical interests would come studies of fertility that simply disregard one of the spouses, since one view represents the position of both. A Western model of childbearing and childrearing is inappropriate in many settings, even in the West. In their cross-cultural study of childrearing, Whiting and Whiting (1975) remarked on the wide range of patterns of parental responsibility and in particular the unique emphasis in the United States on an unsupported mother. This focus has made it appear acceptable to associate only women with fertility. This focus and the resultant lack of information on men have contributed to stereotypes about male promiscuity. More important, perhaps, is that a lack of information on men implicitly overemphasizes female responsibility for contraceptive use, pregnancy, and childbearing (Lloyd 1996). As a result, policies in this regard focus almost exclusively on girls’ and women’s sexual behavior.

The intellectual context in which demographers work

The assumptions we have described persisted in the face of cultural evidence that challenged them. Although health workers and anthropologists were working in many developing-country settings, their knowledge of local beliefs, understandings, and customs was either unknown or largely dis- regarded by demographers. Demography was slower to address these shortcomings than some other fields, in part because of its lack of coherent theoretical grounding. As an interdisciplinary field, demography is informed by selected theories from other fields that do not need to be reconciled with one another. Demogra- phy is also exceptional in its focus on just a few theories: modernization theory, the new household economics, and demographic transition theory. Its theoretical weakness has been bemoaned by many researchers, particularly scholars trained in anthropology and economics (e.g., Greenhalgh 1990; Hodgson 1983; Manser and Brown 1979; McElroy and Horney 1981; Renne 1996; Riley 1997; Robinson 1997). The simple attribution of children to the women who gave birth to them results not only from the biological realities of fertility but also from

Click to return to Table of Contents M ARGARET E. GREENE / ANN E. BIDDLECOM 85 interpretations of the social realities surrounding it. Demography tends to have an orientation toward the category of sex, even when talking about behaviors that would be more usefully understood as having to do with gender (Riley 1997). Although the term gender has gained in popularity, often its use reflects a terminological sleight of hand: gender is invoked when what is actually being addressed is the biological category of sex. The purportedly biological premises that associate childbearing solely with women on the basis of sexual difference narrow demographic explanations of fertility patterns and change in spite of the joint nature of conception. Ironically, had biological models been more fully emphasized earlier, men would have been integrated into demographic studies; instead, the biology of reproduction was invoked just sufficiently that the underlying premise of women’s centrality was not questioned. Biologically oriented theories of reproduction are among those that have more recently inspired demographic interest in “reproductive strategies,” questioning why men have children at all (e.g., Bergstrom 1996; Kaplan, Lancaster, and Anderson 1998).

The research tools of the demographic trade

Suppositions made in all fields are reproduced and carried out in new contexts through research methodology. Although some early demographic studies did look at men’s fertility (e.g., Tietze 1938, 1943; Stycos, Back, and Hill 1956), they were the exception. The widely accepted reasons for not calculating male fertility rates have been described in a classic text on demographic techniques (Shryock and Siegel 1976): first, men’s reproductive spans are not as clearly defined as women’s; second, women are easier to interview because they are more usually at home than men; and third, if they are not living with both parents, children are more likely to be living with their mothers than their fathers. To the extent that demography is only about counting, using women as the reference point is indeed sufficient: it is simpler to collect and analyze information about childbearing from one sex. Mothers remember events such as miscarriages and deaths in early childhood more clearly than fathers do, and there is no ambiguity as to whether a child is theirs or not (see Bachrach et al. 1992; Becker 1996; and Watkins 1993 for reviews of these issues). A keystone of demography is accurate measurement; thus it is understandable that the observed link between mother and child, where there is less room for ambiguity, would be preferred in the documentation of fertility. Yet since demographers do not limit themselves to counting but also attempt to explain and to predict fertility behavior, this methodological justification is patently weak. Another reason for excluding men in models of reproductive behavior is methodological complexity. The incorporation of information on husbands

Click to return to Table of Contents 86 M ALE REPRODUCTIVE ROLES and wives into a single quantitative model is difficult because of the high collinearity between variables that are the same for each spouse (e.g., correspondence in their levels of education). Modeling is further complicated by unions that depart from effectively permanent monogamy such as polygyny or extramarital partnerships. Because many surveys in developing countries have been externally funded by developed-country consortiums, broad comparisons at the national level with the goal of promoting family planning have been emphasized over context-specific studies. National governments generally share the preference for national-level data: they want to know about the entire country for planning purposes and are wary of emphasizing one region over another. But there is a price measured in depth and detail to be paid for the broad strokes of national surveys. More culturally specific research would point directly to the varied roles of men and women in different contexts. Data collected from men may have certain flaws that differ from the shortcomings of data on women: men, unlike women, may potentially know little about their own progeny and may undercount them (Lloyd and Gage-Brandon 1992). In an article comparing spouses’ accounts of their reproductive histories in one US community, Fikree, Gray, and Shah (1993) found that men accurately reported the number and timing of live births, but were less reliable in reporting other events such as spontaneous or induced abortions. In questions of paternity and men’s recall of information about children, there is more room for ambiguity, especially among men who father children out of wedlock or whose ties to mothers are tenuous for other reasons such as divorce (Cherlin, Griffith, and McCarthy 1983; Rendall et al. 1999). Data collected from women can, however, also be im- perfect for many reasons, both unintentional and willful (see, for example, Jones and Forrest 1992 on women’s underreporting of pregnancies ending in abortion).

A growing interest in men’s roles

Despite the barriers described above, demographic research on men’s roles in reproduction has grown in the 1990s. Indeed, there is an insistent push from within research and policy-oriented circles to include men in studies on fertility and family planning (Becker 1996; Green, Cohen, and Belhadj- El Ghouayel 1995; Lloyd 1996; Lockwood 1996; Watkins 1993). Several factors account for this rising interest. First, feminist thinking has had direct effects on the way demographers view men. Feminists have written extensively about the social meaning of women’s childbearing roles and their exploitation through marriage (e.g., Ehrenreich 1983; Ginsburg and Rapp 1995; Greer 1984; Rich 1986) and this thinking has profoundly affected Western society. Feminism has,

Click to return to Table of Contents M ARGARET E. GREENE / ANN E. BIDDLECOM 87 however, had more influence on demography indirectly through its impact on health and population policy and its definition of what is important to study. By identifying policy issues such as child support and women’s reproductive health, among others, feminism called attention to areas neglected in demographic research. As Presser (1997) has argued, demography is ideologically undisposed to absorb and make use of relevant feminist ideas: “Women are given special (and usually exclusive) attention in fertility research, but their wellbeing as compared to the wellbeing of men is not the central issue; rather, the issue is the factors that determine their reproductive behavior” (1997: 303). One of feminism’s contributions has been to weaken this disposition, thereby promoting the study of the sexes in contrast with each other, not just the analysis of women’s characteristics as determinants of fertility. Second, the women’s health movement has directed more demographic attention to men. The movement was especially influential at the 1994 In- ternational Conference on Population and Development (ICPD) in shifting the focus from family planning programs to reproductive health (see Hodgson and Watkins 1997 for a discussion), a move that has dismayed some demographers concerned with a more exclusive emphasis on the issue of rapid population growth and its economic consequences (see, for example, De- meny 1994; Westoff 1994). The ICPD’s Programme of Action and comments from leaders reflecting the spirit of this document stress the need to make men more aware of their responsibilities to the family and the wider community in the matter of family planning and reproductive health (United Nations 1995). The Programme of Action specifically states:

Special efforts should be made to emphasize men’s shared responsibility and promote their active involvement in responsible parenthood, sexual and reproductive behaviour, including family planning; prenatal, maternal and child health; prevention of sexually transmitted diseases, including HIV; prevention of unwanted and high-risk pregnancies; shared control and contribution to family income, children’s education, health and nutrition; and recognition and promotion of the equal value of children of both sexes. Male responsibilities in family life must be included in the education of children from the earliest ages. Special emphasis should be placed on the prevention of violence against women and children. (1995: paragraph 4.27)

The failure of classic demographic transition theory to explain fertility change across a range of settings is a third factor in the increasing interest in men’s roles. For decades the centerpiece of demographic research was demographic transition theory, a theory of population change that supported an interventionist perspective in international population work by pointing to a path of mortality and fertility decline that eventually every country should take. This, in turn, helped to justify financial support for women-

Click to return to Table of Contents 88 M ALE REPRODUCTIVE ROLES oriented fertility control research (Hodgson 1988). Criticisms of the theory have focused on its oversimplification of the factors influencing fertility change and on its crude and normative character in its efforts to unify the mortality and fertility experience of countries around the globe and to predict the direction in which they are headed (Hodgson 1983; Szreter 1993). Responses to the criticism have sparked more culturally specific research on reproductive behavior and on the varied reproductive roles of men and women in different cultural contexts. Once transition theory began to be questioned, difficulties in describing particular reproductive patterns were increasingly interpreted as reflecting conceptual shortcomings rather than merely weaknesses of the data. Marital arrangements such as polygyny (Speizer 1995), marital infidelity (Orubuloye, Caldwell, and Caldwell 1992), and marital instability (Rao and Greene 1993) that do not follow the uniform pattern laid out by demographic transition theory and make traditional demographic research on women’s fertility alone less appropriate illustrate these points well. Fourth, recent methodological developments have also catalyzed an interest in men and the social relations between men and women. The development of analytic models in economics, anthropology, and sociology that successfully encompass more than one actor has been the first major change. Anthropology’s analytic separation and clarification of the biological and social components of childbearing and childrearing has shed a great deal of light on fertility (Townsend 1999). Econometric models now often incorporate conflict or bargaining into analyses of fertility (see Manser and Brown 1979; McElroy and Horney 1981). Influential conceptual models that address issues of power and negotiation within couples include those by Beckman (1983), Hollerbach (1980, 1983), and Manser and Brown (1979) (for a review see Folbre 1988); Dwyer and Bruce’s (1988) volume focuses on women’s management of money in developing countries and their use of fertility as a bargaining tool with their husbands. Lastly, a number of researchers are working on innovative strategies for handling the problems posed when husband and wife are included in the same model (Smith and Morgan 1994; Speizer and Yates 1996; Thomson 1989, 1997a). Combining quantitative data and qualitative ethnographic data represents another methodological development. Its increasing use helped to bring into focus the varied social aspects of fertility and broadened our understanding of the social mechanisms of demographic phenomena that are often not amenable to survey measurement (Caldwell and Caldwell 1987; Greenhalgh 1990; Schneider and Schneider 1996). A desire to understand the broader context of fertility decisionmaking has also led to the expansion of the units of analysis, conceptually if not methodologically. For example, there is evidence that not only couples but also extended families participate in decisions affecting fertility and contra-

Click to return to Table of Contents M ARGARET E. GREENE / ANN E. BIDDLECOM 89 ceptive use (e.g., Das Gupta n.d.; Zulu 1997) and that spousal interaction and the larger social group influence reproductive behavior (Phillips et al. 1997). Research on social networks has encouraged measures and models of the spread of ideas about fertility and fertility control (Bongaarts and Watkins 1996; Montgomery and Casterline 1996). Furthermore, the inclusion of male respondents and the study of male roles have become important components of efforts to collect data that measure women’s status and the wellbeing of children (Mason 1996). Nevertheless, the dominant emphasis in fertility research continues to be on women as the basic units of reproduction, and the underlying concern lies with reducing their fertility rather than with their power and autonomy as significant outcomes. Although many demographers voice an interest in women’s status, questions on such matters as the effects of fertility decline differentiated by sex have been neglected. The most common way of measuring women’s status—to the complete exclusion of men’s circumstances—reduces its meaning. The intellectual and methodological changes described here have nevertheless brought men into demographers’ research agenda on fertility, albeit often as accessories to women rather than as the objects of study. Ques- tions about women’s status reflect a shift in the search for understanding reproductive behavior: from seeing women as individuals acting alone to seeing them (and men) as part of families and households, where reproductive strategies are based on the preferences of multiple actors. The recent underlying conceptualization of men’s roles in the domain of fertility can be interpreted as a feminist one, involving considerations of power, autonomy, and decisionmaking. With respect to standard tools for analyzing fertility like the proximate determinants framework, this translates into research questions about differentials concerning not only when women marry, for example, but also whom they marry and the implications of these men’s experiences for later reproductive behavior (Basu 1999). We conclude this section with a comment on the unexpected partnership that has developed between demography and feminism, a theme ex- plored at length by Hodgson and Watkins (1997). The underlying conception of men reflected in demographic research may be characterized as close to a feminist one, although many feminists and demographers might not concur. Feminists and demographers interested in fertility control have identified men as “problems” in similar ways: neo-Malthusian demographers see men as potential obstacles to women’s exercise of their fertility preferences; and feminists see them as potential obstacles to women’s exercise of their rights, one of which is, of course, the exercise of their fertility preferences. The uneasy alignment of demographic and feminist interests promotes the aims of both groups, potentially allowing some demographers to characterize their work as feminist. The alignment may also benefit femi-

Click to return to Table of Contents 90 M ALE REPRODUCTIVE ROLES nists because the financial resources available for population concerns are often much greater than those for women’s issues narrowly defined.

What do we know about men’s reproductive roles?

As a consequence of this growing interest in men’s roles, one can no longer assert that men are missing from the demographic literature on reproduction. The number of articles on men has increased greatly in recent years and much of this growth consists of studies that examine both men and women. A review of the literature on family planning over roughly the past two decades shows consistently about three female references to every male reference, with a very slow annual increase in studies on men alone (Stycos 1996). The number of articles covering both sexes, however, increased by nearly half over this same period. Many large-scale data collection projects also seek to interview both men and women. While the World Fertility Surveys interviewed men in just a few countries, the Demographic and Health Surveys (DHS) were far more active in this endeavor: the DHS has conducted more than 50 surveys of men, particularly in sub-Saharan Africa, as of 1999. Although studies of men’s reproductive attitudes and behaviors have grown in number, they are dominated by a problem-oriented approach. In short, men are a focus of study because they contribute to a perceived social concern. Diverse examples of this approach may be given. The AIDS epidemic spurred interest in men’s sexual behaviors and condom use; rising rates of single motherhood generated interest in men’s economic and time investment in children or its absence; and continued high fertility despite family planning program efforts turned attention to fertility decisionmaking and the dominance of men in those decisions. A problem-oriented approach often dictates the questions about men’s roles in research studies and the structure of such studies even when men are not explicitly treated as problematic. One of the consequences is that the substantive areas covered are relatively few and the picture of male reproductive roles is incomplete. The questions asked about male roles tend to focus predominantly on a few issues such as contraceptive methods, thus further promulgating a partial view of men’s roles. We now review recent studies that focus on men to determine the main questions being asked about men’s reproductive roles and whether the empirical evidence bears out a problem-oriented approach to these roles (regardless of the approach adopted by the studies). The review encom- passes mainstream studies of fertility and family planning in developing countries, though we point to some studies in developed countries for contrast. The reader is referred to a collection of US-based studies on men published by the Alan Guttmacher Institute (1996), a review by Goldscheider

Click to return to Table of Contents M ARGARET E. GREENE / ANN E. BIDDLECOM 91 and Kaufman (1996), and the literature on fatherhood (see Booth and Crouter 1998; Coltrane 1996; Lamb 1987; and Marsiglio 1995). We will also note substantive areas where empirical evidence is still lacking and research attempts have only recently begun. The review is organized by a series of commonplace, problem-oriented questions about men’s reproductive roles, each followed by a brief synopsis of existing studies and the research gaps that remain. The questions we pose roughly encapsulate assumptions, often negative in their implications about men’s varying involvement throughout the reproductive process, beginning with their knowledge about fertility control and ending with their investments in children.

Are men uninformed about fertility control?

Were we to assume that childbearing and pregnancy are primarily women’s concerns, then it would not be surprising to find that men did not know much about contraceptive methods in general (and female-controlled methods in particular). Nor would we expect men to know much about the female reproductive cycle. In fact, men are as knowledgeable as women about contraceptive methods (Ezeh, Seroussi, and Raggers 1996), sometimes better informed than women about male methods (Hulton and Falkingham 1996; Mbizvo and Adamchak 1991), and sometimes less informed than women about female methods (Kalipeni and Zulu 1993). This knowledge is usually defined as men’s awareness of contraceptive methods—phrased in surveys as having “heard of” a particular method—and does not refer to other aspects of contraceptive knowledge or how to use a given method. There is much less evidence about male views of abortion, an important element in fertility control given that an estimated one in four pregnancies worldwide is terminated deliberately (Kulczycki, Potts, and Rosen- field 1996). While one might assume that men, especially unmarried men, do not know about their partners’ unwanted pregnancies or abortions, some evidence indicates otherwise. In a comparison of a number of Latin Ameri- can surveys in which men were asked about abortion, 32 to 60 percent of men aged 15–24 years reported that their partners had had abortions (Morris 1993). We remain, though, with an extremely limited picture of men’s knowledge and perceptions of this procedure, one that can involve substantial financial and health costs, as well as psychological and ethical considerations. Most studies that examine men’s views of contraception do so in limited ways, asking about overall approval of contraception or of a few methods such as condom use or vasectomy (Sarkar 1993). There are very high levels of general approval of contraception among men in most developing countries, and differences between men’s and women’s approval of contraceptive use tend to be small (Ezeh, Seroussi, and Raggers 1996). The relatively high levels of men’s approval of contraception may be influenced by

Click to return to Table of Contents 92 M ALE REPRODUCTIVE ROLES the desire to project a “modern” image to an interviewer or to provide a socially desirable answer; however, adequate tests of such a conjecture (e.g., comparing interviewer-administered versus self-administered question- naires) have not been made in developing-country settings. There are far fewer figures on men’s perceptions of other dimensions of contraceptive use, such as health side effects, efficacy, ease of use, or privacy. One study in the Philippines that focused in more detail on perceived costs of contraception found that men expressed strong views about methods and their various attributes and that their views were similar to women’s overall, although the level of disagreement among matched spouses was substantial (Biddlecom, Casterline, and Perez 1997). A study of lower- income Egyptian men’s opinions of contraception found that they had specific concerns about the health and sexual side effects of contraceptives for their wives, such as fatigue brought on by pill use or breakthrough bleed- ing from using the IUD (Ali 1996: 104). Men’s views of contraception may also vary with the reasons for using contraceptives. For example, a study in Dakar, Senegal found that acceptance of contraception among men was significant, even among men from the most conservative backgrounds, when it was for the purpose of spacing births (Posner and Mbodji 1989). The numerous studies on whether or not men know of contraceptive methods stand in contrast with very few studies on how men acquire this knowledge and whether the knowledge means anything in practical terms. Some efforts have been made to learn about men’s general sources of information on family planning, including media exposure and their social networks (Adamchak and Mbizvo 1991; Agyeman et al. 1996). Given that in most settings little effort has been made to educate men about reproduction and family planning, a better understanding of what men are learning and from whom or where they learn it would be of interest. Judging simply by contraceptive knowledge, it is evident that men are learning about both male and female contraceptive methods, but the degree to which they are informed about reproductive health, especially the experiences of their partners, is relatively unknown.

Do men avoid responsibility for fertility control?

The issue of male responsibility underlies the recent emphasis in population policy and programs on encouraging men to be supportive partners (Green 1990; Green, Cohen, and Belhadj-El Ghouayel 1995; Hawkins 1992; United Nations 1995). While we know that men have at least heard of both male- and female-controlled methods as often as women, that knowledge may not result in using methods, supporting partners who do, or seeking treatment when there are deleterious health side effects. One of the simplest indicators that men assume responsibility for fertility control is contraceptive use, especially of methods that directly require

Click to return to Table of Contents M ARGARET E. GREENE / ANN E. BIDDLECOM 93 their participation. Married men almost always report higher rates of contraceptive use than do married women (Ezeh, Seroussi, and Raggers 1996). This may reflect genuine differences in contraceptive use between men and women (i.e., men may be more likely to have more sexual partners than women), it may reflect different propensities to report accurately, or it may indicate divergent interpretations of the couple’s reproductive behavior. One study of five countries found that spousal differences in reported contraceptive use were much less likely among couples who were monogamously married or in which the wife had a secondary or higher education; and, surprisingly, men’s reports of extramarital sexual relations had no effect on differential use (Ezeh and Mboup 1997). In another study of spousal disagreement and contraceptive use, only one-third of couples concurred in their reported use of contraceptives: the lowest agreement was for those using condoms, abstinence, and rhythm, which men typically reported using while their wives did not (Koenig, Simmons, and Misra 1984). Expla- nations for this discrepancy focused on marriage duration and interview context, spousal fertility preferences, spousal interaction, and approval of family planning. The authors concluded that wives tend to underreport contraceptive use in survey interviews, mainly because of familial and normative values that run counter to women’s use of contraceptives and women’s subordinate status within the family (Koenig, Simmons, and Misra 1984: 281). Other studies provide mixed evidence on the effect that the sex of the interviewer has on how men and women answer survey questions. In Ni- geria, for example, women’s reports of contraceptive use were not significantly lower when interviewed by men rather than women, with the exception of a culturally conservative area in northern Nigeria (Becker, Feyisetan, and Makinwa-Adebusoye 1995). In contrast, a study in Nepal found that both men and women tended to underreport sensitive behaviors when the interviewer was male rather than female (Axinn 1991). Many men directly assume responsibility for preventing pregnancy through coital-dependent methods such as condoms, withdrawal, and periodic abstinence. Studies of men’s acceptance of vasectomy and condoms are numerous, reflecting the public health problem of sexually transmitted diseases (see Bertrand et al. 1989; Grady et al. 1993; Pleck, Sonenstein, and Ku 1990; Ringheim 1993; Ross and Huber 1983). A revealing study in Thai- land shows that while men may think condom use is good in general, their views of actually using condoms in sexual relations with spouses may be negative because of the association between condoms and promiscuity, disease, and commercial sex (Knodel and Pramualratana 1996). Studies in Uganda and Tanzania also found little support among men for condom use within marriage (Blanc et al. 1996; Pool et al. 1996). Despite the widespread use of traditional methods (some of which require men’s involvement) over generations, less attention has been devoted to understanding men’s use of methods such as withdrawal and periodic

Click to return to Table of Contents 94 M ALE REPRODUCTIVE ROLES and postpartum abstinence. In a number of sub-Saharan African countries, more than 50 percent of current contraceptive practice is recourse to “traditional” methods, periodic abstinence being predominant (Ezeh, Seroussi, and Raggers 1996). Withdrawal accounts for 9 percent of overall contraceptive practice in developing countries and is widely used in some countries (e.g., 26 percent of all married women of reproductive age in Turkey report withdrawal as their current method: Rogow and Horowitz 1995). Given that withdrawal is free and involves no effort to obtain, it may play a role in the sexual activity of adolescents or as a stopgap method in the absence of acceptable alternatives. Data from the United States, Spain, and Turkey show that high rates of withdrawal are reported by married and unmarried adolescents, particularly at the beginning of sexual relationships (Rogow and Horowitz 1995). It is sometimes argued that the use of male-controlled methods is necessary for men to assume responsibility for contraception, and that the programmatic and contraceptive development emphasis on female methods has reinforced men’s ability to avoid a connection between sexual behavior and reproductive responsibility. Prior to the 1960s, most of the available contraceptive methods (the condom, vasectomy, and withdrawal) were male-controlled. Since then, contraceptive technology has advanced with respect to female-controlled methods, but male methods remain the same (Csillag 1996; Ringheim 1993, 1996). While the call for more investment in male contraceptive methods is not new (Diller and Hembree 1977), only about 8 percent of the world’s contraceptive research budget is allotted for the development of new male contraceptives (Sachs 1994). A lively debate exists as to whether contraceptive development should shift its attention to male- controlled methods (e.g., Harrison and Rosenfield 1996), and program-oriented research often highlights the inadequacy of contraceptive choices for men and the need for more alternatives (Chikamata 1996; Marsiglio 1985; Mbizvo and Adamchak 1992; Ringheim 1993, 1996). One implicit assumption of involving men in family planning programs and encouraging men to assume greater responsibility for fertility control is that male involvement will make them more aware of and empathetic toward women’s reproductive needs and rights. The “supportive partner” language of the 1994 ICPD Programme of Action makes this assumption. In- deed, one review of reproductive health programs found that interventions, such as increasing condom use among couples in which one partner is HIV- positive, were more effective when male partners were involved than when only women were involved (Becker 1996). On the other hand, one could also assume that male involvement sub- verts women’s authority, and in this sense men should be less involved in fertility control. For example, men in a number of countries have legally been able to veto their wives’ use of family planning services regardless of their wives’ preferences (see Cook and Maine 1987 for a cross-national re-

Click to return to Table of Contents M ARGARET E. GREENE / ANN E. BIDDLECOM 95 view). Another example can be found in women’s use of contraception without their husbands’ knowledge. Estimates of surreptitious contraceptive use in three sub-Saharan African countries range from 7 percent to 20 percent of all female contraceptive use, and secret use tends to be more common when overall contraceptive prevalence is low (Biddlecom and Fapohunda 1998). Qualitative data suggest that for some women, not only is their husbands’ involvement in family planning undesirable but they would face harsh consequences if their contraceptive use were detected by their husbands (Biddlecom and Fapohunda 1998; Blanc et al., 1996; Renne 1993; Rutenberg and Watkins 1997; Watkins, Rutenberg, and Wilkinson 1997). Despite radically varying views of male involvement in fertility control and the active role men play in using contraceptives, research is lacking on how men themselves view the connection between contraceptive use (for preventing pregnancy or preventing the spread of disease) and their involvement in other aspects of reproductive responsibility (see Awusabo- Asare and Anarfi 1997; Schneider and Schneider 1991). This link is critical to understand because as societies shift from a predominant reliance on traditional methods to the use of modern methods, men may play a diminished role in fertility control and may also take less responsibility for the consequences of sexual intercourse. Given the possibility of fathering children with more than one woman, are fathers likely to support children from various unions? In sub-Saharan Africa, the declining practice of postpartum abstinence accounts for a large part of fertility change in the region, reflecting changes in marriage regimes and the nature of marital sexual relations. It is important, then, to understand why men and women are aban- doning or replacing this male-involved method and the consequences this has for men’s roles as supportive sex partners and fathers.

Do men pose obstacles to women’s contraceptive use?

One of the main justifications for including men in demographic studies of reproduction has been the assertion that they inhibit women who want to use contraception. This has inspired numerous studies of the reproductive behavior of couples, with a special emphasis on the extent of spousal disagreement. Rarely, though, is the justification posed the other way around—that women might prevent men from using contraception—and even more rarely that men might deny women who want to have more children. The literature is instead grounded in the assumption that men stand in the way of women’s desires for smaller families. To make the more general assumption that men have some influence on women’s fertility, and to determine the degree and means of that influence would be a sounder basis for exploring this subject. Despite the emphasis on men’s pronatalist behavior, existing evidence does not generally support this characterization. For example, only a small fraction of women who state in survey interviews that they want to delay

Click to return to Table of Contents 96 M ALE REPRODUCTIVE ROLES or limit childbearing claim that their partners’ opposition is the main reason they do not intend to use contraception (Westoff and Bankole 1995). But men’s influence is likely to be more indirect than can be indicated by a single survey question (see Casterline, Perez, and Biddlecom 1997). We examine the assumption that men inhibit women’s contraceptive use in three main areas of spousal relations: differing fertility preferences (pronatalism of one of the partners), communication, and decisionmaking. Pronatalism. A longstanding assumption about men’s fertility preferences is that men want more children than do women. One argument is that because men do not bear the physical or often even the economic costs of repeated childbearing borne by women, they are likely to be more pronatalist. The evidence for this assumption is mixed. The fertility preferences of men as a group are similar to those of women. In a thorough review of men’s and women’s fertility preferences in developing countries, Mason and Taj (1987) found that when gender differences occurred they were typically small—the average difference in ideal family size was less than one-fifth of a child. In a more recent review of 17 Demographic and Health Surveys of men and women, Ezeh, Seroussi, and Raggers (1996) documented as wide a variation in men’s fertility preferences as in women’s. Men’s ideal family size ranged from around 9 children in West Africa, to 5 children in East Africa, to about 3.5 in North Africa and Asia. Documented gender differences in fertility preferences were very small except in West Africa, where men’s ideal family size exceeded women’s by 2 to 4 children. West Africa is also a region with high levels of polygyny, which probably explains some of this gap. Of course, disagreements between men and women on the number of children wanted may not be nearly as critical for subsequent reproductive outcomes as differences in the desired timing of another child. The small differences found between men’s and women’s fertility preferences at the aggregate level can obscure substantial disagreement between men and women at the couple level (Mott and Mott 1985). For example, studies in Malaysia and Taiwan showed that overall congruence between men and women on family size preferences and even sex preferences was high, but that agreement was low among couples (Coombs and Chang 1981; Coombs and Fernandez 1978). In an extensive review of couple studies, Becker (1996) evaluated the correspondence between husbands and wives on a variety of reproductive measures across surveys in developing and developed countries. Spousal agreement on subjective matters ranged from 60 to 70 percent (the desire for more children, for example, registered a median level of concordance among couples of 68 percent). The direction of spousal disagreement is likely to be as important as the level of disagreement. Spousal differences in desired family size do not always imply that husbands are more pronatalist. In a study in India, for

Click to return to Table of Contents M ARGARET E. GREENE / ANN E. BIDDLECOM 97 example, couples for the most part shared preferences for additional children and the ideal number of sons, and when there was disagreement husbands tended to be less pronatalist than wives, mainly because of men’s lesser dependence on sons for old-age support (Jejeebhoy and Kulkarni 1989). This study also illustrates the fact that husbands and wives can have critically different vested interests in childbearing decisions. Preferences for children of a certain sex, usually boys, may also make men more pronatalist than women. There is a tendency for men to prefer sons over daughters (Mason and Taj 1987), but this varies across countries. For example, Pebley, Delgado, and Brineman (1980) found that the predominant preference among both men and women in Guatemala was for equal numbers of sons and daughters. Spousal communication. Spousal disagreement on reproductive matters relates to the ways in which men and women communicate their preferences. Spousal disagreement can be due more to the lack of communication between spouses than to the articulated opposition of one spouse to the other’s desires (see Omondi-Odhiambo 1997). The result is that men may have a less direct influence on reproductive decisions than is usually assumed. In West Africa nearly 75 percent of men reported that they had never discussed family planning with their wives; in East Africa fewer than 40 percent of men said they had never discussed it; and in North Africa the percentage was even lower (Becker 1996; Ezeh, Seroussi, and Raggers 1996). Communication can also be nonverbal, especially where there is no tradition of discussion between spouses about sexual intercourse or contraception (Balmer et al. 1995). Failure to communicate about sex and other reproductive matters can lead to a failure to act on commonly held preferences (van de Walle and Maiga 1991). Depending on how decisions are made, this can also mean that behavior remains unchanged. A study in Uganda examined the ways in which negotiating occurs within sexual unions (Blanc et al. 1996). Detailed questions were asked about communication and how disagreements were resolved, and comparisons were made between partners. The authors found that both communication and open disagreement between spouses were uncommon: roughly one-third of respondents had discussed family size or child spacing with their partner, although most respondents believed they had a clear understanding of their partner’s desires. Moreover, each partner tended to claim responsibility for decisions, and women were more likely than men to perceive disagreement with their partner over reproductive issues. These varied findings on spousal communication indicate that more verbal discussion, an increasingly common programmatic recommendation, is neither a universally accepted nor a necessarily beneficial path of action to adopt in

Click to return to Table of Contents 98 M ALE REPRODUCTIVE ROLES order to promote contraceptive use or involve men in reproductive health matters. Decisionmaking. One of the most common motivations for research on men is the assumption that they hold a dominant role in reproductive decisionmaking. Many studies of decisionmaking concentrate on the extent of spousal agreement and its association with a particular outcome and from this infer men’s relative influence in reproductive decisionmaking (Dodoo 1993). Another approach has been to examine the reciprocal influence of spouses on each other’s attitudes or preferences (Ezeh 1993). Theoretical models of reproductive decisionmaking are numerous, but most applications use data from the United States (e.g., Beckman 1984; Beckman et al. 1983; Hollerbach 1980, 1983; Smith and Morgan 1994; Thomson 1990, 1997a; Thomson, McDonald, and Bumpass 1990). This literature focuses less on whether men dominate decisionmaking and more on how spousal disagreement gets resolved and the specific spousal characteristics and desires that affect the couple’s reproductive behavior. Couple disagreement may lead to a continuation of ongoing behavior (Beach et al. 1982) or it may be resolved in favor of the spouse who holds more power. There is little empirical evidence from developing countries that adequately tests these propositions using longitudinal data. One exception is a study in Nigeria which found that among couples with four or fewer children, a subsequent birth was more likely if the husband was more pronatalist than the wife (versus if the wife was more pronatalist); and if there were five or more children, another birth was more likely if the wife was more pronatalist (Bankole 1995). The author argued that a woman was better able to defend her fertility preferences (and, conversely, a man was less likely to insist on his preferences) once she amply demonstrated her ability to bear children. Other studies of decisionmaking draw on specific survey questions such as who is the main decisionmaker, who initiated the decision to use contraception, or who has final say on a given matter. These kinds of questions have become standard fare in a number of nationally representative surveys of men and illustrate for the most part that men perceive reproductive decisions to be made jointly, though when they deviate from this they more often claim responsibility for themselves. For example, 55 percent of men interviewed in a survey in Egypt said that they and their wives decided together on the use of family planning methods and 37 percent said that they alone had the last word (El-Zanaty et al. 1993). In a study in Sudan, 45 percent of ever-married men said family planning decisions should be made jointly by couples while 34 percent said it was the husband’s right alone (Khalifa 1988). Even in the United States, where fertility is low and where more than 75 percent of men (aged 20–39 years) believe that men and women share equal responsibility for decisions about contraception,

Click to return to Table of Contents M ARGARET E. GREENE / ANN E. BIDDLECOM 99 men are twice as likely to claim that they have a greater responsibility in contraceptive decisions as they are to say that women do (15 percent versus 7 percent) (Grady et al. 1996). Several provisos need to be stated with respect to research on men and reproductive decisionmaking. First, the picture conveyed about men’s sense of their relative dominance in reproductive decisionmaking assuredly varies by the kinds of questions asked. “Final say,” “last word,” “who initiated,” and “main influence” all capture different stages and aspects of decisionmaking and are affected by respondents’ attempts to present a certain image to the interviewer. Second, taking responsibility for making decisions is not the same as taking responsibility for implementing them. A survey of men in Zimbabwe, for example, showed that although 54 percent said the husband should have the major say in the decision to use contraceptives, the majority said that the responsibility for obtaining family planning information and supplies rested with their wives (Mbizvo and Adamchak 1991). Third, having a husband who dominates in decisionmaking or opposes using contraception does not necessarily prevent women from using contraception. Women often use contraceptives without their husbands’ knowledge. In one study in Uganda, about 15 percent of women who were using contraception said they were doing so without their partners’ knowledge (Blanc et al. 1996). This could be a permanent practice to circumvent a husband who staunchly disagrees with using contraception, or it could be a short-term strategy to persuade an ambivalent husband that contraceptive use is not a disruptive practice—so much so that the wife could do it without his noticing a difference. Men, too, feel that there are circumstances that warrant wives taking action without husbands’ knowledge. Again, this is with respect to preventing pregnancy—instances in which women co- vertly try to become pregnant are rare.

Are men sexually promiscuous?

Sexually transmitted diseases, particularly AIDS, have radically transformed demography’s conceptualizations of reproduction by broadening the focus from fertility alone to include reproductive health, a shift from counting women’s reports of births to documenting women’s and men’s reports of sexual behavior (Mbizvo and Bassett 1996; Mundigo 1995). Action-oriented research has followed suit to encompass broader objectives such as encouraging supportive sexual partnerships and responsible sexual behavior (Green, Cohen, and Belhadj-El Ghouayel 1995; Johns Hopkins Center for Commu- nication Programs 1997). A number of studies have examined men’s sexual behavior, especially their experience with sexual networks and commercial sex. These studies deal mostly with countries or regions where HIV/AIDS has had a noticeable

Click to return to Table of Contents 100 M ALE REPRODUCTIVE ROLES impact on the population (Caldwell et al. 1993; Cleland and Way 1994; Dyson 1992; Knodel et al. 1996; Morris et al. 1996; Orubuloye et al. 1995; VanLandingham et al. 1993). In general, men have a stronger cultural prerogative than women in initiating and negotiating sexual relationships (Balmer et al. 1995; Mason 1994), though women often draw on culturally sanctioned means of controlling sexual activity, even within marriage (see Blanc et al., 1996; Tsui, de Silva, and Marinshaw 1991). Multiple partners are much more common among men than among women. In Tanzania almost 50 percent of unmarried men are reported to have had multiple partners (compared to about 20 percent of unmarried women) (Rutenberg, Blanc, and Kapinga 1994). In a survey in Kenya, 32 percent of sexually active men and 11 percent of sexually active women reported casual or commercial sex in the preceding 12 months with someone other than their regular partner (Carael et al. 1992). A study in Nigeria found that 54 percent of married men and 39 percent of married women have had extramarital sexual relations (Isuigo-Abanihe 1994). Both men and women often note that men take alternative sexual partners when their wives are abstinent postpartum or are breastfeeding (Olukoya and Elias 1996; Orubuloye, Caldwell, and Caldwell 1991). In one study in Ghana, more than two-thirds of men acknowledged they were having sex with girlfriends while their wives were breastfeeding (Anarfi 1993). In general, men are more sexually active than women, judging by a host of indicators such as premarital sexual experiences, multiple partners, and use of commercial sex. The fact that these sexual behaviors are linked to unwanted pregnancies and the transmission of sexual diseases, often to monogamous female partners, lends cogency to the problem-oriented por- trayal of men’s roles alluded to earlier (see Brandt 1985 for a related discussion of sexual behavior and venereal disease; Mason 1994). One often- mentioned example is the situation where women, potentially at risk of HIV infection from their partners, are unable to persuade their partners to use condoms (Worth 1989). The fact that in many countries men have greater sexual prerogatives than women, as well as the right to enforce these prerogatives, directs attention to the topic of sexual coercion and violence. This topic has begun to receive attention in demographic research and data collection in developing countries, although many of the studies are small-scale and do not relate the experience of violence to specific reproductive outcomes such as unwanted pregnancy or abortion (see Dixon-Mueller 1993; Heise, Moore, and Toubia 1995; Rao 1997). Some studies have made headway in calculating nationally representative estimates of violence and sexual coercion. While one might suspect underreporting of violence, especially in a survey interview, one out of three ever-married women reported in the 1995 Egypt Demographic and Health Survey that they had been beaten at least once

Click to return to Table of Contents M ARGARET E. GREENE / ANN E. BIDDLECOM 101 since marriage, and nearly all said it was the husband who had administered the beating. Of those who had been beaten and who had had a birth, nearly one-third reported being beaten during pregnancy (El-Zanaty et al. 1996). Men themselves admit to violence against their partners. In an in-depth study of couple negotiation in Uganda, about 40 percent of male partners reported that they had physically harmed their partner (Blanc et al. 1996). Because many of these studies on sexual behavior are of adult men, we know little about the nature of sexual activity in adolescence in developing countries, a stage in life where problematic behaviors may develop in the first place. (In contrast, the research literature on sexual activity among adolescents in the United States is extensive; see among others, Ku, Sonen- stein, and Pleck 1993 and Marsiglio 1993.) One study of 12 Latin American countries and cities showed that more adolescent boys had had premarital sexual experience than girls, ranging from 30 to 78 percent of 15-to- 19-year-old boys (Morris 1994). Of course, such frequencies do not in themselves say much about male sexuality. For example, a more in-depth study in Cameroon found that young men as well as young women had financial motives for their sexual relationships (Calvès, Cornwell, and Enyegue 1996). Usually only young women are described as engaging in sexual relationships for material or financial benefits. In short, adolescent social behavior and awareness of fertility, disease risks, and sex role expectations remain relatively unexplored topics in developing countries (see Mensch, Bruce, and Greene 1998).

Do men underinvest in their children?

A further aspect of men’s reproductive roles is fatherhood and men’s commitment to their children. Fathers’ potential contributions to their children’s lives include their presence in the home, the time they devote to childrearing and domestic responsibilities, and their financial support of the household. Men’s involvement with their children is affected by the nature of the tie between the biological father and mother. With no sexual access to the mother of his child, a father may be less willing to invest in that child (Kaplan, Lancaster, and Anderson 1996). Rising levels of divorce and of childbearing outside marriage have meant that more men are living apart from their biological children than previously. This has inspired studies that apply a “deficit” model to examine how much the absence of a biological father negatively affects the child’s wellbeing due to lack of financial support, social interaction, and so on. Numer- ous studies and ongoing research in the United States adopt this approach (see Booth and Crouter 1998; Garfinkel, McLanahan, and Robins 1994 among the many US studies). There is growing evidence in developing coun-

Click to return to Table of Contents 102 M ALE REPRODUCTIVE ROLES tries that this approach is relevant there as well (Bruce, Lloyd, and Leonard 1995). In sub-Saharan Africa, for example, the proportion of school-age children who live in households without their biological father is substantial: about 40 percent in Ghana, Kenya, and Malawi and about one-third in Tanzania, Cameroon, and Zambia (Lloyd and Blanc 1996; Lloyd and Gage-Brandon 1994). While the emphasis is often on deficiencies in men’s financial contributions to their children, at least as important is the “deteriorating father–child relationship that often occurs when he is absent” (Foumbi and Lovich 1997: 21). Attention has also been given to the potential negative effects of the presence of some fathers: those who are violent, drink the family’s resources away, or simply do not contribute to the household, even though they are present (Brachet-Marquez 1994). Studying fatherhood is complicated by these same trends of divorce and nonmarital childbearing. First, many men have biological children from more than one sexual union. For example, in a 1988 survey of adult men in the Congo (formerly Zaire), where polygamy is illegal (though about 8 percent of married men reported having more than one wife), 36 percent of men currently in union had fathered at least one child with a woman other than their current wife or partner, and these men had an average of 2.8 children with other partners (Magnani et al. 1995). In a study of paternity over the life course in Canada, more than 20 percent of men experienced fatherhood outside of marriage (Juby and Le Bourdais 1998). Sec- ond, since the emphasis is overwhelmingly on biological fathers, men’s investments in children other than their own tend to be overlooked. Men actively support children through fosterage, informally caring for family members’ children, or becoming stepfathers by marrying women with children from other unions (Bernhardt and Goldscheider 1997; Juby and Le Bourdais 1998; Townsend 1999). The deficit model of fathers’ financial contributions as applied in developing countries is usually tied to a fertility reduction argument; that is, if the financial burden of children were more equally shared between men and women, men would have a financial stake in controlling their fertility in ways that would lead to a fertility decline (Bruce 1994; Bruce, Lloyd, and Leonard 1995). There is some support for this argument. A study in Nigeria found a negative relationship between the proportion of child support expenses a husband paid and the number of children ever born (Fapohunda and Todaro 1988). The focus on men’s proportionate contribution negates the general assumption that the husband is the primary economic provider and gives some idea of the relative cost of children to each spouse. A Ghanaian study found that more-egalitarian marriages with flex- ible division of domestic tasks and a reduction of shared responsibilities with kin meant costlier parenting for men and hence a greater desire among men for fewer children (Oppong 1987). The author argued that policies should

Click to return to Table of Contents M ARGARET E. GREENE / ANN E. BIDDLECOM 103 take into account the allocation of child costs and responsibilities and who manages to avoid shouldering them (i.e., a “free-rider effect” among men) (1987: 178). An argument linking the costs of children to paternal commitment can be made in the other direction as well. As children become more expensive, men may also be more likely to default on their responsibilities toward those children. Levels of child support delinquency among fathers in the United States lend credence to this argument. In developed countries, institutional factors such as equity in parental leave at the time of birth or the enforcement of child support payments are drawing attention to fathers’ involvement with their children; welfare reform in the United States, for example, has increased concern with “fragile families” (Mincy and Pouncy 1997). In developing countries, however, newer research on fathers has been motivated more by concern with the gender equity objectives articulated at Cairo. When institutional mechanisms ensure child support, men’s and women’s attitudes toward paternal commitment may present barriers to the enforcement of such laws if the relationship between a mother and father is strained. For example, a review of the effectiveness of child support laws in six countries in southern Africa showed that the large majority of men and women thought that the father alone or both biological parents had a duty to support their children (Armstrong 1992). However, men’s and women’s views of paternal commitment diverged once children from extramarital sexual unions and children from previous marriages were considered. Fewer men than women felt that men should support out-of-wedlock children, but both women and men expressed doubts about mandating a father’s support of his children when the couple were no longer married (Armstrong 1992). Of course, the presence of a child can also be used to enforce or legitimate the claims of one biological parent on the resources of the other parent. In other words, having children may have less to do with investing in those children than in ensuring access to resources of one parent by the other (Lockwood 1996). While the rising costs of children appear to increase the probability of paternal default in supporting and rearing them, other global changes are also implicated in the retreat from responsible fatherhood. In southern Af- rica, data show that while men may at first leave their homes temporarily to find work elsewhere, the passage of time diminishes their commitment, and sometimes their ability, to send money home (Bruce, Lloyd, and Leonard 1995). As evidence of men’s declining commitment to responsible fatherhood, researchers point to the increasing global incidence of children born out of wedlock. Engle and Breaux (1998), for example, note that global economic and social change has led to “a retreat from family obligations...; men have less to gain from and less to give to their families” (1998: 12). Elsewhere, we learn that “in some West African societies that encompass

Click to return to Table of Contents 104 M ALE REPRODUCTIVE ROLES polygamy, men are reportedly forming more ‘outside’ relationships in order to maintain multiple sexual partners without undertaking the financial commitments that marrying would involve” (Anderson 1997: 25–26). As we noted earlier, most studies, and therefore most interventions, have been based on a “deficit” model of fathering, in which men are seen more as obstacles to than as resources for improving the circumstances of women and children (Barker 1996). Yet a number of programs have sought to involve men more closely with their children’s wellbeing. For example, by educating men and drawing on their concern for the health of their un- born children, a campaign to combat maternal syphilis in Zambia involved 60 to 80 percent of the partners of infected women (De Graft-Johnston 1995). Another program directed at adolescent Mexican fathers found that three-quarters of fathers who had been present at their children’s birth were still involved in their children’s lives four years later, a significantly higher proportion than found among fathers who had not been present (Engle and Alatorre Rico 1994). Outcomes as varied as involvement in child care, children’s school attendance, and commitment of financial resources for children are more easily attained when mothers and fathers are mutually supportive.

Future directions for research on men

The problem-centered approach to demographic accounts of the reproductive roles of men is limited in three ways. First, this approach usually ig- nores how men view their own reproductive roles and how they perceive these roles as they relate to women’s concerns. The limited research literature on men’s support of their children highlights this point. Second, emphasis is placed on how men differ from women (negatively so), although the weight of empirical evidence suggests that on a host of relevant indicators men and women have much in common. In some cases, the differences among men, whether by age or other characteristics, in reproductive attitudes may be greater than differences between men and women (Basu 1999; Renne 1993). Third, this approach fails to examine male behavior within the broader constraints and obligations that characterize men’s lives, such as changes in the wage economy and men’s abilities to provide economically for their families. As we see it, two demographic trends are likely to become increasingly salient for future research on men’s roles in fertility and family planning. The first is the loosening link between marriage and childbearing, which requires knowing more about men to explain couples’ childbearing and reproductive behavior. The second is the cumulative divergence in the reproductive experiences of men and women over their lifetimes, which requires studying men and women as individuals, not just as members of current sexual unions. In situations characterized by a separation of marriage and fertility and by an increasing asymmetry in the lifetime marriage and

Click to return to Table of Contents M ARGARET E. GREENE / ANN E. BIDDLECOM 105 fertility experiences of men and women, we need to consider the unit of analysis (i.e., couples or individuals) that best suits our research questions. Much of the literature we reviewed assumes that the couple consti- tutes the only partnership for each member and is a lasting union within which people’s statements of fertility preferences and contraceptive use are definitive. Many unions fail to meet one or both of these criteria, as illustrated by polygynous unions in Africa and fluid consensual unions in the Caribbean and much of Latin America (Pebley and Goldman 1986; Rao and Greene 1993; Speizer and Yates 1996). In such situations, increasing trends in divorce and migration are forcing the analytic separation of current sexual partnerships from childbearing. Not only are unions often unstable, but reproductive and sexual partnerships may not be lifelong. For example, in families that include stepchil- dren as well as biological offspring, calculations of parents’ and stepparents’ fertility preferences reflect multiple unions and nonresident children (see Thomson 1997b). In many developing countries, economic changes have led to increases in migration, separating spouses and encouraging alternative or parallel sexual partnerships (Anarfi 1993). As a result of these social trends, the fertility histories of men will increasingly diverge from those of their women partners. While we have described the narrow approaches to demographic research on men and the aspects of male reproductive roles that remain in- distinct as a result, recommending the expansion of research on men per se is not our objective. Rather, we note that to understand sexual behavior, childbearing decisions, and other areas of reproductive health, a more comprehensive approach is necessary. What might emerge from a broader approach and where might such an approach begin? Earlier we discussed anthropology’s useful contribution to the analysis of the social and biological components of childbearing and childrearing and the meaning of parenthood, including fatherhood. The next step would be the integration of work on masculinity, femininity, and sexuality with conventional demographic research. Such integration might lead to a closer exploration of sexuality and fatherhood in shaping reproductive behavior of interest to demographers. Closer attention to the cultural construction of fatherhood and the ways that men and women think about the meaning of parenthood should provide useful direction in the expansion of research on both men and women. A broader approach would facilitate closer linkages with the social change agenda of the ICPD Programme of Action, and would allow for a more precise understanding not only of men’s roles and motivations but of women’s as well. By expanding social and behavioral approaches to the study of reproduction, we should be able to explain more and create more opportunities for programmatic interventions. Ironically, it is the policy sphere, where fertility-control strategies originally constrained analytic ap-

Click to return to Table of Contents 106 M ALE REPRODUCTIVE ROLES proaches to men’s reproductive roles, that has now moved beyond demography’s research agenda. The broader social change agenda of the Pro- gramme of Action advocates the transformation of inequitable gender relations and acknowledges the links between cultural beliefs and practices and demographic behaviors. In this sense, the inclination of the field to treat marriage as a partnership, with consensus between men and women on childbearing and other health-related decisions, is no longer just a working assumption in research, but also an explicit goal of the population field. A critical review by Watkins (1993) argues that research on fertility has focused on the factors that induce women to delay marriage and childbearing and to have fewer children. Demographers are beginning to ask some of the same questions about men and their motivations for having— or not having—children (Axinn 1992). As a first step, we need to understand men’s commitment to parenthood, how this varies across the life course, and how this compares to women’s experience; and we need to incorporate this information into analyses of childbearing and contraceptive use. Studies based on individual men and women are a promising way to generate the kind of information needed. Fertility declines around the world have called attention to changes in childbearing preferences and parental roles and have prompted demographers to look at the costs and benefits of fertility for men as well as for women. A better grasp of what children mean to men and women separately, and how this meaning has changed over time, will improve our understanding of men’s and women’s reproductive strategies. The inclusion of men both as individuals and as members of couples will enable demographers to transcend some of the current assumptions about marriage and fertility and to interpret more effectively the changes taking place throughout the world in reproductive behavior and family formation.

Note

The authors contributed equally to this ar- from Caroline Bledsoe, Judith Bruce, John ticle. The study was supported in part by Casterline, Ann Leonard, Cynthia B. Lloyd, grants from the United Nations Population Mark VanLandingham, and Susan Cotts Fund and the Rockefeller Foundation. The Watkins. authors gratefully acknowledge comments

References

Adamchak, Donald J. and Michael Mbizvo. 1991. “Family planning information sources and media exposure among Zimbabwean men,” Studies in Family Planning 22(5): 326–331. Agyeman, Dominic K., Peter Aglobitse, John B. Casterline, and Mark R. Montgomery. 1996. Social Structure and the Diffusion of Fertility Behavior. Final Report Submitted to the Rocke- feller Foundation. New York.

Click to return to Table of Contents M ARGARET E. GREENE / ANN E. BIDDLECOM 107

Alan Guttmacher Institute. 1996. Readings on Men: From Family Planning Perspectives, 1987– 1995. New York: Alan Guttmacher Institute. Ali, Kamran Asdar. 1996. “Notes on rethinking masculinities: An Egyptian case,” in Sondra Zeidenstein and Kirsten Moore (eds.), Learning About Sexuality: A Practical Beginning, pp. 98–109. New York: Population Council. Anarfi, J. K. 1993. “Sexuality, migration and AIDS in Ghana—A socio-behavioral study,” Health Transition Review 3 (Supplement): 45–67. Anderson, David. 1997. Men, Reproduction and Fatherhood. Liège, Belgium: International Union for the Scientific Study of Population. Armstrong, Alice. 1992. “Maintenance payments for child support in southern Africa: Us- ing law to promote family planning,” Studies in Family Planning 23(4): 217–228. Awusabo-Asare, Kofi and John K. Anarfi. 1997. “Postpartum sexual abstinence in the era of AIDS in Ghana: Prospects for change,” Health Transition Review 7: 257–270. Axinn, William G. 1991. “The influence of interviewer sex on responses to sensitive questions in Nepal,” Social Science Research 20: 303–318. ———. 1992. “Family organization and fertility limitation in Nepal,” Demography 29: 503– 521. Bachrach, Christine A., V. Jeffrey Evans, Sylvia Ann Ellison, and Kathy S. Stolley. 1992. “What price do we pay for single sex fertility surveys?” paper presented at the Annual Meeting of the Population Association of America, 30 April–2 May, Denver, CO. Balmer, D. H. et al. 1995. “The negotiating strategies determining coitus in stable heterosexual relationships,” Health Transition Review 5: 85–95. Bankole, A. 1995. “Desired fertility and fertility behavior among the Yoruba of Nigeria: A study of couple preferences and subsequent fertility,” Population Studies 49: 317–328. Barker, Gary. 1996. The Misunderstood Gender: Male Involvement in the Family and in Reproduc- tive and Sexual Health in Latin America and the Caribbean. Report for the John D. and Catherine T. MacArthur Foundation Population Program. Chicago. Basu, Alaka Malwade. 1999. “Women’s education, marriage and fertility in South Asia: Do men really not matter?” in C. H. Bledsoe et al. (eds.), Critical Perspectives on Schooling and Fertility in the Developing World, pp. 267–286. Washington, DC: National Academy Press. Beach, L. R., A. Hope, B. D. Townes, and F. L. Campbell. 1982. “The expectation-threshold model of reproductive decision making,” Population and Environment 5: 95–108. Becker, Stan. 1996. “Couples and reproductive health: A review of couple studies,” Studies in Family Planning 27(6): 291–306. Becker, Stan, Kale Feyisetan, and Paulina Makinwa-Adebusoye. 1995. “The effect of the sex of interviewers on the quality of data in a Nigerian family planning questionnaire,” Studies in Family Planning 26(4): 233–240. Beckman, Linda J. 1983. “Communication, power, and the influence of social networks in couple decisions on fertility,” in Rodolfo A. Bulatao and Ronald D. Lee (eds.), Determi- nants of Fertility in Developing Countries, Volume 2, pp. 415–443. New York: Academic Press. ———. 1984. “Husbands’ and wives’ relative influence on fertility decisions and outcomes,” Population and Environment 7: 182–197. Beckman, Linda J., Rhonda Aizenberg, Alan B. Forsythe, and Tom Day. 1983. “A theoretical analysis of antecedents of young couples’ fertility decisions and outcomes,” De- mography 20(4): 519–533. Bergstrom, Theodore C. 1996. “Economics in a family way,” Journal of Economic Literature 34: 1903–1934. Bernhardt, Eva M. and Frances Goldscheider. 1997. “Family change and the value of children: Child qualities preferred by biological and informal fathers in the U.S. and Swe- den,” paper presented at the Annual Meeting of the Population Association of America, 27–29 March, Washington, DC.

Click to return to Table of Contents 108 M ALE REPRODUCTIVE ROLES

Bertrand, Jane T. et al. 1989. “Attitudes toward voluntary surgical contraception in four districts of Kenya,” Studies in Family Planning 20(5): 281–288. Biddlecom, Ann E., John B. Casterline, and Aurora E. Perez. 1997. “Spouses’ views of contraception in the Philippines,” International Family Planning Perspectives 23(3): 108–115. Biddlecom, Ann E. and Bolaji M. Fapohunda. 1998. “Covert contraceptive use: Prevalence, motivations, and consequences,” Studies in Family Planning 29(4): 360–372. Blanc, Ann K. et al. 1996. Negotiating Reproductive Outcomes in Uganda. Calverton, MD: Macro International Inc. and Institute of Statistics and Applied Economics [Uganda]. Bogue, Donald J. 1969. Principles of Demography. New York: John Wiley and Sons. Bongaarts, John and Susan Cotts Watkins. 1996. “Social interactions and contemporary fertility transitions,” Population and Development Review 22(4): 639–682. Booth, Alan and Ann C. Crouter (eds.). 1998. Men in Families: When Do They Get Involved? What Difference Does It Make? Mahwah, NJ: Lawrence Erlbaum Associates. Brachet-Marquez, Viviane. 1994. “Absentee fathers: A case-based study of family law and child welfare in Mexico,” Working Paper of the Population Council and the Interna- tional Center for Research on Women. Brandt, Allan M. 1985. No Magic Bullet: A Social History of Venereal Disease in the United States Since 1880. New York: Oxford University Press. Bruce, Judith. 1994. “Reproductive choice: The responsibilities of men and women,” Repro- ductive Health Matters 4: 68–70. Bruce, Judith, Cynthia B. Lloyd, and Ann Leonard. 1995. Families in Focus: New Perspectives on Mothers, Fathers, and Children. New York: Population Council. Caldwell, John C. and Pat Caldwell. 1987. “The cultural context of high fertility in sub- Saharan Africa,” Population and Development Review 13(3): 409–437. Caldwell, John C. et al. (eds.). 1993. Sexual Networking and HIV/AIDS in West Africa. Supple- ment to Health Transition Review Volume 3. Calvès, Anne-Emmanuèle, Gretchen T. Cornwell, and Parfait Eloundou Enyegue. 1996. “Adolescent sexual activity in Sub-Saharan Africa: Do men have the same strategies and motivations as women?” Population Research Institute Working Papers in Afri- can Demography No. AD96-04. State College, PA: Pennsylvania State University. Carael, M., J. Cleland, J.-C. Deheneffe, and L. Adeokun. 1992. “Research on sexual behaviour that transmits HIV: The GPA/WHO collaborative surveys—Preliminary findings,” in Tim Dyson (ed.), Sexual Behaviour and Networking: Anthropological and Socio-Cultural Stud- ies on the Transmission of HIV, pp. 65–87. Liège, Belgium: Editions Derouaux-Ordina. Casterline, John B., Aurora E. Perez, and Ann E. Biddlecom. 1997. “Factors underlying unmet need for family planning in the Philippines,” Studies in Family Planning 28(3): 173–191. Cherlin, A., J. Griffith, and J. McCarthy. 1983. “A note on maritally-disrupted men’s reports of child support in the June 1980 Current Population Survey,” Demography 20(3): 385–389. Chikamata, Davy M. 1996. “Male needs and responsibilities in family planning and reproductive health,” Planned Parenthood Challenges 2: 8–10. Cleland, John and Peter Way (eds.). 1994. AIDS Impact and Prevention in the Developing World: Demographic and Social Science Perspectives. Supplement to Health Transition Review Vol- ume 4. Coltrane, Scott. 1996. Family Man: Fatherhood, Housework, and Gender Equity. New York: Ox- ford University Press. Cook, Rebecca J. and Deborah Maine. 1987. “Spousal veto over family planning services,” American Journal of Public Health 77(3): 339–344. Coombs, Lolagene C. and Ming-Cheng Chang. 1981. “Do husbands and wives agree? Fer- tility attitudes and later behavior,” Population and Environment 42(2): 109–127. Coombs, Lolagene and Dorothy Fernandez. 1978. “Husband–wife agreement about reproductive goals,” Demography 15: 57–74. Csillag, Claudio. 1996. “Male contraceptive pill to start trial in Brazil,” The Lancet 348: 608.

Click to return to Table of Contents M ARGARET E. GREENE / ANN E. BIDDLECOM 109

Das Gupta, Monica (no date). “Kinship systems and demographic outcomes,” unpublished paper. De Graft-Johnston, J. 1995. Maternal Syphilis Prevention Programme in Lusaka-Urban, Zam- bia—Mid-Project Assessment Report. UNICEF/FHI. May–June. Demeny, Paul. 1994. “Two proposals for the agenda at Cairo,” International Family Planning Perspectives 20(1): 28–30. Diller, L. and W. Hembree. 1977. “Male contraception and family planning: A social and historical review,” Fertility and Sterility 28(12): 1271–1279. Dixon-Mueller, Ruth. 1993. “The sexuality connection in reproductive health,” Studies in Family Planning 24(5): 269–282. Dodoo, F. Nii-Amoo. 1993. “A couple analysis of micro level supply/demand factors in fertility regulation,” Population Research and Policy Review 12: 93–101. Dwyer, Daisy H. and Judith Bruce (eds). 1988. A Home Divided: Women and Income in the Third World. Stanford, CA: Stanford University Press. Dyson, Tim (ed.). 1992. Sexual Behaviour and Networking: Anthropological and Socio-Cultural Studies on the Transmission of HIV. Liège, Belgium: Editions Derouaux-Ordina. Ehrenreich, Barbara. 1983. Hearts of Men: American Dreams and Flight from Commitment. Gar- den City, NY: Anchor Press. El-Zanaty, Fatma, Hussein A. A. Sayed, Hassan H. M. Zaky, and Ann A. Way. 1993. Egypt Demographic and Health Survey 1992. Calverton, MD: National Population Council [Egypt] and Macro International Inc. El-Zanaty, Fatma et al. 1996. Egypt Demographic and Health Survey 1995. Calverton, MD: Na- tional Population Council [Egypt] and Macro International Inc. Engle, Patrice L. and Javier Alatorre Rico. 1994. “Workshop on Responsible Fatherhood.” Workshop report, unpublished manuscript. Engle, Patrice L. and Cynthia Breaux. 1998. “Fathers’ involvement with children: Perspec- tives from developing countries,” Social Policy Report 12(1): 1–21. Ezeh, Alex Chika. 1993. “The influence of spouses over each other’s contraceptive attitudes in Ghana,” Studies in Family Planning 24 (3): 163–174. Ezeh, Alex Chika and Gora Mboup. 1997. “Estimates and explanations of gender differentials in contraceptive prevalence rates,” Studies in Family Planning 28(2): 104–121. Ezeh, Alex Chika, Michka Seroussi, and Hendrik Raggers. 1996. Men’s Fertility, Contraceptive Use, and Reproductive Preferences. Calverton, MD: Macro International Inc. Fapohunda, Eleanor R. and Michael P. Todaro. 1988. “Family structure, implicit contracts, and the demand for children in southern Nigeria,” Population and Development Review 14(4): 571–594. Fikree, F. F., R. H. Gray, and F. Shah. 1993. “Can men be trusted? A comparison of pregnancy histories reported by husbands and wives,” American Journal of Epidemiology 138(4): 237–242. Folbre, Nancy. 1988. “The black four of hearts: Toward a new paradigm of household economics,” in Daisy H. Dwyer and Judith Bruce (eds.), A Home Divided: Women and In- come in the Third World, pp. 248–264. Stanford, CA: Stanford University Press. Foumbi, Joseph and Ronnie Lovich. 1997. “Role of Men in the Lives of Children: A Study of How Improving Knowledge About Men in Families Helps Strengthen Programming for Children and Women.” New York: UNICEF. Garfinkel, Irwin, Sara S. McLanahan, and Philip K. Robins (eds.). 1994. Child Support Re- form and Child Well-Being. Washington, DC: Urban Institute Press. Ginsburg, Faye D. and Rayna Rapp (eds.). 1995. Conceiving the New World Order: The Global Politics of Reproduction. Berkeley: University of California Press. Goldscheider, Frances K. and Gayle Kaufman. 1996. “Fertility and commitment: Bringing men back in,” in John B. Casterline, Ronald D. Lee, and Karen A. Foote (eds.), Fertil- ity in the United States: New Patterns, New Theories. Supplement to Volume 22, Population and Development Review, pp. 87–99. New York: Population Council.

Click to return to Table of Contents 110 M ALE REPRODUCTIVE ROLES

Grady, William R., Daniel H. Klepinger, John O. G. Billy, and Koray Tanfer. 1993. “Con- dom characteristics: The perceptions and preferences of men in the United States,” Family Planning Perspectives 25(2): 67–73. Grady, William R., Koray Tanfer, John O. G. Billy, and Jennifer Lincoln-Hanson. 1996. “Men’s perceptions of their roles and responsibilities regarding sex, contraception and childrearing,” Family Planning Perspectives 28(5): 221–226. Green, Cynthia P. 1990. Male Involvement Programs in Family Planning: Lessons Learned and Implications for AIDS Prevention. Global Programme on AIDS. Geneva: World Health Organization. Green, Cynthia P., Sylvie I. Cohen, and Hedia Belhadj-El Ghouayel. 1995. Male Involvement in Reproductive Health, Including Family Planning and Sexual Health. Technical Report 28, United Nations Population Fund. New York: United Nations Population Fund. Greenhalgh, Susan. 1990. “Toward a political economy of fertility: Anthropological contributions,” Population and Development Review 16(1): 85–106. Greer, Germaine. 1984. Sex and Destiny: The Politics of Human Fertility. New York: Harper and Row. Gutmann, Matthew C. 1994. “The meanings of macho: Changing male identities in Mexico City.” Unpublished doctoral thesis. Department of Anthropology, University of Cali- fornia at Berkeley. Harrison, Polly F. and Allan Rosenfield. 1996. Contraceptive Research and Development: Looking to the Future. Washington, DC: National Academy Press. Hawkins, K. (ed.). 1992. Male Participation in Family Planning: A Review of Programme Ap- proaches in the Africa Region. London: International Planned Parenthood Federation. Heise, Lori, Kirsten Moore, and Nahid Toubia. 1995. Sexual Coercion and Reproductive Health: A Focus on Research. New York: Population Council. Hertrich, Veronique. 1998. “Are men’s and women’s answers to be equally trusted? A dual data collection on maternity and fertility issues in a population in Mali,” Population: An English Selection 10(2): 303–318. Hodgson, Dennis. 1983. “Demography as social science and policy science,” Population and Development Review 9(1): 1–34. ———. 1988. “Orthodoxy and revisionism in American demography,” Population and Devel- opment Review 14(4): 541–569. Hodgson, Dennis and Susan Cotts Watkins. 1997. “Feminists and neo-Malthusians: Past and present alliances,” Population and Development Review 23(3): 469–523. Hollerbach, Paula. 1980. “Power in families, communication, and fertility decision-making,” Population and Environment 3: 146–173. ———. 1983. “Fertility decision-making processes: A critical essay,” in Rodolfo A. Bulatao and Ronald D. Lee (eds.), Determinants of Fertility in Developing Countries, Volume 2, pp. 797–828. New York: Academic Press. Hulton, Louise and Jane Falkingham. 1996. “Male contraceptive knowledge and practice: What do we know?” Reproductive Health Matters 7: 90–100. Isiugo-Abanihe, Uche C. 1994. “Extramarital relations and perceptions of HIV/AIDS in Ni- geria,” Health Transition Review 4: 111–125. Jejeebhoy, Shireen J. and Sumati Kulkarni. 1989. “Reproductive motivation: A comparison of wives and husbands in Maharashtra, India,” Studies in Family Planning 20(5): 264–272. Johns Hopkins Center for Communication Programs. 1997. “Reaching men worldwide: Les- sons learned from family planning and communication projects, 1986–1996.” Work- ing Paper No. 3. Baltimore, MD: Johns Hopkins Center for Communication Programs/ Population Information Program. Jones, Elise F. and Jacqueline Darroch Forrest. 1992. “Underreporting of abortion in surveys of U.S. women: 1976 to 1988,” Demography 29(1): 113–126. Juby, Heather and Céline Le Bourdais. 1998. “The changing context of fatherhood in Canada: A life course analysis,” Population Studies 52(2): 163–175.

Click to return to Table of Contents M ARGARET E. GREENE / ANN E. BIDDLECOM 111

Kalipeni, E. and E. M. Zulu. 1993. “Gender differences in knowledge and attitudes toward modern and traditional methods of child spacing in Malawi,” Population Research and Policy Review 12(2): 103–121. Kaplan, Hillard S., Jane B. Lancaster, and Kermyt G. Anderson. 1998. “Human parental investment and fertility: The life histories of men in Albuquerque,” in Alan Booth and Ann C. Crouter (eds.), Men in Families: When Do They Get Involved? What Difference Does It Make? Mahwah, NJ: Lawrence Erlbaum Associates. Khalifa, Mona A. 1988. “Attitudes of urban Sudanese men toward family planning,” Studies in Family Planning 19(4): 236–243. Knodel, John and Anthony Pramualratana. 1996. “Prospects for increased condom use within marriage in Thailand,” International Family Planning Perspectives 22(3): 97–102. Knodel, John, Mark VanLandingham, Chanpen Saengtienchai, and Anthony Pramualratana. 1996. “Thai views of sexuality and sexual behaviour,” Health Transition Review 6: 179– 202. Koenig, M. A., G. B. Simmons, and B. D. Misra. 1984. “Husband–wife inconsistencies in contraceptive use responses,” Population Studies 38: 281–298. Ku, Leighton, Freya L. Sonenstein, and Joseph H. Pleck. 1993. “Neighborhood, family, and work: Influences on the premarital behaviors of adolescent males,” Social Forces 72(2): 479–503. Kulczycki, Andrzej, Malcolm Potts, and Allan Rosenfield. 1996. “Abortion and fertility regulation,” The Lancet 347: 1663–1668. Lamb, Michael E. 1987. The Father’s Role: Cross-Cultural Perspectives. Hillsdale, NJ: Lawrence Erlbaum. Lesthaeghe, Ron J. 1989. “Social organization, economic crises and the future of fertility control in Africa,” in Ron J. Lesthaeghe (ed.), Reproduction and Social Organization in Sub-Saharan Africa, pp. 475–505. Berkeley: University of California Press. Lloyd, Cynthia B. 1996. “Family and gender issues for population policy,” in Population and Women: Proceedings of the United Nations Expert Group Meeting on Population and Women, pp. 65–80. New York: United Nations. Lloyd, Cynthia B. and Ann K. Blanc. 1996. “Children’s schooling in sub-Saharan Africa: The role of fathers, mothers, and others,” Population and Development Review 22(2): 265–298. Lloyd, Cynthia B. and Anastasia J. Gage-Brandon. 1992. “Does sibsize matter?: The implications of family size for children’s education in Ghana,” Research Division Working Paper No. 45. New York: Population Council. ———. 1994. “High fertility and children’s schooling in Ghana: Sex differences in parental contributions and educational outcomes,” Population Studies 48(2): 293–306. Lockwood, Matthew. 1996. “Institutional and cultural determinants of demand for reproductive health services in sub-Saharan Africa: A review and implications for research,” unpublished paper. Magnani, R. J., J. T. Bertrand, B. Makani, and S. W. McDonald. 1995. “Men, marriage and fatherhood in Kinshasa, Zaire,” International Family Planning Perspectives 21(1): 19–25. Manser, M. and M. Brown. 1979. “Bargaining analyses of household decisions,” in C. B. Lloyd, E. S. Andrews, and C. L. Gilroy (eds.), Women in the Labor Market, pp. 3–26. New York: Columbia University Press. Marsiglio, William. 1985. “Husbands’ sex-role preferences and contraceptive intentions: The case of the male pill,” Sex Roles 12(5-6): 655–663. ———. 1993. “Adolescent males’ orientation toward paternity and conception,” Family Plan- ning Perspectives 25(1): 22–31. ——— (ed.). 1995. Fatherhood: Contemporary Theory, Research and Social Policy. Thousand Oaks, CA: Sage Publications. Mason, Karen Oppenheim. 1994. “HIV transmission and the balance of power between women and men: A global view,” Health Transition Review (Supplement) 4: 217–240.

Click to return to Table of Contents 112 M ALE REPRODUCTIVE ROLES

———. 1996. “A comparative study of the status of women in five Asian countries,” paper presented at the International Symposium: Life and Earth in the 21st Century, March, Tokyo, Japan. Mason, Karen Oppenheim and Anju Malhotra Taj. 1987. “Differences between women’s and men’s reproductive goals in developing countries,” Population and Development Re- view 13(4): 611–638. Mbizvo, Michael T. and Donald J. Adamchak. 1991. “Family planning knowledge, attitudes, and practices of men in Zimbabwe,” Studies in Family Planning 22(1): 31–38. ———. 1992. “Male fertility regulation: A study on acceptance among men in Zimbabwe,” Central African Journal of Medicine 38(2): 52–57. Mbizvo, M. T. and M. T. Bassett. 1996. “Reproductive health and AIDS prevention in Sub-Saharan Africa: The case for increased male participation,” Health Policy and Plan- ning 11(1): 84–92. McElroy, M. B. and M. J. Horney. 1981. “Nash bargained household decisions,” Interna- tional Economic Review (22): 33–50. Mensch, Barbara S., Judith Bruce, and Margaret E. Greene. 1998. The Uncharted Passage: Girls’ Adolescence in the Developing World. New York: Population Council. Mincy, Ronald B. and Hillard Pouncy. 1997. “There must be fifty ways to start a family: Social policy and the fragile families of low-income, noncustodial fathers,” in Wade Horn et al. (eds.), The Fatherhood Movement: A Call to Action. Berkeley: University of California Press. Montgomery, Mark R. and John B. Casterline. 1996. “Social learning, social influences, and new models of fertility,” in John B. Casterline, Ronald D. Lee, and Karen A. Foote (eds.), Fertility in the United States: New Patterns, New Theories. Supplement to Volume 22, Population and Development Review, pp. 151–175 New York: Population Council. Morris, Leo. 1994. “Sexual behavior of young adults in Latin America,” Advances in Popula- tion: Psychosocial Perspectives 2: 231–252. Morris, Martina, Chai Podhisita, Maria J. Wawer, and Mark S. Hancock. 1996. “Bridge populations in the spread of HIV/AIDS in Thailand,” AIDS 10: 1265–1271. Mott, Frank L. and Susan H. Mott. 1985. “Household fertility decisions in West Africa: A comparison of male and female survey results,” Studies in Family Planning 16(2): 88– 99. Mundigo, Axel I. 1995. Men’s Roles, Sexuality, and Reproductive Health. Chicago: The John D. and Catherine T. MacArthur Foundation. Olukoya, Adepeju A. and Christopher Elias. 1996. “Perceptions of reproductive tract morbidity among Nigerian women and men,” Reproductive Health Matters 7: 56–65. Omondi-Odhiambo. 1997. “Men’s participation in family planning decisions in Kenya,” Popu- lation Studies 51: 29–40. Oppong, C. (ed.). 1987. Sex Roles, Population and Development in West Africa: Policy-Related Studies on Work and Demographic Issues. Portsmouth, NH: Heinemann. Orubuloye, I. O, John C. Caldwell, and Pat Caldwell. 1991. “Sexual networking in the Ekiti District of Nigeria,” Studies in Family Planning 22(2): 61–73. ———. 1992. “Diffusion and focus in sexual networking: Identifying partners and partners’ partners,” Studies in Family Planning 23(6): 343–351. Orubuloye, I. O., John C. Caldwell, Pat Caldwell, and Shail Jain (eds.). 1995. The Third World AIDS Epidemic. Supplement to Health Transition Review, Volume 5. Pebley, A. R., H. Delgado, and E. Brineman. 1980. “Family sex composition preferences among Guatemalan men and women,” Journal of Marriage and the Family 42(2): 437– 447. Pebley, Anne R. and Noreen Goldman. 1986. “Legalization of consensual unions in Mexico,” Social Biology 33(3–4): 199–213. Phillips, James F. et al. 1997. “The determinants of contraceptive innovation: A case–control study of family planning acceptance in a traditional African society,” Policy Re- search Division Working Paper No. 93. New York: Population Council.

Click to return to Table of Contents M ARGARET E. GREENE / ANN E. BIDDLECOM 113

Pleck, J. H., F. L. Sonenstein, and L. C. Ku. 1990. “Contraceptive attitudes and intention to use condoms in sexually experienced and inexperienced adolescent males,” Journal of Family Issues 11: 294–312. Pool, Robert, Mary Maswe, J. Ties Boerma, and Soori Nnko. 1996. “The price of promiscuity: Why urban males in Tanzania are changing their sexual behaviour,” Health Tran- sition Review 6: 203–222. Posner, J. K. and F. Mbodji. 1989. “Men’s attitudes about family planning in Dakar, Sene- gal,” Journal of Biosocial Science 21(3): 279–291. Presser, Harriet B. 1997. “Demography, feminism, and the science–policy nexus,” Popula- tion and Development Review 23(2): 295–331. Rao, V. 1997. “Wife-beating in rural South India: A qualitative and econometric analysis,” Social Science and Medicine 44(8): 1169–1180. Rao, V. and M. E. Greene. 1993. “Marital instability, inter-spousal bargaining, and their implications for fertility in Brazil,” Population Research Center Working Paper, Uni- versity of Chicago. Rendall, Michael S. et al. 1999. “Incomplete reporting of men’s fertility in the United States and Britain: A research note,” Demography 36(1): 135–144. Renne, Elisha P. 1993. “Gender ideology and fertility strategies in an Ekiti Yoruba village,” Studies in Family Planning 24(6): 343–353. ———. 1996. “Shifting boundaries of fertility change in Southwestern Nigeria,” Health Tran- sition Review 6: 169–190. Rich, Adrienne Cecile. 1986. Of Woman Born: Motherhood as Experience and Institution. New York: W. W. Norton. Riley, Nancy E. 1997. “Similarities and differences: Anthropological and demographic perspectives on gender,” in David I. Kertzer and Tom Fricke (eds.), Anthropological Demog- raphy: Toward a New Synthesis, pp. 115–138. Chicago: University of Chicago Press. Ringheim, Karin. 1993. “Factors that determine prevalence of use of contraceptive methods for men,” Studies in Family Planning 24(2): 87–99. ———. 1996. “Whither methods for men? Emerging gender issues in contraception,” Re- productive Health Matters 7: 79–89. Robinson, Warren C. 1997. “The economic theory of fertility over three decades,” Popula- tion Studies 51: 63–74. Rogow, Deborah and Sonya Horowitz. 1995. “Withdrawal: A review of the literature and an agenda for research,” Studies in Family Planning 26(3): 140–153. Rosen, R. H. and T. Benson. 1982. “The second class partner: The male role in family planning decisions,” in G. L. Fox (ed.), The Childbearing Decision, pp. 97–124. Beverly Hills, CA: Sage Publications. Ross, John A. and Douglas H. Huber. 1983. “Acceptance and prevalence of vasectomy in developing countries,” Studies in Family Planning 14(3): 67–73. Rutenberg, Naomi, Ann K. Blanc, and S. Kapiga. 1994. “Sexual behavior, social change, and family planning among men and women in Tanzania,” Health Transition Review 4 (Supplement): 173–196. Rutenberg, Naomi and Susan Cotts Watkins. 1997. “The buzz outside the clinics: Conversa- tions and contraception in Nyanza Province, Kenya,” Studies in Family Planning 28(4): 290–307. Sachs, A. 1994. “Men, sex, and parenthood,” World Watch 7(2): 12–19. Sarkar, N. N. 1993. “Sterilisation: Characteristics of vasectomy acceptors in Delhi,” Journal of Biosocial Science 25(1): 45–49. Schneider, Jane C. and Peter T. Schneider. 1991. “Sex and respectability in an age of fertility decline: A Sicilian case study,” Social Science and Medicine 33(8): 885–895. ———. 1996. Festival of the Poor: Fertility Decline and the Ideology of Class in Sicily, 1860–1980. Tucson: University of Arizona Press. Shryock, Henry S. and Jacob S. Siegel. 1976. The Methods and Materials of Demography. San Diego, CA: Academic Press.

Click to return to Table of Contents 114 M ALE REPRODUCTIVE ROLES

Smith, H. L. and S. P. Morgan. 1994. “A response model for agreement and disagreement between husbands and wives regarding fertility intentions,” paper presented at the Annual Meeting of the Population Association of America, 5–7 May, Miami, FL. Speizer, I. S. 1995. “A marriage trichotomy and its applications,” Demography 32(4): 521– 532. Speizer, I. S. and A. J. Yates. 1996. “Couples as innovators in Sub-Saharan Africa,” paper presented at the Annual Meeting of the Population Association of America, 9–11 May, New Orleans, LA. Stycos, J. Mayone. 1996. “Men, couples, and family planning: A retrospective look,” Cornell University Population and Development Program Working Paper Series No. 96.12. Ithaca, NY: Cornell University. Stycos, J. Mayone, Kurt Back, and Reuben Hill. 1956. “Problems of communication between husband and wife on matters relating to family limitation,” Human Relations 1: 207–215. Szreter, Simon. 1993. “The idea of demographic transition and the study of fertility change: A critical intellectual history,” Population and Development Review 19(4): 659–701. Thomson, Elizabeth. 1989. “Dyadic models of contraceptive choice, 1957 and 1975,” in D. Brinberg and J. Jaccard (eds.), Dyadic Decision Making, pp. 268–285. New York: Springer-Verlag. ———. 1990. “Two into one: Structural models of couple behavior,” in T. Draper and A. Marcos (eds.), Family Variables: Conceptualization, Measurement and Use, pp. 129–142. Newbury Park, CA: Sage Publications. ———. 1997a. “Couple childbearing desires, intentions, and births,” Demography 34(3): 343– 354. ———. 1997b. “Hers, his and their children: Influences on U.S. couple childbearing decisions,” paper presented at the Annual Meeting of the Population Association of America, 27–29 March Washington, DC. Thomson, Elizabeth, Elaine McDonald, and Larry L. Bumpass. 1990. “Fertility desires and fertility: His, hers, and theirs,” Demography 27(4): 579–588. Tietze, Christopher. 1938. “The measurement of differential reproduction by paternity rates,” Eugenics Review 30(2): 101–107. ———. 1943. “Differential reproduction in the United States: Paternity rates for occupational classes among the urban white population,” American Journal of Sociology 49(3): 242–247. Townsend, Nicholas. 1997. “Reproduction in anthropology and demography,” in David I. Kertzer and Tom Fricke (eds.), Anthropological Demography: Toward a New Synthesis, pp. 96–144. Chicago: University of Chicago Press. ———. 1999. “Male fertility as a lifetime of relationships: Contextualizing men’s biological reproduction in Botswana,” in Caroline Bledsoe, Susana Lerner, and Jane I. Guyer (eds.), Fertility and the Male Life Cycle in the Era of Fertility Decline, pp. 343–364. New York: Oxford University Press. Tsui, Amy O., S. Victor de Silva, and Ruth Marinshaw. 1991. “Pregnancy avoidance and coital behavior,” Demography 28(1): 101–117. United Nations. 1995. Report of the International Conference on Population and Development, Cairo, 5–13 September 1994. New York. van de Walle, Francine and Marian Maiga. 1991. “Family planning in Bamako, Mali,” Inter- national Family Planning Perspectives 17(3): 84–90, 99. VanLandingham, M. J. et al. 1993. “Sexual activity among never-married men in Northern Thailand,” Demography 30(3): 297–313. Watkins, Susan Cotts. 1993. “If all we knew about women was what we read in Demogra- phy, what would we know?” Demography 30(4): 551–577. Watkins, Susan Cotts, Naomi Rutenberg, and David Wilkinson. 1997. “Orderly theories, disorderly women,” in G. W. Jones, R. M. Douglas, J. C. Caldwell, and R. M. D’Souza

Click to return to Table of Contents

M ARGARET E. GREENE / ANN E. BIDDLECOM 115

(eds.), The Continuing Demographic Transition, pp. 213–245. New York: Oxford Univer- sity Press. Westoff, Charles F. 1994. “What’s the world’s priority task? Finally, control population,” New York Times Magazine (6 February): 30, 32. Westoff, Charles F. and Akinrinola Bankole. 1995. Unmet Need: 1990–1994. DHS Compara- tive Studies No. 16. Calverton, MD: Macro International. Whiting, Beatrice B. (ed.). 1963. Six Cultures: Studies of Child Rearing. New York: John Wiley and Sons. Whiting, Beatrice B. and John W. M. Whiting, in collaboration with Richard Longabaugh. 1975. Children of Six Cultures: A Psycho-cultural Analysis. Cambridge, MA: Harvard Uni- versity Press. World Health Organization. 1996. “Cervical cancer control in developing countries: Memo- randum from a WHO meeting,” Bulletin of the World Health Organization 74(4): 345– 351. Worth, Dooley. 1989. “Sexual decision-making and AIDS: Why condom promotion among vulnerable women is likely to fail,” Studies in Family Planning 20(6): 297–307. Zulu, Eliya. 1997. “Men’s role in childbearing decisions relating to modern versus traditional contraceptives: A look at evidence from Malawi,” paper presented at the An- nual Meeting of the Population Association of America, 27–29 March, Washington, DC.

Click to print article Click to return to Table of Contents Click to return to Table of Contents

NOTES AND COMMENTARY

Rethinking the African AIDS Epidemic

JOHN C. CALDWELL

THE AFRICAN AIDS epidemic is a contemporary health crisis of staggering proportions and one with which African society and governments, as well as the international health system, have signally failed to cope. This need not have been so, and need not continue to be so. We now know enough about the social context of the epidemic, and the interventions that would probably succeed, to begin to limit the epidemic’s impact without waiting for the development of vaccines or depending on antiretroviral drugs for prolonging life. On a global scale HIV/AIDS is one of many problems. Estimates by UNAIDS and WHO (1999) up to the end of 1999 show that a total of 50 million people have either died of AIDS or are infected and probably ultimately doomed to die of the disease. This number compares with estimated mortality from the Black Death in fourteenth-century Europe of 20 million and a similar number of deaths globally in the influenza pandemic in the second decade of the twentieth century. There is no guarantee that eventual AIDS mortality will not double or rise even higher: it is not a classic epidemic that burns itself out. Both the Black Death and influenza stayed in individual localities no longer than two to three months, while AIDS has been in parts of Africa for nearly two decades. This difference may be related to the fact that the latency period from infection to symptoms is only a matter of days in the case of bubonic plague and influenza, while with AIDS it is closer to a decade. AIDS has probably reduced the world’s current annual population growth rate from 1.5 percent to 1.4 percent. But there are parts of the world where the situation is comparable to the Black Death, which killed around one-third of Europe’s population. Al- most 70 percent of persons with HIV/AIDS and over 80 percent of those who have died of the disease are found in sub-Saharan Africa, and over 50 percent of those now infected are located in mainland East and Southern

POPULATION AND DEVELOPMENT REVIEW 26(1):117–135 (MARCH 2000) 117

Click to print article Click to return to Table of Contents 118 R ETHINKING THE AFRICAN AIDS EPIDEMIC

Africa, whose population of about 265 million is less than 5 percent of the world’s total (Schwartländer et al. 1999: 2452–2454). There, at the end of 1997, the adult (15–49 years) HIV prevalence rate was 11.4 percent, although it exceeds 25 percent in Zimbabwe and Botswana (UNAIDS and WHO 1998: 64–65). Because the average duration from HIV infection to death in Africa is only about 8.5 years, these seroprevalence levels translate into lifetime chances of dying from AIDS of around 40 percent in Eastern and Southern Africa as a whole and 70 percent in Zimbabwe and Botswana (Blacker and Zaba 1997). Since 1997 the prevalence rate for the region has risen, especially rapidly in Southern Africa. Behavioral change has limited the AIDS epidemic elsewhere, but the approaches employed have been tried nowhere in Africa on a national scale and in only a few places on a smaller scale. Some of the most striking examples of such change come from the homosexual epidemic in Western countries. The example used here is the Australian epidemic because it is well documented, and early success in controlling the epidemic was achieved through cooperation between the government and the gay community (Dowsett 1990, 1993, 1999; Ballard 1998). Members of the gay community had earlier begun to recognize themselves as a group with specific sex preferences about which they could talk unashamedly, and increasingly took the attitude that they were at particular risk of HIV infection and that only gay solidarity in confronting the epidemic could ward off disaster. The government responded with a sympathetic publicity campaign and with such practical steps as implementing needle exchanges for intravenous drug users. Efforts intensified after 1984, with HIV incidence subsequently falling by 70 percent over the next four years and by 86 percent over nine years (National Centre in HIV Epidemiology and Clinical Research 1998: 8). At first it was believed that the major causes of such a rapid decline must have been a substantial reduction in sexual relationships with high-risk partners and a greater resort to monogamous pairing. Later research has shown that control of the epidemic was achieved mainly through a high level of compliance with condom use and strong gay community support in emphasizing the necessity for such compliance. Peer group education was essential. Emphasis was placed on different sexual cultures and on a safe sex culture (Dowsett 1999). Effective programs have also been carried out in developing countries, the most pertinent and noteworthy example probably being Thailand. There, HIV transmission was predominantly heterosexual and much of the transmission took place in brothels. Health inspectors achieved a very high level of condom use by threatening brothel owners that the police would close the premises if it was shown that prostitution was taking place without condom use. Both government-planned and spontaneous publicity given by the media to the epidemic led to clients at the brothels more readily using

Click to return to Table of Contents J OHN C. CALDWELL 119 condoms and to fewer men going to such places. In the next two years levels of sexually transmitted disease (STD) fell steeply among prostitutes, and HIV prevalence among army recruits declined by two-thirds (Hanenberg et al. 1994). This article examines why such approaches are not being used in Af- rica, how they might be used, and whether they would be likely to be effective. Just how different is Africa? The evidence outlined here has been derived from ongoing collaborative social and behavioral research in the region over the last 12 years, beginning in West Africa, subsequently involving parallel research in Uganda, Nigeria, and Ghana, and including other groups in conferences and workshops.1 The project first looked at the nature of sexual activity and its levels and causes (Caldwell, Caldwell, and Quiggin 1989, 1991; Caldwell, Orubuloye, and Caldwell 1992b; Orubuloye, Caldwell, and Caldwell 1991, 1992, 1993a). Some work was done on the geographic spread of the epidemic (Caldwell and Caldwell 1993). In total, this research described a society where land had not been owned or inherited by individual families and where a high level of disease always threatened premature death. The result was that traditional religion, unlike the major religions of Asia and Europe from which the society was long insulated, placed its greatest emphasis on fertility rather than on confining female sexual activity within marriage (Caldwell and Cald- well 1987; Caldwell, Caldwell, and Quiggin 1989). This meant that women were in less danger from relatives or society than in the Old World agrarian societies if they engaged in premarital or extramarital sex or if they resorted to prostitution either full-time or occasionally (Orubuloye et al. 1994; Orubuloye, Caldwell, and Caldwell 1994). There was a high level of demand by men for sexual relations with women other than their wives for two reasons. The first reason is that the world’s highest levels of polygyny can be sustained only if husbands are much older than their wives (Cald- well 1963; Goldman and Pebley 1989). This means that most men still cannot marry until their late 20s, and, until 100 years ago, possibly could not do so until their late 30s (Peel 1983: 119). Society has traditionally allowed single men discreet access to sex, usually with relatives (Caldwell, Orubuloye, and Caldwell 1991), rather than face the social instability that its denial would almost certainly cause. The second reason is that over large parts of the region wives are frequently unavailable for sexual relations because of long periods of postpartum sexual abstinence occurring often because of high levels of fertility (Caldwell and Caldwell 1977; Page and Lesthaeghe 1981; Schoenmaeckers et al. 1981). Later research in our project focused on women’s limited control over their sexual activity (Orubuloye, Cald- well, and Caldwell 1993b) and on men’s perceived need for sexual relations with more than one woman (Orubuloye, Caldwell, and Caldwell 1997). The latter is implied by the centrality of polygyny as a social institution, and

Click to return to Table of Contents 120 R ETHINKING THE AFRICAN AIDS EPIDEMIC most men and women believe that men are biologically programmed to need sexual relations with a variety of women. A recent study of men in hotels and bars who suddenly decide to have sex with prostitutes revealed that most of them talked of uncontrollable needs rather than pleasure (Cald- well, Orubuloye, and Caldwell 1999: 117). Both behavioral research and the parity between males and females in HIV infection increasingly showed the sub-Saharan epidemic, uniquely in the world, to be almost exclusively heterosexual. The existence of even one epidemic of this type is surprising because the HIV transmission rate during one act of vaginal intercourse between otherwise healthy persons is so low: perhaps one per thousand acts from the female to the male partner and one in 300 from male to female. This epidemic has been made possible by a number of factors: (1) a higher level of sex outside marriage than occurs in Old World agrarian societies; (2) a high level of prostitution, caused partly by the lack of wives or sexually accessible wives and partly by the mobility of the population and an excess of males in many urban areas; (3) a resulting high level of STDs—the world’s highest level according to a WHO study (1987: 967)—which act as cofactors for infection, thus removing the qualification above about the partners being “otherwise healthy”; (4) the persistence of ulcerating untreated and uncured STDs because of poverty and the world’s poorest health facilities; and (5) a low level of condom use, even in commercial sex. On their own, even this range of conditions is not sufficient to sustain a major AIDS epidemic: the majority of West African countries exhibit adult HIV levels below 2.5 percent (UNAIDS and WHO 1998). At least one additional factor is operating in East and Southern Af- rica: whole ethnic groups not practicing male circumcision (Bongaarts et al. 1989; Moses et al. 1990; Caldwell and Caldwell 1993).

The failure to control the African epidemic

The reasons given to explain how a major heterosexually driven AIDS epidemic came about in sub-Saharan Africa are insufficient to explain why it has persisted. All the conditions listed above are amenable to change by either individual or government initiative. There has in fact been no change recorded at the national level except in Uganda, where HIV infection rates appear to have fallen among the population under 25 years of age (Konde- Lule 1995). This lack of change has for some years now been the focus of our program; our conclusion is that this situation should not have been allowed to persist and is reversible. It is not an insoluble African problem. Our first attempt to draw together our findings and thinking on the matter was in an article titled “Underreaction to AIDS in sub-Saharan Af- rica” (Caldwell, Orubuloye, and Caldwell 1992a). This guided later research and was supplemented in the mid-1990s by an attempt to identify attitudes toward men’s sexual behavior (Orubuloye, Caldwell, and Caldwell 1997).

Click to return to Table of Contents J OHN C. CALDWELL 121

The present article is largely informed by two further experiences. The first was the move to focus our research and that of others, mostly in Africa but with some research for comparison in Asia, on resistances to behavioral change to reduce HIV/AIDS infection in predominantly heterosexual epidemics in developing countries. This effort has resulted in the presentation of research reports at a conference in Canberra in April 1999 and their subsequent collection in a book (Caldwell et al. 1999). The second was my participation in the 11th International Conference on AIDS and STDs in Africa, in September 1999 in Lusaka, Zambia, in following Track 2, “HIV/ AIDS and socio-economic consequences,” which included statements on policies and plans by governments and international organizations, and then summarizing the presentations in a report to the conference’s final session. The “Underreaction” article noted that no one blamed governments for inaction. Nowhere had there been riots or even demonstrations, and this is still the case even in countries where over one-quarter of adults are seropositive and where most of the deaths are from AIDS. This is one of the reasons why governments are not disposed to take effective action. They fear creating for themselves more trouble than the AIDS epidemic already causes them. Heads of state do not wish to be associated with the epidemic, and, in spite of much protest from the Lusaka conference, not a single president or prime minister attended it, even for the opening and closing ceremonies. The other reason for inaction is that many politicians regard the disease largely as their electorate does. Much of the action that has been taken has been by foreign-funded nongovernmental organizations (NGOs) or as part of programs partly planned and largely funded by international donors. The problems that African governments have in communicating with their citizens, with outsiders, and among themselves were exposed at the Lusaka meeting. After 20 years of the AIDS epidemic, policy discussions are still full of abstract planning language, with promises to organize, decen- tralize, and base the work within the community. There was no desire among government representatives to discuss precisely what would work in the local setting and why. This was jarringly different from the down-to-earth language and examples offered at family planning conferences and immu- nization workshops. There is a fear of alienating their followers by intrud- ing into sexual matters and by speaking aloud on such subjects. There is a fear of failure. Above all there is a fear of confronting those who regard the only solution to be confining sexual activity within marriage, either because it is the law of God or because it will Westernize the family and modernize society. Three-quarters of Christian leaders in Nigeria believe that AIDS is a divine punishment (Orubuloye, Caldwell, and Caldwell 1993a: 99–100), and the proportion is unlikely to be lower in East and Southern Africa or among the laity of the congregations. The Catholic Church holds that the use—and hence the distribution—of condoms is forbidden. Most adults regard it as

Click to return to Table of Contents 122 R ETHINKING THE AFRICAN AIDS EPIDEMIC immoral to provide adolescents with condoms and are reluctant to admit that the great majority are sexually active: either these views are held by the politicians themselves or else the leaders are daunted by those who do hold them. Even though AIDS has already killed well over 10 million Africans and at least 20 million more are moving toward death (UNAIDS and WHO 1998), there is little political gain to be had by helping to stop those numbers from multiplying. On the other hand, we now know things that were not clear a decade ago. The first is that the unrestricted continuation of the epidemic is not a failure of the AIDS educational effort. The Demographic and Health Survey program has shown that among men 98 percent knew of AIDS in 1991–92 in Tanzania, 99 percent in 1992 in Zambia and in 1998 in Kenya. For women at those dates the proportions were 93 percent in Tanzania and 99 percent in Zambia and Kenya. A more anthropological approach can show an inadequate knowledge of some methods of transmission or of the nature of the pathogen, but almost all Africans know that there is a sinister new disease, AIDS, that it is sexually transmitted, that it is more likely to be caught if one has multiple partners or participates in commercial sex, and that the disease kills most people it infects. A decade ago it was believed that such knowledge should be sufficient to contain the epidemic. In this sense the educational approach has failed, although massive education will still be needed, partly to show that the message has not been reconsidered and partly because it might achieve other ends such as strengthening the fear of death. We also know that the tide can probably be turned against the epidemic by the same means—or at least variations of them—used to contain it elsewhere. Two international firms in the Ivory Coast have provided their workers and their families with good health services and have made condoms readily available to any individuals in these families (Chevalier 1999). STD levels and apparently HIV incidence have fallen as steeply as in Thai- land. The same approach has been taken in the villages of the Niger River Delta in Nigeria from which the oil companies draw their workers, with the result that STD prevalence fell by 40 percent in one year (Feleyimu 1999). Efforts to ensure a higher level of condom use during sex with prostitutes might well succeed in Africa even though the circumstances are less propitious than in Thailand. In Africa there is the problem of different degrees of commercialism in sexual relations, but even the restriction of efforts to the most commercial of relations, where sex is paid for on the spot and where the young women work in bars and hotels where prostitution is allowed or encouraged, would be worthwhile. These women have the highest number of different partners and consequently usually the highest seroprevalence as a group, and a successful condom intervention here would have a disproportionate impact on the epidemic. In Africa prostitutes are

Click to return to Table of Contents J OHN C. CALDWELL 123 usually not employees but separately rent their own rooms to provide sexual services. Nevertheless, these rooms are usually found in a limited number of premises whose owners are susceptible to argument even from individual research teams as in Ado-Ekiti and Shagamu in Nigeria (Orubuloye and Oguntimehin 1999a; Dada 1999) and would be much more likely to com- ply with to well-organized pressure from government. Police and other officials can cooperate even in programs organized by private researchers (Esu- Williams 1995). Admittedly many of the prostitutes’ clients complain that condoms rob them of a feeling of intimacy, but the clients’ sensitivities may be blunted by the fact that commercial sex is often said to be an urgent need (Caldwell, Orubuloye, and Caldwell 1999: 117–118), that it is often an adven- ture for the young, or that is often undertaken with a degree of drunken- ness. Research in South Africa has shown that this reduction in intimacy is a major reason why prostitutes prefer the use of condoms and may well cooperate in their use being made almost mandatory. They want to feel less intimate with their customers than with their husbands or other regular partners (Varga 1997: 81). Research has shown that the level of sexual relations is high among single people and that their average number of different partners is greater than among the older and married (Orubuloye, Caldwell, and Caldwell 1991; Meekers 1994). HIV incidence levels are highest among the young, much being explained by adolescent sex. It is becoming increasingly clear from both family planning and AIDS research that, irrespective of the official policies of governments or NGOs, the unmarried, especially adolescents, find it almost impossible to obtain condoms (or other contraceptives) from health services or family planning clinics (Arkutu 1995; Stanback et al. 1997; Olowu 1998; Konje et al. 1998; Mturi and Hinde 1998). We postulate that seroprevalence would now be falling rather than rising in East and Southern Africa if governments provided strong positive leadership in identifying the AIDS epidemic as needing the government involvement and social mobilization typical of a state of total war (the mortality figures are similar to those one would expect from prolonged armed conflict); in exerting effective pressure to raise the level of condom use in completely commercial sexual relationships to 90 or 95 percent; and in ensuring that sexually active adolescents have easy access to condoms and are encouraged to use them. The reasons why neither society as a whole nor governments in particular are moving in this direction are interrelated and are the subject of the rest of this article.

Relevant aspects of the society

We reported in the early 1990s that one element in the failure to control the AIDS epidemic was an extraordinarily stoical attitude toward death (Cald-

Click to return to Table of Contents 124 R ETHINKING THE AFRICAN AIDS EPIDEMIC well, Orubuloye, and Caldwell 1992a: 1178–1179). Subsequent research, especially investigation focused on this point in the research leading up to the 1999 Canberra conference, supported this interpretation (Orubuloye and Oguntimehin 1999b; Caldwell, Orubuloye, and Caldwell 1999; Awusabo-Asare et al. 1999; Amuyunzu-Nyamongo et al. 1999). According to health transition theory, one reason for the bravery in the face of death relates to only a limited progression in health transition (Simons 1989; Caldwell 1993). With the transition from an agrarian society based on family production and usually characterized by high mortality, to a more urbanized, nonfarming economy or market-oriented farming characterized by employment by nonrelatives and usually low mortality, individuals become more sensitive to the risk of death and this in itself is a powerful force in reducing mortality further. The earlier society was insensitive to death partly because it was less avoidable and so common that individuals and communities had to accept it. In agrarian joint families, daughters-in-law showed their primary loyalty by supporting the family’s productive effort and deferred to the patriarch and matriarch of the family. The unacceptable daughter-in-law was one who placed a major stress on caring for her children’s health, or her own health or even her husband’s (Caldwell and Caldwell 1992). Where we worked in rural South India an illiterate daughter-in-law was in danger if she drew attention to her child’s illness before her mother-in-law did so (Caldwell, Reddy, and Caldwell 1983: 196). Over time in the West death came to be regarded as absolutely the worst outcome, qualitatively different from all other outcomes. Individuals felt ever more strongly the duty to intervene to reduce the risk of death to their children, their spouses, themselves, and others (Simons 1989). Mod- ern education contains a strong message that this should be so. Sub-Sa- haran Africa has consistently exhibited the highest mortality in the world, with a life expectancy of 37 years in the early 1950s and perhaps 49 years now (United Nations 1999a; Population Reference Bureau 1999), and the lowest educational level (World Bank 1997: 226–227). Much of the recent research suggests that this high level of mortality is still an important reason for a distinctly careless attitude to life in Kenya, Zambia, Ghana, and Zimbabwe (Höjer 1999; Anarfi 1999; Awusabo-Asare et al. 1999; Mupemba 1999). Some men practicing high-risk sex say that, if the latency period until onset of AIDS is nearly a decade, they are not worried because they are likely to die of something else in such a long time. There are, however, other forces: for example, a widespread belief that at least some role is played by predestination, that one’s time to die was determined long ago, an attitude perhaps strongest in West Africa (Fortes 1983: 7). Recent research findings in this regard were summarized by Awusabo-Asare et al. (1999) in the Ghanaian proverb, “All die be die”; by Orubuloye and Oguntimehin (1999b) by the Nigerian attitude that “Death

Click to return to Table of Contents J OHN C. CALDWELL 125

. . . will come when it is due”; and for East Africa by Amuyunzu-Nyamongo and colleagues (1999) in the philosophy that “Everyone will die anyway.” In southern Nigeria only one-fifth of respondents said that they were afraid of death, even premature death, and half linked this to a philosophy of predestination (Caldwell, Orubuloye, and Caldwell 1999: 120). They knew it was not their time to die because of their robust spirits, often as displayed by their sexuality. Also pervasive is the belief that the underlying cause of premature death is witchcraft, performed by an enemy or by an expert at the behest of an enemy; and even many who recognize the role of infection and pathogens believe these are merely the intermediary mechanisms. In these circumstances there is little point in avoiding one type of infection only to find that the malevolent forces settle for another mechanism. In the more urbanized, educated, Christianized, and Muslimized areas these beliefs are waning, but two-fifths of southern Nigerian respondents thought it possible that these forces played a role and half thought so in rural areas (ibid.: 118). The great majority thought AIDS was different from other diseases, not merely in its near-universal mortality, but in that it was caused by malevolent forces or was a divine punishment. This largely explains the extraordinary silence about the disease, a matter to which we return because the silence is a basic cause of individual and government inaction. Two further points should be made, one about males, the other about adolescents. The first is that most Africans believe that males are biologically programmed to require sex with more than one woman (Orubuloye, Caldwell, and Caldwell 1997). It could hardly be otherwise in a society where polygyny is a central institution. In most of West and Middle Africa over 40 percent of currently married women are in polygynous marriages (Lesthaeghe et al. 1989: 276–277), and a much higher proportion will be in such marriages in the course of a lifetime. In countries of East Africa the most common proportion is now 20–39 percent but it probably was higher in the past. Many men feel that high-risk sex is unavoidable, as was clear from research in Ado-Ekiti, Nigeria on men in bars and hotels where sex was offered (Caldwell, Orubuloye, and Caldwell 1999; Orubuloye and Oguntimehin 1999b). The late age of men at marriage—a product of the polygynous system—and the large numbers of wives sexually inaccessible at any specific time aggravate the situation. In much of West, Middle, and East Africa the male singulate mean age at first marriage exceeds 25 years and in Southern Africa exceeds 27 years (Lesthaeghe, Kaufmann, and Meekers 1989: 272–273). The second point is the situation of adolescents. In Ghana, Awusabo- Asare et al. (1999) argue that risk-taking, especially sexual risk-taking, almost defines adolescence. Anarfi (1999) found that sex among Accra street children was necessary for mutual support, companionship, and binding

Click to return to Table of Contents 126 R ETHINKING THE AFRICAN AIDS EPIDEMIC the group together. Across sub-Saharan Africa there are almost irresistible peer-group pressures on both male and female adolescents to have sexual relations, and on girls by their boyfriends to demonstrate their love or to maintain the relationship by giving in to the boys’ sexual advances (Cald- well and Caldwell 1987: 239–241; Varga 1999: 22–25; Preston-Whyte 1999: 146–148). Probably common in urban areas of the region is the situation found in Ibadan City, Nigeria, where fathers were distant and frequently absent and mothers busy, so that girls depended on their boyfriends for affection and adolescents of both sexes needed peer groups to at least partly substitute for families (P. Caldwell and J. C. Caldwell 1987). There is a need for policies that realistically address adolescent culture, lifestyle, and sexuality (Dowsett et al. 1998).

The current policy situation

Africans have been educated by AIDS programs to know that the disease is deadly and is largely spread among them by high-risk sexual activities. The epidemic cannot be defeated by more education. Both the educated and the religious leaders find it hard to accept that Africans have chosen to maintain their sexual system and to accept the risk of AIDS in the face of evidence that a primary means of avoiding it is to restrict sex to marriage. Most Africans feel that this is no choice at all, but few publicly put it into words. We reported that research in Nigeria on sexual activity in the late 1980s could proceed only when we assured people that we were linked neither with the churches nor with governments, institutions that were felt to take a jaundiced view of sexual activity and to be willing to use the AIDS epidemic to limit it (Orubuloye, Caldwell, and Caldwell 1991: 63). The situation is misunderstood and the problems aggravated because of the silence surrounding the AIDS epidemic. This silence is extraordinary and had not been predicted. It goes far toward explaining the passivity of the people in the face of the ravages of the epidemic, the failure of governments to speak out, and the fact that governments have not faced great protest against their inactivity. There is less public or media discussion of AIDS in Zimbabwe, with an adult seroprevalence level approaching 30 percent, than there is in Thailand with a level of 2 percent. The situation is still much the same as it was in the early 1990s when summarized in Caldwell, Orubuloye, and Caldwell (1992a). There is little public and even less private discussion of AIDS. People repeatedly attend funerals of AIDS victims without discussing the probability that the cause of death was AIDS and without challenging the statements of family members of the deceased that the cause was something quite different even though the timing and nature of the death are suspicious. In Nigeria, where the adult seroprevalence level was already more than 4 percent at the end

Click to return to Table of Contents J OHN C. CALDWELL 127 of 1997 (UNAIDS and WHO 1998) and where people frequently go to funerals, research shows that hardly any of the respondents were convinced that they had attended an AIDS funeral. Most could name only one person who they were certain had died of the disease and that was Fela, the fa- mous Lagos-based musician (Caldwell, Orubuloye, and Caldwell 1999: 116– 117). This was because his brother, a prominent medical figure, had announced the cause of death to the press, but the rest of the family denied this diagnosis as had Fela himself during his period of sickness, at least partly because he claimed the disease did not exist. The silence is similar in the press. It mentions AIDS deaths but does not bring immediacy to the situation with stories of the affliction or death of known people with the disease. The situation is similar with known contacts: at a time when the seroprevalence level among Lagos prostitutes was said to be 20 percent (Ransome-Kuti 1992), no prostitute in a large-scale survey knew anyone infected, probably because sick girls silently left for home (Orubuloye, Cald- well, and Caldwell 1994: 114–115). The silence exists partly because people have been taught that AIDS is a sexually transmitted disease associated with sexual activity outside marriage. The churches have taught that this is a shameful thing and many believe the infection was a punishment for sin and are reluctant to disclose that any of their relatives bear such witness. The silence should, however, be seen in a broader context. Sex outside marriage is widely practiced in the region but it is rarely the subject of macho boasting and usually not talked about at all. One reason is that men feel they must seek extramarital sexual relations when their wives are practicing postpartum abstinence but that family stability is maintained by never saying so (P. Caldwell and J. C. Caldwell 1987: 243–244). Wives are not supposed to recognize their husbands’ extramarital sexual activities and the pretense is well maintained. In a study of the Ondo Town area of Nigeria, only 10 percent of the extramaritally sexually active husbands said it was likely that their wives knew of their activities (Orubuloye, Caldwell, and Caldwell 1992: 349). Such a small percentage of knowledgeable wives is an absurd assessment. There is an inability to discuss sexual activity across generational or gender lines, especially within marriage. Fathers find it almost impossible to discuss their sons’ sexual activities, possibly because they are not supposed to discuss within the family their own extramarital sexual activity. In South Africa mothers bring their sexually active daughters to family planning clinics without having discussed with the daughters why they need to go, and without being able to tell the clinic staff why they have brought their daughters. In these circumstances most parents do not wish the government to admit the existence of adolescent sex, let alone facilitate it or reduce the risks by providing condoms. Many church leaders concur. This attitude has presented both AIDS and family planning programs with great difficulties.

Click to return to Table of Contents 128 R ETHINKING THE AFRICAN AIDS EPIDEMIC

Families are also silent about members suffering from HIV/AIDS because they fear isolation or ostracism from neighbors or, in a few reported cases, violence. But the main reason for the silence is almost certainly the apprehension about treating in a cavalier and open way such an unusual disease, in which witchcraft or divine punishment may be an element. It would be asking for trouble. Witchcraft or other occult manifestations are either whispered about or not mentioned at all. Only two out of five Nige- rian respondents were certain that AIDS was just another disease with no other-worldly component (Caldwell, Orubuloye, and Caldwell 1999: 118) but even they are likely to adopt the attitude that Westerners used to take about cancer—something that really should not be discussed. The silence, in turn, makes AIDS seem even more unusual. The silence is also made possible by the long latency period of the disease. It almost takes an act of faith to relate the appearance of AIDS symptoms to sexual encounters that occurred almost a decade ago, given that at least 95 percent of seropositive Africans do not know of their infection until the symptomatic AIDS stage. These circumstances have allowed some Africans to deny the link or even the existence of AIDS (Orubuloye and Oguntimehin 1996b: 108). Here again, it is their being accustomed to a high level of mortality that is the critical issue. Condom use is still a long way from defeating the sub-Saharan Afri- can AIDS epidemic. Sub-Saharan Africans are probably more hostile to the use of condoms than people in any other major region of the world. The Demographic and Health Surveys show rapid increase in condom use, but the figures are usually for ever-use and are never for constant use, and the rises begin from a very low base. For 1998, the United Nations (1999b) estimates current condom use (by women of reproductive age) to be 14 percent in developed regions on average, reaching 46 percent in Japan; 4 percent in Latin America; 3 percent in Asia; and only 1 percent in Africa and Oceania, the latter dominated by Melanesia. African males typically complain of the loss of sensual enjoyment, and women fear injury or ill health, often citing stories of women who died of infection after condoms were sucked into their wombs. Nevertheless, use of condoms is increasing because of worry about AIDS, and, with greater use, the fear of them is diminishing. During the 1990s and especially during the Abacha regime in Nigeria (1993–98), foreign technical aid was reduced in such a way as to restrict the inflow of most contraceptives but to increase the inflow of condoms to combat AIDS. The result was that Nigerians increasingly used condoms also as contraceptives (on the earliest part of this period, see Caldwell, Orubuloye, and Cald- well 1992b). Many adolescents everywhere are ready to use condoms if they can be obtained, because nothing else meets the need when sex is suddenly decided upon and the girl is not using a contraceptive method.

Click to return to Table of Contents J OHN C. CALDWELL 129

Some NGOs, most with foreign connections and funding, have been successful in the social marketing of condoms and have shown there is a demand, probably greater than they can meet. However, there is opposition to condom use and misinformation campaigns are common, often associated with Christian, particularly Catholic groups, and often, too, with foreign connections. The charges against condoms range from exaggerated accounts of the likelihood of the HIV retrovirus passing through them to claims of the deliberate puncturing of some of them and the charge that their use is mainly responsible for the epidemic, sometimes by their being covered by tiny spikes that inject the pathogens. It is surprising how many people have read or heard of these charges and how many have been at least partly convinced of their truth. Given the magnitude of the epidemic, one of its most bizarre aspects is the reliance placed on NGOs. Governments, unwilling to be very active themselves, are happy to have issues of sexuality and AIDS tackled by these organizations; or at the very least they are susceptible to pressure by foreign donors to allow the NGOs to operate. International donors trust the NGOs to have a realistic attitude toward the epidemic and are therefore anxious that they should continue their work even if national governments suc- cumb to external pressure to mount major programs. At the 1999 Lusaka Conference on AIDS and STDS in Africa, the code word for this situation, understood by both the international donors and the African officials, was “multisectoral,” and it was repeatedly argued by donors that future programs should remain multisectoral. Many NGOs do excellent work, but relying on them alone is hardly sufficient to meet a crisis on this scale.

Defeating the AIDS epidemic

The African AIDS epidemic can be defeated by means already known and easily implemented. Its defeat should be an international priority justifying international pressure on, and inducements offered to, the African governments involved. At the end of 1997 almost 10 million sub-Saharan Africans had died of AIDS and another 20 million were infected and waiting to die (UNAIDS and WHO 1998). If the present level of government inaction continues, it is likely that 50 million Africans will die before there is an effective vaccine, and numbers could go much higher still. To begin to contain the epidemic will not necessitate stopping all HIV transmission. What is necessary, as with all epidemics, is to ensure that the average number of people infected by each person already infected falls below one, preferably well below one. This can be achieved by reducing transmission where it is at the highest levels, through ensuring both that a very high proportion of strictly commercial sexual relations are accompanied by the use of condoms and that a high proportion of premarital adolescent

Click to return to Table of Contents 130 R ETHINKING THE AFRICAN AIDS EPIDEMIC sexual activity involves the use of condoms. Both would require much highly motivated activity by national governments. These activities would be most likely to succeed if they were accompanied by (1) a continuing education program focusing as now on the danger of HIV infection, but stressing more the necessity of using condoms and the horror of unnecessary death; (2) an efficient system for distributing condoms by different routes, some allowing the recipients to preserve their anonymity; and (3) a much greater provision for STD treatment so as to reduce the role of cofactors in HIV transmission. Governments have not moved to a kind of wartime footing in order to carry out these steps for a number of reasons. First, there is a lack of reality in the society and in governments about the enormous scale of deaths and impending deaths. Second, there is an astonishing reluctance by governments and many church leaders to recognize the African sexual system. This was the case a decade ago among many African academics, researchers, and intellectuals when we published “The social context of AIDS in sub-Saharan Africa” (Caldwell, Caldwell, and Quiggin 1989) and “The family and sexual networking in sub-Saharan Africa” (Caldwell, Caldwell, and Oruboloye 1992), but most seem to have changed their views. In contrast, governments and some outside agencies misinterpret the failure of the educational effort to curb the epidemic as a sign that the effort was not good enough or was not understood, rather than that it met with deliberate resistance from people satisfied with their sexual system and unprepared to change. Third, there is still no real recognition by governments or many religious leaders of the rights of civil society, in which people deserve help in avoiding death even if their sexual behavior is at odds with the preached orthodoxy. This summary captures the essence of the situation, and is the one on which action should be based, but it still needs some modification. The first point is that there has apparently been a decline in HIV levels in one Afri- can country, Uganda, and note should be taken of its experience. As yet the evidence for that decline comes from a very small number of urban antena- tal clinics, and no evaluation has been published of possible changes in the composition of the clients. A recent analysis of the situation concluded that behavioral change must be occurring, but only because this was the residual category after other changes had been excluded (Kilian et al. 1999). As- suming that a decline in seroprevalence has taken place, the reason is almost certainly the strong leadership of President Yoweri Museveni since 1986 and his insistence that AIDS should be discussed and identified as a national crisis needing action. On the other hand, there is little evidence of unusually high levels of condom use among adolescents or in commercial sex. Uganda was one of the first two or three countries in sub-Saharan Af- rica to suffer from a major AIDS epidemic, and effective behavioral change took another dozen years to occur. The longer experience of AIDS deaths

Click to return to Table of Contents J OHN C. CALDWELL 131 may have had an impact, and AIDS mortality, selective in terms of high- risk sexual behavior, may also have played a role. Before the change came, Uganda had lost almost 10 percent of its population to AIDS deaths, with another 5 percent currently infected, probably the world’s greatest losses and together similar to the Soviet Union’s proportional losses in World War II (UNAIDS and WHO 1998). This is not, then, such a success story as suggested by those placing all their hopes on behavioral change. The second point is that there is not the homogeneity of risk that is found among homosexual partners in the West. Most women do not lead sex lives as fraught with risk as those of their husbands (Orubuloye, Cald- well, and Caldwell 1991). Probably about half of those with the disease were infected by their husbands. Therefore, they need support in persuading their husbands to lead less risky sex lives. They also need to be able to reject husbands suspected of high-risk sexual behavior (see Orubuloye, Caldwell, and Caldwell 1993b). But their even more immediate need is to reduce the likelihood of their husbands being infected, utilizing all the means suggested here. The central plank in the victory over AIDS is the recognition by African governments of social and sexual reality. Millions of people are being allowed to die on the grounds that the only way they can be saved is by adopting a more “moral” way of life, indeed a way of life that does not conform to their morality. An unreal view of society still stalks African regional AIDS meetings. Working within a framework of reality will be difficult. In the case of sexually active adolescents, if they are to recognize themselves as a sexual community with special needs in the way Australia’s gay community did, then they will have to be induced to provide their own organizations and leadership (in East and Southern Africa age grade organizations may provide a basis). Such organizations may prove to be the best way of channel- ing condoms to them. Something similar may be part of the solution in the case of prostitutes, except that the government will have to exert its strength to protect them by forcing the cooperation of landlords, bar and hotel owners, and ultimately their clients. The vigor will have to come from the heads of state and their more powerful ministers, and the encouragement may have to come from donors. At present, reality is slipping away, as the Lusaka Conference demonstrated, toward placing hope in millions of villagers successfully living on courses of antiretroviral medications or turning to indigenous herbal medicine. Finally, strong informational programs must continue to point out the reduction in the risk of AIDS from changed sexual behavior, but this should not be presented as the only option and vigorous efforts will be needed to make the other options as risk-free as possible.

Click to return to Table of Contents 132 R ETHINKING THE AFRICAN AIDS EPIDEMIC

Notes

This work has benefited from assistance from Kofi Awusabo-Asare, University of Cape Rachel Colombo, Wendy Cosford, and Elaine Coast, Ghana; James Ntozi, Makerere Uni- Hollings and advice from Pat Caldwell, John versity, Uganda; and John and Pat Caldwell, Ballard, Penny Kane, Jacob Malungo, and Australian National University. The SIDA/ Colin Caldwell. The research program has SAREC Adviser has been Per Bolme. The been primarily funded by SIDA/SAREC, conferences on research findings have been Swedish technical aid, with earlier assistance recorded in Caldwell et al. 1993; Orubuloye from Health Sciences, Rockefeller Founda- et al. 1994; Orubuloye et al. 1995; Ntozi et tion and the Australian National University. al. 1997; Setel, Chirwa, and Preston-Whyte 1 The principal investigators have been 1997; Orubuloye, Caldwell, and Ntozi 1999; I. O. Orubuloye, State University, Ado-Ekiti, and Caldwell et al. 1999. Nigeria; John Anarfi, University of Ghana;

References

Anarfi, John K. 1999. “Initiating behavioural change among street-involved youth: Find- ings from a youth clinic in Accra,” in Caldwell et al., 1999, pp. 81–90. Arkutu, A. 1995. “Family planning in sub-Saharan Africa: Present status and future strategies,” International Journal of Gynecology and Obstetrics 50 (Supplement 2): S27–S34. Amuyunzu-Nyamongo, M. et al. 1999. “Barriers to behaviour change as a response to STD including HIV/AIDS: The East African experience,” in Caldwell et al., 1999, pp. 1–11. Awusabo-Asare, Kofi, Albert M. Abane, Delali M. Badasu, and John K. Anarfi. 1999. “‘All die be die’: Obstacles to change in the face of HIV infection in Ghana,” in Caldwell et al., 1999, pp. 125–132. Awusabo-Asare, Kofi, J. Ties Boerma, and Basia Zaba (eds.). 1997. Evidence of the Socio-demographic Impact of AIDS in Africa. Supplement 2 to Health Transition Review 7. Canberra: Australian National University. Ballard, John. 1998. “The constitution of AIDS in Australia: Taking ‘government at a distance’ seriously,” in Mitchell Dean and Barry Hindess (eds.), Governing Australia: Stud- ies in Contemporary Rationalities of Government. Cambridge: Cambridge University Press, pp. 125–138. Blacker, John and Basia Zaba. 1997. “HIV prevalence and lifetime risk of dying of AIDS,” in Awusabo-Asare, Boerma, and Zaba, 1997, pp. 45–62. Bongaarts, John, Priscilla Reining, Peter Way, and Francis Conant. 1989. “The relationship between male circumcision and HIV infection in African populations,” AIDS 3(6): 373–377. Caldwell, John C. 1963. “Fertility decline and female chances of marriage in Malaya,” Popu- lation Studies 17(1): 20–32. ———. 1993. “Health transition: The cultural, social and behavioural determinants of health in the Third World,” Social Science and Medicine 36(2): 125–135. Caldwell, John C. et al. (eds.). 1993. Sexual Networking and HIV/AIDS in West Africa. Supple- ment to Health Transition Review 3. Canberra: Australian National University. Caldwell, John C. et al. (eds.). 1999. Resistances to Behavioural Change to Reduce HIV/AIDS Infection in Predominantly Heterosexual Epidemics in Third World Countries. Canberra: Aus- tralian National University. Caldwell, John C. and Pat Caldwell. 1977. “The role of marital sexual abstinence in determining fertility: A study of the Yoruba in Nigeria,” Population Studies 31(2): 193–217. ———. 1987. “The cultural context of high fertility in sub-Saharan Africa,” Population and Development Review 13(3): 409–437.

Click to return to Table of Contents J OHN C. CALDWELL 133

———. 1992. “Family systems: Their viability and vulnerability. A study of intergenera- tional interactions and their demographic implications,” in Elza Berquó and Peter Xenos (eds.), Family Systems and Cultural Change. Oxford: Clarendon Press, pp. 46–66. ———. 1993. “The nature and limits of the sub-Saharan African AIDS epidemic: Evidence from geographic and other patterns,” Population and Development Review 19(4): 817– 848. Caldwell, John C., Pat Caldwell, and I. O. Orubuloye. 1992. “The family and sexual networking in sub-Saharan Africa: Historical regional differences and present-day implications,” Population Studies 46(3): 385–410. Caldwell, John C., Pat Caldwell, and Pat Quiggin. 1989. “The social context of AIDS in sub- Saharan Africa,” Population and Development Review 15(2): 185–234. ———. 1991. “The African sexual system: Reply to Le Blanc et al.,” Population and Develop- ment Review 17(3): 506–515. Caldwell, John C., I. O. Orubuloye, and Pat Caldwell. 1991. “The destabilization of the traditional Yoruba sexual system,” Population and Development Review 17(2): 229–262. ———. 1992a. “Underreaction to AIDS in sub-Saharan Africa,” Social Science and Medicine 34(11): 1169–1182. ———. 1992b. “Fertility decline in Africa: A new type of transition?” Population and Devel- opment Review 18(2): 211–242. ———. 1999. “Obstacles to behavioural change to lessen the risk of HIV infection in the African AIDS epidemic: Nigerian research,” in Caldwell et al., 1999, pp. 113–124. Caldwell, John C., P. H. Reddy, and Pat Caldwell. 1983. “The social component of mortality decline: An investigation in South India employing alternative methodologies,” Popu- lation Studies 37(2): 185–205. Caldwell, Pat and John C. Caldwell. 1987. “Fertility control as innovation: A report on in- depth interviews in Ibadan, Nigeria,” in Etienne van de Walle and J. Akin Ebigbola (eds.), The Cultural Roots of African Fertility Regimes. Philadelphia: University of Penn- sylvania, and Ile-Ife, Nigeria: Obafemi Awolowo University, pp. 233–251. Chevalier, Evelyn. 1999. “HIV/AIDS and workplace interventions in Côte d’Ivoire,” report to 11th International Conference on AIDS and STDs in Africa, Lusaka, 12–16 September. Dada, Kayode. 1999. “A condom distribution scheme to commercial sex workers in Shagamu, Nigeria,” report to 11th International Conference on AIDS and STDs in Africa, Lusaka, 12–16 September. Dowsett, Gary W. 1990. “Reaching men who have sex with men—an overview of AIDS education: Community intervention and community attachment strategies,” Austra- lian Journal of Social Issues 25(3): 186–198. ———. 1993. “Sustaining safe sex: Sexual practices, HIV and social context,” AIDS 7 (Supple- ment 1): S257–S262. ———. 1999. “Understanding cultures of sexuality: Lessons learned from HIV/AIDS education and behaviour change among gay men in Australia,” in Caldwell et al., 1999, pp. 223–231. Dowsett, Gary W. et al. 1998. “Changing gender relations among young people: The global challenge for HIV/AIDS prevention,” Critical Public Health 8(4): 291–309. Esu-Williams, Eka. 1995. “Sexually transmitted diseases and condom interventions among prostitutes and their clients in Cross River State,” in Orubuloye et al., 1995, pp. 223–228. Feleyimu, Bode. 1999. “Report on project of the Chevron Oil Company in the Niger Delta, Nigeria,” to 11th International Conference on AIDS and STDs in Africa, Lusaka, 12– 16 September. Fortes, Meyer. 1983. Oedipus and Job in West African Religion. Cambridge: Cambridge Univer- sity Press. Goldman, Noreen and Anne Pebley. 1989. “The demography of polygyny in sub-Saharan Africa,” in Lesthaeghe, 1989, pp. 212–237. Hanenberg, R. S., W. Rojanapithayakorn, P. Kunasol, and D. Sokal. 1994. “Impact of Thai-

Click to return to Table of Contents 134 R ETHINKING THE AFRICAN AIDS EPIDEMIC

land’s HIV-control programme as indicated by the decline of sexually transmitted diseases,” Lancet 344(8917): 243–245. Höjer, Bengt. 1999. “The community–health services interface: The critical issue for AIDS prevention,” in Caldwell et al., 1999, pp. 59–63. Kilian, Albert H. D. et al. 1999. “Reductions in risk behaviour provide the most consistent explanation for declining HIV-1 prevalence in Uganda,” AIDS 13(3): 391–398. Konde-Lule, Joseph K. 1995. “The declining HIV seroprevalence in Uganda: What evidence?” in Orubuloye et al., 1995, pp. 27–33. Konje, Justin C., Folashade Oladini, Emmanual O. Otolorin, and Oladapo O. Ladipo. 1998. “Factors determining the choice of contraceptive methods at the Family Planning Clinic, University College Hospital, Ibadan, Nigeria,” British Journal of Family Planning 24(3): 107–110. Lesthaeghe, Ron J. (ed.). 1989. Reproduction and Social Organization in Sub-Saharan Africa. Berkeley: University of California Press. Lesthaeghe, Ron, Georgia Kaufmann, and Dominique Meekers. 1989. “The nuptiality regimes in sub-Saharan Africa,” in Lesthaeghe, 1989, pp. 238–337. Meekers, Dominique. 1994. “Sexual initiation and premarital childbearing in sub-Saharan Africa,” Population Studies 48(1): 47–64. Moses, Stephen et al. 1990. “Geographical patterns of male circumcision practices in Africa: Association with HIV seroprevalence,” International Journal of Epidemiology 19(3): 693– 697. Mturi, Akin J. and P. R. Andrew Hinde. 1998. “Fertility and family planning in Tanzania,” paper presented to IUSSP Seminar on Reproductive Change in Sub-Saharan Africa, Nairobi, November. Mupemba, Karen. 1999. “The Zimbabwe HIV prevention program for truck drivers and commercial sex workers: A behaviour change intervention,” in Caldwell et al., 1999, pp. 133–137. National Centre in HIV Epidemiology and Clinical Research. 1998. HIV/AIDS and Related Diseases in Australia: Annual Surveillance Report 1998. Sydney: University of New South Wales. Ntozi, James P. M., John K. Anarfi, John C. Caldwell, and Shail K. Jain (eds.). 1997. Vul- nerability to HIV Infection and Effects of AIDS in Africa and Asia/India. Supplement to Health Transition Review 7. Canberra: Australian National University. Olowu, Folarin. 1998. “Quality and costs of family planning as elicited by an adolescent mystery client trial in Nigeria,” African Journal of Reproductive Health 2(1): 49–60. Orubuloye, I. O., John C. Caldwell, and Pat Caldwell. 1991. “Sexual networking in the Ekiti District of Nigeria,” Studies in Family Planning 22(2): 61–73. ———. 1992. “Diffusion and focus in sexual networking: Identifying partners and partners’ partners,” Studies in Family Planning 23(6): 343–351. ———. 1993a. “The role of religious leaders in changing sexual behaviour in southwest Nigeria in an era of AIDS,” in Caldwell et al., 1993, pp. 93–104. ———. 1993b. “African women’s control over their sexuality in an era of AIDS: A study of the Yoruba of Nigeria,” Social Science and Medicine 37(7): 859–872. ———. 1997. “Perceived male sexual needs and male sexual behaviour in southwest Nige- ria,” Social Science and Medicine 44(8): 1195–1207. Orubuloye, I. O., John C. Caldwell, Pat Caldwell, and Shail Jain (eds.). 1995. The Third World AIDS Epidemic. Supplement to Health Transition Review 5. Canberra: Australian National University. Orubuloye, I. O., John C. Caldwell, Pat Caldwell, and Gigi Santow (eds.). 1994. Sexual Net- working and AIDS in Sub-Saharan Africa: Behavioural Research and the Social Context. Canberra: Australian National University. Orubuloye, I. O., John C. Caldwell, and James P. M. Ntozi (eds.). 1999. The Continuing HIV/ AIDS Epidemic in Africa: Responses and Coping Strategies. Canberra: Australian National University.

Click to return to Table of Contents

J OHN C. CALDWELL 135

Orubuloye, I. O., Pat Caldwell, and John C. Caldwell. 1994. “Commercial sex workers in Nigeria in the shadow of AIDS,” in Orubuloye et al., 1994, pp. 101–116. Orubuloye, I. O. and Folakemi Oguntimehin. 1999a. “Intervention for the control of STDs including HIV among commercial sex workers, commercial drivers and students in Nigeria,” in Orubuloye, Caldwell, and Ntozi, 1999, pp. 121–129. ———. 1999b. “Death is pre-ordained, it will come when it is due: Attitudes of men to death in the presence of AIDS in Nigeria,” in Caldwell et al., 1999, pp. 101–111. Page, Hilary J. and Ron Lesthaeghe (eds.). 1981. Child-Spacing in Tropical Africa: Traditions and Change. London: Academic Press. Peel, J. D. Y. 1983. Ijeshas and Nigerians: The Incorporation of a Yoruba Kingdom, 1890s–1970s. Cambridge: Cambridge University Press. Population Reference Bureau. 1999. World Population Data Sheet, 1999. Washington, DC. Preston-Whyte, Eleanor. 1999. “Reproductive health and the condom dilemma: Identify- ing situational barriers to HIV protection in South Africa,” in Caldwell et al., 1999, pp. 139–155. Ransome-Kuti, Olikoye. 1992. Report to the Meeting of the National AIDS Committee and Media Executives. Lagos: Ministry of Health. Schoenmaeckers, R., I. H. Shah, R. Lesthaeghe, and O. Tambashe. 1981. “The child-spacing tradition and the postpartum taboo in tropical Africa: Anthropological evidence,” in Page and Lesthaeghe, 1981, pp. 25–71. Schwartländer, Bernard et al. 1999. “Country-specific estimates and models of HIV and AIDS: Methods and limitations,” AIDS 13(17): 2445–2458. Setel, Philip W., Wiseman Chijere Chirwa, and Eleanor Preston-Whyte (eds.). 1997. Sexual Networking, Knowledge and Risk: Contextual Social Research for Confronting AIDS and STDs in Eastern and Southern Africa. Supplement 3 to Health Transition Review 7. Canberra: Australian National University. Simons, John. 1989. “Cultural dimensions of the mother’s contribution to child survival,” in John C. Caldwell and Gigi Santow (eds.), Selected Readings in the Cultural, Social and Behavioural Determinants of Health. Canberra: Australian National University, pp. 132– 145. Stanback, John, Andy Thompson, Karen Hardee, and Barbara Janowitz. 1997. “Menstrua- tion requirements: A significant barrier to contraceptive access in developing countries,” Studies in Family Planning 28(3): 245–250. UNAIDS and WHO. 1998. Report on the Global HIV/AIDS Epidemic, June 1998. Geneva: Joint United Nations Programme on HIV/AIDS, and World Health Organization. ———. 1999. AIDS Epidemic Update: December 1999. Geneva: Joint United Nations Programme on HIV/AIDS, and World Health Organization. United Nations. 1999a. World Population Prospects: The 1998 Revision, Vol. 1, Comprehensive Tables. ST/ESA/SER.A/177. New York. ———. 1999b. World Contraceptive Use 1998, Fact Sheet. ST/ESA/SER.A/175. New York. Varga, Christine A. 1997. “The condom conundrum: Barriers to condom use among commercial sex workers in Durban, South Africa,” African Journal of Reproductive Health 1(1): 74–88. ———. 1999. “South African young people’s sexual dynamics: Implications for behavioural responses to HIV/AIDS,” in Caldwell et al., 1999, pp. 13–34. World Bank. 1997. World Development Report 1997. New York: Oxford University Press. World Health Organization. 1987. “Infections, pregnancies, and infertility: Perspectives on prevention,” Fertility and Sterility 47(6): 964–968.

Click to print article Click to return to Table of Contents Click to return to Table of Contents DATA AND PERSPECTIVES

Long-Term Population Projections and the US Social Security System

RONALD LEE

IT IS WIDELY believed that future population aging in the United States will cause a financial crisis for the Social Security System and for the federal budget of which it is a part. With a wave of retirees commencing with the retirement of the baby boom generations—those born between 1946 and 1964—and continuing after they are gone, the costs of Social Security benefits are projected to exceed the payroll tax revenues earmarked for the system after 2014; and in 2022 tax revenues plus interest on the Social Se- curity Trust Fund will fall short of payouts, so the Fund will begin to decline. The value of the Fund is projected to fall to zero in 2034 if policy is left unchanged. These problems have been prominent in news reports, in the minds of citizens old and young, and in political debates. Problems of this sort are particularly acute for systems, such as the US Social Security System, where in practical terms pension obligations are backed up by the willingness of future workers to pay the necessary taxes to support the elderly, rather than by accumulated real assets. The problem is widely recognized in the United States, in large part because Social Secu- rity actuaries routinely and commendably prepare long-term projections of financial balance over a 75-year horizon, based on the decisions of the Board of Trustees (the governing body of the Social Security System) about assumptions and published in the Trustees’ Annual Report (most recently, Board of Trustees 1999). Indeed, the public pension programs of many other industrial nations, where such long-term projections are not done and where the magnitude of the problems often goes unnoticed, have far more serious problems than the US program. The problems are more serious in a number of other countries for several reasons. First, their public pension systems provide more generous benefits after retirement, up to 70 or 80 per-

POPULATION AND DEVELOPMENT REVIEW 26(1):137–143 (MARCH 2000) 137

Click to return to Table of Contents 138 P OPULATION PROJECTIONS AND US SOCIAL SECURITY cent of preretirement wage levels compared to an average of 40 percent by US Social Security. Second, retirement is typically several years earlier in European nations. Third, fertility is lower, with an average current total fertility rate (TFR) of about 1.4 in Europe and Japan versus about 2.0 in the United States. Fourth, life expectancy is higher in Japan and in many Euro- pean countries outside of Eastern Europe than it is in the United States. Producing a financial projection over a 75-year horizon is a daunting task, fraught with uncertainty. To assist the actuaries in carrying out these projections, a Technical Panel on Assumptions and Methods has been appointed every four years in recent decades. The most recent version of this panel was given a somewhat broader charge, not only to consider the basic projections but also to consider how to assess the financial implications of the various policy proposals that have been offered, including proposals that some fraction of the receipts of the Trust Fund be invested in the stock market, or that individual retirement accounts be established to partially or fully substitute for the unfunded system. In November 1999, the panel1 issued its report (The 1999 Technical Panel), which received considerable attention from the press because its recommendations implied a deterioration in the estimated financial status of the system. Whereas the most recent Trustees Report indicated that long-term balance could be restored by raising the payroll tax rate by 2.07 percentage points—from the current 12.4 percent to 14.47 percent—the Technical Panel report suggested that the necessary increase would be .5 percentage points larger, that is, 2.6 percentage points. Adopting this recommendation would mean that the policy reforms, which had been carefully tailored to cover the imbalance of 2.07 percentage points, would fail to cover the shortfall according to the new estimate. The increased estimate of the shortfall was largely due to the panel’s recommendation to project a larger gain than was formerly assumed in life expectancy in the future. This alone accounted for an increase of .51 percentage points. The other main recommendations were offsetting. A recommendation that the projected rate of increase of covered wages (calculated in real terms) be raised from .9 percent per year to 1.1 percent per year led to a .2 percentage points improvement in the imbalance, while the recommendation of a lower real rate of return on the government bonds held in the Trust Fund—2.7 percent per year instead of 3.0 percent per year— made the imbalance .2 percentage points worse. In this note I discuss only the demographic recommendations of the panel, although I commend the report in its entirety to anyone interested in the finances of the system. I first discuss briefly the recommendations about assumptions on future fertility and immigration and then at more length the mortality assumptions. Financial projections for the system are extremely sensitive to the assumed trajectory of fertility. For a number of years, the Trustees Report has

Click to return to Table of Contents R ONALD LEE 139 assumed that the TFR will ultimately be 1.9 children per woman, with a range of 1.6 to 2.2. The Technical Panel found this to be a reasonable assumption, although it recommended careful monitoring of trends, particularly regarding ethnic differences in fertility. Since the average TFR in Eu- rope is 1.4, and Japan’s TFR is similar, and since some industrial countries have fertility near a TFR of 1.1 or 1.2, one might conclude that 1.9 is too high. In many European countries, however, the average age of childbearing has been rising by .1 or .2 years per calendar year, which would depress the period TFR by perhaps 10 percent or 20 percent below the level of completed fertility for cohorts (see, for example, Bongaarts and Feeney 1998, discussing this effect). Similar forces in the United States appear to have been at least partially responsible for TFRs around 1.7 or 1.8 during the 1970s and 1980s. Furthermore, surveys indicate that women in Europe, like women in the United States, expect to have about 2 children. Econo- mies in Europe have been affected by high unemployment, particularly for the young, housing shortages, and other impediments to family formation. It seems possible, therefore, although far from certain, that the European situation reflects factors that, although quite persistent, are not permanent. Alternatively, one might think that assuming a TFR of 1.9 for the United States is too low over the long run. Fertility of non-Hispanic white women in the United States has been about 1.8 in recent years, and it is the higher fertility of minority groups including immigrants that has raised the aggregate level to slightly over 2.0. But the population share of minorities in the United States is projected to rise from the current 28 percent to 47 percent by 2050, given continuing immigration. The increasing share of this population component will raise its influence on the average level of fertility, perhaps causing a rising TFR in coming decades. Some studies suggest that there is relatively little convergence of minority-group fertility levels toward the non-Hispanic white level. It is also important to recognize, however, that the major sending countries for US immigrants are themselves in the midst of fertility transition, and indeed Mexico’s TFR is projected by the United Nations to fall to 2.2 by 2015. One might expect, therefore, that immigration would cease to fuel high fertility for the dominant Latino and Asian ethnic groups in the future. For these reasons, the panel decided that an ultimate fertility of 1.9 children per woman was a reasonable assumption, while acknowledging that there is no compelling basis for choosing that level over levels somewhat higher or lower. There is a great deal of uncertainty in forecasting fertility, as a review of the past record of forecasts demonstrates. The panel also accepted the Trustees’ assumption that net immigration would be 900,000 persons per year, while recommending that the range of high–low assumptions bracketing that figure be considerably widened. Rates of immigration will be largely determined by future policy. Even

Click to return to Table of Contents 140 P OPULATION PROJECTIONS AND US SOCIAL SECURITY though the annual number of immigrants has been rising nearly linearly over the past several decades, the actuaries are not permitted to extrapolate that rising trend into the future because their assumptions must be consistent with existing law. In any event, as shown by the sensitivity analyses in the Trustees’ annual reports and confirmed by Lee and Miller (2000), the finances of the system are fairly insensitive to variations in the assumed annual numbers of immigrants, variations that matter far less than plausible variations in fertility or mortality. I now turn to discuss the panel’s assumptions on mortality. Life expectancy at birth in the United States was 76.7 years (sexes combined) in 1998 according to the National Center for Health Statistics. In recent years, the Trustees Report has assumed that it would rise to 81.8 years by 2075, or by about five years. The panel recommended that this assumption be raised by about an additional four years (3.7 years for 2070), in line with the Trust- ees’ “High cost” assumption for mortality improvement. The panel considered three kinds of evidence in making this recommendation: international trends in mortality, historical trends in the United States, and the possibility of biomedical limits on longevity. I will touch briefly on each of these. Life expectancy in Japan in 1997 for males and females combined was 80.5 years. According to the Trustees’ intermediate assumption, this level will not be reached in the United States until 2052, or 55 years later than in Japan. The 1995 life expectancy levels for France and Sweden would similarly not be reached in the United States for another 20 or 30 years. An influential study of mortality decline above age 80 for countries with good data (Kannisto et al. 1994) found that the average rate of decline in the probability of dying in nine countries has accelerated from the 1950s to the 1980s. During the 1980s, the average rate of that decline in 19 countries was 1.8 percent for women in their 80s and 1.2 percent for women in their 90s, while for men the corresponding figures were 1.2 percent and 1.0 percent. The long-term rates of decline assumed by the Trustees Report are one-half to one-third of these rates, at .6 percent per year for women in both age groups and .5 percent for men. Based on the US record, a continuation of the rates of mortality decline observed since 1900 would lead to life expectancy of just over 86 years in 2075, slightly higher than the 85.9 years implied by the assumptions on mortality recommended by the panel. A continuation of rates of decline observed in the second half of the twentieth century (since 1950) would lead to an almost identical gain. Only if the base period is limited to the last two decades does the outcome change much, and even in this case a life expectancy of 84 years would be attained. The panel also considered whether biomedical limits might prevent achievement of an average life expectancy this high. It was noted that the study by Kannisto et al. found no significant association of rates of mortal-

Click to return to Table of Contents R ONALD LEE 141 ity decline with the levels of mortality previously achieved, although one would expect a negative association if a limit were being approached. Re- markable biomedical breakthroughs have been announced frequently in recent years with unknown implications for extension of life, but with potentially highly favorable consequences. Studies of nonhuman populations have shown that age-specific death rates may reach a plateau or even decline at extreme old age, rather than continue to increase exponentially. While some analysts such as Olshansky, Carnes, and Grahn (1998) emphasize the difficulties in achieving substantial gains in life expectancy, the panel believed it was prudent for the Trustees to raise the projected improvement in mortality as described. A National Academy of Sciences workshop on forecasting life expectancy, which included biologists and physicians as well as demographers and statisticians, supported projections of the sort the panel recommended, at least for the short and medium term, and also supported a method for carrying them out (Stoto and Durch 1993; Lee and Carter 1992). The panel made a second important recommendation about assumptions as to the future course of mortality: that the pattern of decline of age- specific death rates should be relatively flat after age 25 or so, with the assumed decline at older ages as rapid as the decline at younger ages. This additional recommendation, which may sound like a minor technical foot- note, in fact adds substantially to the cost imposed on the Social Security System by the increase in life expectancy. When mortality falls, survival improves not only at the oldest ages, but also in childhood and during the working years. During the earlier part of the twentieth century, mortality fell most strongly in childhood. From childhood on, rates of decline fell steadily during the working years, and were slowest for the elderly. Under these circumstances, mortality decline is less costly for Social Security, because mortality decline in the early years of life raises the growth of the potential labor force, and the increased number of workers helps to pay the pension cost of the increased number of the elderly. But in the second half of the century, the age pattern of mortality decline shifted, becoming flat- ter: decline in the working ages slowed, and decline at older ages accelerated. This pattern in the US mortality experience is also observed in other industrial nations such as Japan, Sweden, and France. The relatively slug- gish improvement of mortality among younger adults in part reflects the AIDS epidemic and more frequent deaths due to accidents and violence, but this is not the whole story. The panel noted that the rate of decline of mortality for women has slowed substantially in the last two decades, particularly at older ages, relative both to the historical trend in the United States and to the corresponding international experience. It is an important task for research to discover why this is so, and to ascertain whether this slowdown is likely to continue

Click to return to Table of Contents 142 P OPULATION PROJECTIONS AND US SOCIAL SECURITY in the future, for the outcome would have a major effect on the size of the projected older population over the long term. The panel’s recommendation to increase the projected gains in life expectancy assumes that the decline is a transitory phenomenon. It may be conjectured, for instance, that the slowdown reflects a delayed response to the practice of smoking by women. In addition to making recommendations about fertility, immigration, and mortality, the panel stressed the great uncertainty surrounding the long- run calculations. The Trustees project an intermediate imbalance in the So- cial Security Trust Fund amounting to 2.07 percentage points of the wage bill subject to payroll tax as described earlier, but they also project high and low figures for this estimate ranging from a shortfall of –4.97 percentage points to a surplus of +.23 percentage points. This range is thus 5.2 percentage points wide, while the net effect of the changes recommended by the panel for the demographic projection is to adjust the intermediate imbalance of the Trust Fund by only .5 percentage points, or less than one-tenth of the range of uncertainty as provided by the Trustees Report. In light of these relative magnitudes, it would be a mistake to attach too much importance to the precise results of the intermediate demographic projection. To be sure, the policies proposed to “fix” Social Security are all pointed at this intermediate projection, which therefore assumes great political importance. But this simply reflects a shortcoming of the proposed remedial measures. No policy should stand or fall on its ability to strike any particular point in this range of uncertainty. We need policies that can adapt to a variety of possible futures and that are robust to the uncertainty of the future as it unfolds, either in the demographic or in the economic domain. The panel’s mortality projection recommendation would apply even more strongly to the pension programs of most European countries and Japan. An analysis by Keilman (1997) of the past record of population projections in Europe indicates that there was a systematic tendency to underestimate the size of the elderly population. A recent study by Tuljapurkar, Li, and Boe (1999) has found that in the G7 countries, mortality projections along the lines recommended by the Technical Panel would lead to a life expectancy in 2050 that is 2 to 4 years higher than in current official forecasts, with an even greater gap in the case of Japan. The current projections generally assume that life expectancy will have reached a plateau by 2050, with no further increases to follow, so the gap would grow much larger by 2075. As pension programs around the world grapple with long- term funding problems and the possibilities of reform, it is important that they realistically assess the demographic outlook through well-founded long- term projections. In this endeavor, demographers have much to offer.

Click to return to Table of Contents R ONALD LEE 143

Note

1 The panel was chaired by Eugene Smeeding, and Michael Sze. These members Steuerle and included Andrew Abel, Barry were drawn from a variety of fields includ- Bosworth, Edward Frees, Ronald Lee, Debo- ing demography, economics, actuarial sci- rah Lucas, David McKusick, Martha F. Riche, ence, finance, and public policy. John Rust, Andrew Samwick, Timothy M.

References

Board of Trustees, Federal Old-Age and Survivors Insurance and Disability Insurance Trust Funds. 1999. The 1999 Annual Report. Washington, DC: US Government Printing Of- fice. Bongaarts, John and Griffith Feeney. 1998. “On the quantum and tempo of fertility,” Popu- lation and Development Review 24(2): 271–291. Kannisto, Väinö, Jens Lauritsen, A. Roger Thatcher, and James W. Vaupel. 1994. “Reduc- tions in mortality at advanced ages: Several decades of evidence from 27 countries,” Population and Development Review 20(4): 793–810. Keilman, Nico. 1997. “Ex-post errors in official population forecasts in industrialized countries,” Journal of Official Statistics (Statistics Sweden) 13(3): 245–277. Lee, Ronald D. and Lawrence R. Carter. 1992. “Modeling and forecasting U.S. mortality,” Journal of the American Statistical Association 87(419): 659–671. Lee, Ronald and Timothy Miller. 2000. “Immigration, social security, and broader fiscal impacts,” paper presented at the Annual Meeting of the American Economic Associa- tion, Boston, January, and forthcoming in American Economic Review (May). Olshansky, S. Jay, Bruce A. Carnes, and Douglas Grahn. 1998. “Confronting the boundaries of human longevity—Many people now live beyond their natural lifespans through a form of ‘manufactured time,’” American Scientist 86(1): 52–61. Stoto, Michael A. and Jane S. Durch. 1993. “Forecasting survival, health, and disability: Report on a workshop,” Population and Development Review 19(3): 557–581. The 1999 Technical Panel on Assumptions and Methods. 1999. Report to the Social Security Advisory Board (November). Tuljapurkar, Shripad, Nan Li, and Carl Boe. 1999. “Is there a universal pattern of mortality decline? Evidence and forecasts for the G-7 countries.” Unpublished technical report, Mountain View Research, CA. >www.mvr.org<

Click to return to Table of Contents Click to return to Table of Contents

ARCHIVES

Emile Zola Against Malthusianism

Fertility declined in France earlier than in the rest of Western Europe and remained lower than that of its neighbors throughout the nineteenth century and into the twentieth. France’s birth rate in 1900 was around 22 per 1000, compared to about 29 in Britain and 35 in Germany. Worry over depopulation, absolute or relative, has long been a staple element of French population thought. In the late nineteenth century, that concern was expressed in scholarly but vigorous works like Arsène Dumont’s Dépopulation et civilisation (1890) and Natalité et démocratie (1898) and in political activism through the National Alliance for the Growth of the French Population. Other population ideas, not always compatible, were current as well— most notably, variants of Malthusianism. This was also a time of ferment in social policy debate over the implications of new ideas about public health and hygiene and about heredity and environment. While supporters and opponents of Malthusian views could often be identified with the political right and left, combating depopulation was the cause of all. Equally, imperial ambition was not confined to one side of politics: few contradictions were seen between socialism at home and colonization abroad. (French territorial ambi- tions at this time looked particularly to North and West Africa.) Unsurprisingly, many of these themes also cropped up in contemporary novels—among them, those of Emile Zola. Born in 1840 of Italian and French parents, Zola was one of the best-known writers of his time. His many novels include Nana (1880) and Germinal (1885). He is most celebrated, however, for his passionate open letter J’accuse (1898), denouncing the French high command over the Dreyfus Affair. (Alfred Dreyfus, an army officer, was wrongfully convicted of treason in an atmosphere of anti-Semitism—a judgment eventually reversed.) Always somewhat of a propagandist, Zola, in temporary exile in the after- math of this intervention, embarked on a cycle of four novels on the themes of fertility, work, truth, and justice. Fécondité (Paris: Charpentier-Fasquelle, 1899) was

POPULATION AND DEVELOPMENT REVIEW 26(1):145–152 (MARCH 2000) 145

Click to return to Table of Contents Click to print article 146 A RCHIVES the first of these. Two others were completed: Travail (1901) and Vérité (1903). Justice had barely been begun before his death in 1902. An English translation of Fécondité, by Ernest A. Vizetelly, was published under the euphemistic title Fruit- fulness (London: Chatto and Windus; New York: Doubleday, Page, 1900). (Zola often skirted the margins of what was then considered acceptable language and subject matter, and Vizetelly had been jailed in England for an earlier Zola translation.) Fécondité is a didactic moral fable rather than a significant work of fiction. The Fortnightly Review (London) of January 1900 wrote of it: “The tale is a simple one: the cheerful conquest of fortune and the continual birth of offspring.” Mathieu and Marianne Froment, the central characters, convey Zola’s anti-Malthusian views through their life story—the meaning of which is underlined in the author’s ful- some commentary. At the start of the novel, they are poor but already have four children. By its end, still stalwart and celebrating 70 years of marriage, they have had twelve, seven surviving, together with innumerable grandchildren and great grandchildren. Over the same period, through hard work and prudence, they have gradually amassed a large and highly productive landed estate, Chantebled, much of it acquired from once-rich but feckless (and unprolific) neighbors whose decline in fortune mirrors the Froments’ rise and whose depopulationist views are thereby shown to be groundless. (A fuller précis, also describing the novel’s gothic subplots, is given in Michael S. Teitelbaum and Jay M. Winter, The Fear of Population Decline (Academic Press, 1985), pp. 25–27.) The excerpts below are taken from the 1900 translation. (The page numbers refer to the more accessible 1925 reprint.) The characters mentioned, aside from Mathieu, are:

Beauchêne, a relative of Marianne, owner of a farm equipment factory Boutan, family physician and friend of Mathieu Moineaud, a mechanic in Beauchêne’s factory Santerre, a fashionable novelist Séguin, “a rich, elegant idler” whose estate is gradually lost to the Froments

The absence of a female voice on the matters discussed is faithful to the novel.

Beauchêne waxed passionate on the subject. That question of the birth- rate and the present-day falling off in population was one of those which he thought he had completely mastered, and on which he held forth at length authoritatively. He began by challenging the impartiality of Boutan, whom he knew to be a fervent partisan of large families. He made merry with him, declaring that no medical man could possibly have a disinter- ested opinion on the subject. Then he brought out all that he vaguely knew of Malthusianism, the geometrical increase of births, and the arithmetical in-

Click to return to Table of Contents E MILE ZOLA 147 crease of food-substances, the earth becoming so populous as to be reduced to a state of famine within two centuries. It was the poor’s own fault, said he, if they led a life of starvation; they had only to limit themselves to as many children as they could provide for. The rich were falsely accused of social wrong-doing; they were by no means responsible for poverty. Indeed, they were the only reasonable people; they alone, by limiting their families, acted as good citizens should act. And he became quite triumphant, repeating that he knew of no cause for self-reproach, and that his ever-growing fortune left him with an easy conscience. It was so much the worse for the poor, if they were bent on remaining poor! In vain did the doctor urge that the Malthusian theories were shattered, that the calculations had been based on a possible, not a real, increase of population; in vain too did he prove that the present-day economic crisis, the evil distribution of wealth under the capitalist system, was the one hateful cause of poverty, and that when- ever labour should be justly apportioned amongst one and all the fruitful earth would easily provide sustenance for happy men ten times more numerous than they are now. The other refused to listen to anything, took refuge in his egotism, declared that all those matters were no concern of his, that he felt no remorse at being rich, and that those who wished to become rich had, in the main, simply to do as he had done. ‘Then, logically, this is the end of France, eh?’ Boutan remarked ma- liciously. ‘The number of births ever increases in Germany, Russia, and elsewhere, whilst it decreases in a terrible way among us. Numerically the rank we occupy in Europe is already very inferior to what it formerly was; and yet number means power more than ever nowadays. It has been calculated that an average of four children per family is necessary in order that population may increase and the strength of a nation be maintained. You have but one child; you are a bad patriot.’ [pp. 11–12]

. . .

Mathieu, refraining from any intervention, listened and remained grave; for this question of the birth-rate seemed to him a frightful one, the foremost of all questions, that which decides the destiny of mankind and the world. There has never been any progress but such as has been determined by increase of births. If nations have accomplished evolutions, if civilisation has progressed, it is because the nations have multiplied and subsequently spread through all the countries of the earth. And will not to-morrow’s evolution, the advent of truth and justice, be brought about by the constant onslaught of the greater number, the revolutionary fruitfulness of the toilers and the poor? It is quite true that Mathieu did not plainly say all these things to himself; indeed, he felt slightly ashamed of the four children that he al-

Click to return to Table of Contents 148 A RCHIVES ready had, and was disturbed by the counsels of prudence addressed to him by the Beauchênes. But within him there struggled his faith in life, his belief that the greatest possible sum of life must bring about the greatest sum of happiness. [p. 13]

. . .

[Boutan] went on to speak of the prolificacy of wretchedness, the swarming of the lower classes. Was not the most hateful natality of all that which meant the endless increase of starvelings and social rebels? ‘I perfectly understand you,’ Beauchêne ended by saying, without any show of anger, as he abruptly brought his perambulations to an end. ‘You want to place me in contradiction with myself, and make me confess that I accept Moineaud’s seven children and need them, whereas I, with my fixed determination to rest content with an only son, suppress, as it were, a family in order that I may not have to subdivide my estate. France, “the country of only sons,” as folks say nowadays—that’s it, eh? But, my dear fellow, the question is so intricate, and at bottom I am altogether in the right!’ Then he wished to explain things, and clapped his hand to his breast, exclaiming that he was a liberal, a democrat, ready to demand all really progressive measures. He willingly recognised that children were necessary, that the army required soldiers, and the factories workmen. Only he also invoked the prudential duties of the higher classes, and reasoned after the fashion of a man of wealth, a conservative clinging to the fortune he has acquired. Mathieu meanwhile ended by understanding the brutal truth: Capi- tal is compelled to favour the multiplication of lives foredoomed to wretchedness; in spite of everything it must stimulate the prolificness of the wage- earning classes in order that its profits may continue. The law is that there must always be an excess of children in order that there may be sufficient cheap workers. Then also speculation on the wages’ ratio wrests all nobility from labour, which is regarded as the worst misfortune a man can be condemned to, when in reality it is the most precious of boons. Such, then, is the cancer preying upon mankind. In countries of political equality and economical inequality the capitalist régime, the faulty distribution of wealth, at once restrains and precipitates the birth-rate by perpetually increasing the wrongful apportionment of means. On one side are the rich folk with ‘only’ sons, who continually increase their fortunes; on the other, the poor folk, who, by reason of their unrestrained prolificness, see the little they possess crumble yet more and more. If labour be honoured to-morrow, if a just apportionment of wealth be arrived at, equilibrium will be restored. Otherwise social revolution lies at the end of the road.

Click to return to Table of Contents E MILE ZOLA 149

But Beauchêne, in his triumphant manner, tried to show that he possessed great breadth of mind; he admitted the disquieting strides of depopulation, and denounced the causes of it—alcoholism, militarism, excessive mortality among infants, and other numerous matters. Then he indicated remedies; first, reductions in taxation, fiscal means in which he had little faith; then freedom to will one’s estate as one pleased, which seemed to him more efficacious; and a change, too, in the marriage laws, without forgetting the granting of affiliation rights. However, Boutan ended by interrupting him. ‘All the legislative measures in the world will do nothing,’ said the doctor. ‘Manners and customs, our notions of what is moral and what is not, our very conception of the beautiful in life—all must be changed. If France is becoming depopu- lated, it is because she so chooses. It is simply necessary then for her to choose so no longer. But what a task—a whole world to create anew!’ [pp. 15–17]

. . .

[Santerre and Séguin] looked at [Mathieu] and smiled with an air of compassionate superiority. ‘Depopulation an evil!’ exclaimed Séguin; ‘can you, my dear sir, intelligent as you are, still believe in that hackneyed old story? Come, reflect and reason a little.’ Then Santerre chimed in, and they went on talking one after the other and at times both together. Schopenhauer and Hartmann and Nietzsche were passed in review, and they claimed Malthus as one of themselves. But all this literary pessimism did not trouble Mathieu. He, with his belief in fruitfulness, remained convinced that the nation which no longer had faith in life must be dangerously ill. True, there were hours when he doubted the expediency of numerous families and asked himself if ten thousand happy people were not preferable to a hundred thousand unhappy ones; in which connection political and economic conditions had to be taken into account. But when all was said, he remained almost convinced that the Malthusian hypotheses would prove as false in the future as they had proved false in the past. ‘Moreover,’ said he, ‘even if the world should become densely populated, even if food supplies, such as we know them, should fall short, chem- istry would extract other means of subsistence from inorganic matter. And, besides, all such eventualities are so far away that it is impossible to make any calculation on a basis of scientific certainty. In France, too, instead of contributing to any such danger, we are going backward, we are marching towards annihilation. The population of France was once a fourth of the population of Europe, but now it is only one-eighth. In a century or two

Click to return to Table of Contents 150 A RCHIVES

Paris will be dead, like ancient Athens and ancient Rome, and we shall have fallen to the rank that Greece now occupies. Paris seems determined to die.’ But Santerre protested: ‘No, no; Paris simply wishes to remain stationary, and it wishes this precisely because it is the most intelligent, most highly civilised city in the world. The more nations advance in civilisation the smaller becomes their birth-rate. We are simply giving the world an example of high culture, superior intelligence, and other nations will certainly follow that example when in turn they also attain to our state of perfection. There are signs of this already on every side.’ ‘Quite so!’ exclaimed Séguin, backing up his friend. ‘The phenomenon is general; all the nations show the same symptoms, and are decreas- ing in numbers, or will decrease as soon as they become civilised. Japan is affected already, and the same will be the case with China as soon as Eu- rope forces open the door there.’ Mathieu had become grave and attentive since the two society men, seated before him in evening dress, had begun to talk more rationally. The pale, slim, flat virgin, their ideal of feminine beauty, was no longer in question. The history of mankind was passing by. And almost as if communing with himself, he said: ‘So you do not fear the Yellow Peril, that terrible swarming of Asiatic barbarians who, it was said, would at some fatal moment sweep down on our Europe, ravage it, and people it afresh? In past ages, history always began anew in that fashion, by the sudden shifting of oceans, the invasion of fierce rough races coming to endow weakened nations with new blood. And after each such occurrence civilisation flowered afresh, more broadly and freely than ever. How was it that Babylon, Nineveh, and Memphis fell into dust with their populations, who seem to have died on the spot? How is it that Athens and Rome still agonise to-day, unable to spring afresh from their ashes and renew the splendour of their ancient glory? How is it that death has already laid its hand upon Paris, which, whatever her splendour, is but the capital of a France whose virility is weakened? You may argue as you please and say that, like the ancient capitals of the world, Paris is dying of an excess of culture, intelligence, and civilisation; it is none the less a fact that she is approaching death, the turn of the tide which will carry splendour and power to some new nation. Your theory of equilibrium is wrong. Nothing can remain stationary; whatever ceases to grow, decreases and disappears. And if Paris is bent on dying, she will die, and the country with her.’ ‘Well, for my part,’ declared Santerre, resuming the pose of an elegant pessimist, ‘if she wishes to die, I sha’n’t oppose her. In fact, I’m fully determined to help her.’ ‘It is evident that the really honest, sensible course is to check any increase of population,’ added Séguin.

Click to return to Table of Contents E MILE ZOLA 151

But Mathieu, as if he had not heard them, went on: ‘I know Herbert Spencer’s law, and I believe it to be theoretically correct. It is certain that civilisation is a check to fruitfulness, so that one may picture a series of social evolutions conducting now to decrease and now to increase of population, the whole ending in final equilibrium, by the very effect of culture’s victory when the world shall be entirely populated and civilised. But who can foretell what road will be followed, through what disasters and suffer- ings one may have to go? More and more nations may disappear, and others may replace them, and how many thousands of years may not be needed before the final adjustment, compounded of truth, justice, and peace, is arrived at? At the thought of this the mind trembles and hesitates, and the heart contracts with a pang.’ Deep silence fell whilst he thus remained disturbed, shaken in his faith in the good powers of life, and at a loss as to who was right—he or those two men so languidly stretched out before him. [pp. 35–37]

. . .

[Mathieu’s reflections in maturity:] [I]n presence of those jesters, Beauchêne and Séguin, quite a flood of words rose to Mathieu’s lips. He would have liked to answer them; he would have liked to triumph over the mendacious theories which they still dared to assert even in their hour of defeat. To fear that the earth might become over-populated, that excess of life might produce famine, was this not idi- otic? Others only had to do as he had done: create the necessary subsistence each time that a child was born to them. And he would have pointed to Chantebled, his work, and to all the corn growing up under the sun, even as his children grew. They could not be charged with having come to consume the share of others, since each was born with his bread before him. And millions of new beings might follow, for the earth was vast: more than two-thirds of it still remained to be placed under cultivation, and therein lay endless fertility for unlimited humanity. Besides, had not every civilisation, every progress, been due to the impulse of numbers? The im- providence of the poor had alone urged revolutionary multitudes to the conquest of truth, justice, and happiness. And with each succeeding day the human torrent would require more kindliness, more equity, the logical division of wealth by just laws regulating universal labour. If it were true, too, that civilisation was a check to excessive natality, this phenomenon itself might make one hope in final equilibrium in the far-off ages, when the earth should be entirely populated and wise enough to live in a sort of divine immobility. But all this was pure speculation beside the needs of the hour, the nations which must be built up afresh and incessantly

Click to return to Table of Contents

152 A RCHIVES

enlarged, pending the eventual definitive federation of mankind. [pp. 302– 303]

. . .

[One of the Froments’ sons, who emigrated to Africa, has followed his parents’ example, breeding and prospering:] There was now no longer any mere question of increasing a family, of building up the country afresh, of re-peopling France for the struggles of the future, the question was one of the expansion of humanity, of the re-claiming of deserts, of the peopling of the entire earth. After one’s country came the earth; after one’s family, one’s nation, and then mankind. And what an invading flight, what a sudden outlook upon the world’s immen- sity! All the freshness of the oceans, all the perfumes of virgin continents, blended in a mighty gust like a breeze from the offing. Scarcely fifteen hundred million souls are to-day scattered through the few cultivated patches of the globe, and is that not indeed paltry, when the globe, ploughed from end to end, might nourish ten times that number? What narrowness of mind there is in seeking to limit mankind to its present figure, in admitting simply the continuance of exchanges among nations, and of capitals dying where they stand—as Babylon, Nineveh, and Memphis died—whilst other queens of the earth arise, inherit, and flourish amidst fresh forms of civilisation, and this without population ever more increasing! Such a theory is deadly, for nothing remains stationary: whatever ceases to increase decreases and disappears. Life is the rising tide whose waves daily continue the work of creation, and perfect that of awaited happiness, which shall come when the times are accomplished. The flux and reflux of nations are but periods of the forward march: the great centuries of light, which dark ages at times replace, simply mark the phases of that march. Another step forward is ever taken, a little more of the earth is conquered, a little more life is brought into play. The law seems to lie in a double phenomenon: fruitfulness creating civilisation, and civilisation restraining fruitfulness. And equilibrium will come from it all on the day when the earth, being entirely inhabited, cleared, and utilised, shall at last have accomplished its destiny. And the divine dream, the generous utopian thought soars into the heav- ens; families blended into nations, nations blended into mankind, one sole brotherly people making of the world one sole city of peace and truth and justice! Ah! may eternal fruitfulness ever expand, may the seed of humanity be carried over the frontiers, peopling the untilled deserts afar, and increasing mankind, through the coming centuries until dawns the reign of sovereign life, mistress at last both of time and of space! [pp. 410–411]

Click to print article Click to return to Table of Contents

BOOK REVIEWS

Freedom and Welfare: A Review Essay on Amartya Sen, Development as Freedom*

PAUL STREETEN

It has taken the more-enlightened advanced societies three centuries to achieve the civil, political, and social liberties and rights of human development that they enjoy today. The eighteenth century established civil rights: ranging from freedom of thought, speech, and religion to the rule of law. In the course of the nineteenth century political freedom and participation in the exercise of political power made major strides, as the right to vote was extended to more people. In the twentieth century the welfare state extended human development to the social and economic sphere, by recognizing that minimum standards of education, health, nutrition, wellbeing, and security are basic to the civilized life, as well as to the exercise of the civil and political attributes of citizenship. These battles had not been won easily or without resistance. Reactionary counterattacks and setbacks have followed each progressive thrust. The struggle for civil liberty was opposed, after the French Revolu- tion, by those fearful that it could lead only to tyranny; the fight for political participation by those who feared that it would bring about enslave- ment to the masses. We are now witnessing one of these counterattacks on the economic liberties of the welfare state, and on some fronts partial retreat as a result. The argument again is that the opposite of the intended results is achieved. Just as civil liberty was said to lead to tyranny, and political liberty to slavery, so compassionate concern for the poor, it is now

*New York: Alfred A. Knopf, 1999. xvi + 366 p. $27.50.

POPULATION AND DEVELOPMENT REVIEW 26(1) (MARCH 2000) 153

Click to print article Click to return to Table of Contents 154 PDR 26(1) B OOK REVIEWS said, can lead only to their pauperization. It is not difficult to show that in the war on poverty it is not inevitable that poverty should win, nor is it difficult to demonstrate that well-designed public expenditure programs can contribute substantially to improving the human condition.1 In addressing these and related topics, Amartya Sen has presented us with another brilliant, elegant, and stimulating book. It is eloquently written and forcefully argued. It combines vision, lucidity, and wit. It is a joy to read and reread.2 Sen sees expansion of freedom both as the primary end and as the principal means of development. He illuminates the connection that links different types of freedom with one another. Political freedom promotes economic security. Social opportunities facilitate economic participation. Economic facilities generate resources for social facilities. Different types of freedom strengthen each other. He views freedom as involving both processes and opportunities. Deprivation of freedom can arise, for example, by violating someone’s voting rights and by lack of adequate food. He sees the following much-debated question as fundamentally misdirected: “Do de- mocracy and basic political and civil rights help to promote the process of development?” And his answer is: “Rather, the emergence and consolida- tion of these rights can be seen as being constitutive of the process of development” (pp. 287–288; emphasis in original). In this sense, China cannot be said to have enjoyed (properly defined) full development. In a chapter on “Culture and human rights” Sen defends human rights against three critiques. There is, first, the criticism that rights require a legal authority and that there cannot be pre-legal rights (he calls it the legitimacy critique); second, the view that unless there is a specific locus of a correlated obligation there cannot be a right (the coherence critique); and, third, the skepticism that rights cannot be universal but must vary with different cultures (the cultural critique). In defense against the second criticism he writes: “It may of course be the case that rights, thus formulated, sometimes end up unfulfilled. But it is surely possible for us to distinguish between a right that a person has which has not been fulfilled and a right that the person does not have” (pp. 230– 231). This is true for the right not to be tortured or assaulted, enslaved or murdered, not to be imprisoned without due process of law, the right to equality under the law, the right to freedom of speech and assembly, and so forth. But what if the nonfulfillment of the right is well-nigh universal and inevitable because it calls on resources that the society does not have? This is the case at low levels of development for the claim that there are human rights to the best medical treatment, the best education, the best provision for injury, sickness, old age, and widowhood, free choice of employment, minimum wages, collective bargaining, and more. Yet these are the economic, social, and cultural rights embodied in the United Nations

Click to return to Table of Contents B OOK REVIEWS PDR 26(1) 155

Declaration of Human Rights and the International Covenant on Economic, Social and Cultural Rights. It is obvious that implementation of these rights is impossible, or, if partially possible, these objectives can be achieved only at the expense of other rights (or, better, obligations; see below) or can be achieved for a few at the expense of other members of the community. Would it then not be better to distinguish between “rights,” about which there is something absolute or categorical, and which therefore have to be implemented fully and immediately, and “aspirations,” which are ideals that we hope to attain eventually? The difference is not just one of the choice of words; it has important implications for policy. The language of rights can be and has been abused to promote the advantages of privileged groups. For example, the right to collective bargaining has been used in some developing countries to reinforce the privileges of a small, urban labor aristocracy. Sen continues, “Ultimately, the ethical assertion of a right goes beyond the value of the corresponding freedom only to the extent that some demands are placed on others that they should try to help. While we may be able to manage well enough with the language of freedom rather than of rights (indeed it is the language of freedom that I have been mainly invoking in Development as Freedom), there may sometimes be a good case for suggesting—or demanding—that others help the person to achieve the freedom in question. The language of rights can supplement that of freedom” (p. 231). But the language of obligations or duties or responsibilities is sufficient for these demands, without invoking the language of rights. Obliga- tions, duties, or responsibilities are possible without rights. One example is our obligations to future generations. Nonexisting people cannot have rights, but we can have duties toward them, such as the obligation to bequeath to them the resources, conditions, and institutions for a decent society.3 An- other example is our duty to contribute to charities, where there are no corresponding rights of claimants. And if obligations to ourselves are accompanied by rights of ourselves, this may involve excessively split personalities. John Toye called Amartya Sen, rightly, the distinguished distinguisher. One of Sen’s special gifts is to draw subtle distinctions between concepts that the rest of us fail to see. But in this rare exception, he is a lumper, failing to separate rights from duties. In a low-income country there is an obligation, not to provide the full paraphernalia of a modern welfare state, but for everyone to enjoy a fair share of the community’s resources. In particular, the avoidance of starvation or famine, where enough food is available, is an important objective that is met in all democracies. Sen’s analysis of the link between democratic freedom and the avoidance of famine in poor countries, as a result of the free flow of information and the accountability of politicians, is one of his important findings. A whole chapter is devoted to this issue.

Click to return to Table of Contents 156 PDR 26(1) B OOK REVIEWS

Sen’s reply to the third or cultural critique, and in particular his discussion of “Asian values” that emphasize discipline and obedience rather than freedom, is completely convincing. He asserts the universal value of freedom and tolerance, nay, the rejoicing over and celebration of our cultural differences. It is sad to see that Ashok Singhal of the World Hindu Council has said that the award of the Nobel Prize in economics to Amartya Sen was evidence of a “Christian conspiracy to propagate their religion and wipe out Hinduism from this country.”4 Another couple of concepts that Sen, the distinguished distinguisher, does not always seem to distinguish sufficiently sharply are equity and equality. The two are used almost interchangeably (e.g., p. 119). But inequality is different from inequity. Inequality means that I have more and you have less. Inequity means that I have more because you have less or you have less because I have more. It implies exploitation and an unfair division of benefits. Inequity is unjustified and unjustifiable inequality. We are all against inequitable distribution but many are for inequality, especially if it is deserved or merited, or if it meets differential needs, or if it is a necessary condition for improving the lot of the poor. Inequity also implies what is sometimes called unjustified horizontal inequality. It implies unfair treatment. It would be inequitable to tax every third person coming out of the Ritz, or only red-headed rich people, even if they were all very rich. It would be unfair, and to be treated fairly is a fundamental human right. Not so to enjoy equality, quite apart from the difficulty of deciding equality of what—of receiving a claim on scarce goods and services? of opportunity, or of achievement? of access to the law, of free expression of opinion, or in the sight of God? In a chapter on “Population, food and freedom” Sen shows that not Malthusian pessimism but Malthusian optimism can be the cause of disaster. When policymakers think that the food problem is solved when food per head grows faster than population, while forgetting the need for food entitlements, famines can occur. “A misconceived theory can kill, and the Malthusian perspective of food-to-population ratio has much blood on its hands” (p. 209). Similarly, the simplistic notion that all that is needed is to make the land more fertile and women less fertile misses the point. Food production has increased faster than did population in spite of falling food prices. Expansion of education, especially for girls, better health services, greater independence of women and more job opportunities for them, and reduction of mortality rates are among the prerequisites for solving the food/ population problem. High fertility rates are also adverse to the quality of life of young women. Women’s empowerment raises their ability to influence family decisions and hence to reduce birth rates. Sen once wrote that the tendency to see in population growth an explanation of every calamity that afflicts poor people is now fairly well es-

Click to return to Table of Contents B OOK REVIEWS PDR 26(1) 157 tablished in some circles, and the message that gets transmitted constantly is the opposite of the one on the old picture postcard: “Wish you weren’t here.” But, in spite of thinking that the magnitude of the population problem is often exaggerated, Sen remains concerned about world population growth. There follows in this chapter an interesting discussion of the ac- ceptability and efficacy of coercion. Social and economic development, particularly the empowerment of women, is the best way to reduce fertility: women’s education, their involvement in the labor force (jobs outside the home), the opportunity for women to earn an independent income, women’s property rights, and improvement in their status in the culture. (These are, of course, also desirable in their own right and because they reduce sex inequality.) Sen then compares these influences with coercive policies, such as the one-child family policy, tried in China. He shows that Kerala’s birth rate is actually lower than China’s, and that this has been achieved without any coercion by the government. Both Kerala and Tamil Nadu in fact had no slower a fall of fertility rates than China. An effective and humane policy of fertility decline demands more freedom, not less. This is important for other reasons as well. There are, of course, additional arguments against coercion. Quite apart from the violation of the autonomy of individuals and their reproductive rights, it was shown in India to be counter-productive. “Aside from having little immediate impact on fertility rates, the coercive measures of the emergency period introduced in some regions in India were, in fact, followed by a long period of stagnation in the birthrate . . .” (p. 224; emphasis in original). Such coercive measures can also have an adverse effect on infant mortality, especially that of girls. Adam Smith is Sen’s mentor. Sen writes that this book has “a strongly ‘Smithian’ character” (p. 255) and rightly complains that “while some men are born small and some achieve smallness, Adam Smith has much smallness thrust upon him” (p. 272). He shows that Smith did not believe that self-interest dominates men’s actions but rather that they had the capacity for a sense of justice and for a conception of the good. Sen opposes both the “high-minded sentimentality” of assuming that everyone is intensely moral and the “low-minded sentimentality” that we are only acting for personal advantage. He also shows convincingly that Smith was not an unqualified advocate of free markets and knew when to restrain them. And so does Sen. While welcoming the Indian liberalization in 1991, he regrets that it was not accompanied by remedying the defects in elementary education (half of India’s adult population and two-thirds of its adult women are illiterate) and expanding other social opportunities. He invokes Adam Smith in support. In a market-price-based evaluation, everything except commodity holdings gets a zero direct weight: mortality, morbidity, education, liberties, and recognized rights.

Click to return to Table of Contents 158 PDR 26(1) B OOK REVIEWS

The World Bank, which Sen describes as not invariably his favorite organization (p. xiii), is prone to use military metaphors: strategies, task forces, tactics, campaigns, war on hunger, logistics, attrition, mobilization, frontal assault, overkill, delivery systems. (Should “mission” perhaps be replaced by “sortie”?) Targeting is also a military metaphor. It is misapplied for many reasons; its military connotations are objectionable (is it suggested that certain groups are not only got at, but actually shot at; is self-targeting suicide?) and it implies passivity and excludes the active agency of people. This, like other forms of freedom, is important both as an end of, and as means to, human development. Sen emphasizes repeatedly the need for participation in a freedom- based and freedom-oriented approach to development. For instance, when it comes to a choice between preserving a traditional culture and acquiring more material goods, he rightly says that this should be decided by the people, not by mullahs or other religious leaders or by experts or political authorities. Oscar Wilde once said that “the trouble with socialism is that it takes too many evenings.” The same can be said about participation. Town or village meetings take time, and some may prefer to do other things, including nothing, and regard attendance at them as an infringement on their freedom. In many low-income countries women in particular have their time fully occupied with fetching water and firewood, preparing meals, looking after their children and their plots, and other tasks. The objective then is first to free these women’s time (for example, by inventing a cheap cook- ing stove) so that they can devote themselves to participation in the affairs of their community. Sen has calculated that many more than 100 million women worldwide are “missing.” Normally, and given the same care, women outnumber men. In some areas of the world, however, such as South Asia, China, West Asia, and North Africa, men outnumber women. This is the result of the neglect of female health and nutrition and of female infants. It is an illus- tration of the many dimensions of inequality that have been neglected by economists in favor of income poverty and income inequality. Other variables include unemployment, ill health, lack of education, lack of opportunities, and social exclusion. Gender inequalities, for example, cannot be analyzed in terms of income inequalities since incomes earned are presumed to be shared by all family members.5 Sen discusses different philosophical positions with great clarity and fairness. Passages of abstract reasoning are spiced with wonderful tales. But when he says as an argument against utilitarianism that “we do not necessarily want to be happy slaves or delirious vassals” (p. 62), does he not give the game away? If we do not want to be happy slaves, the wanted freedom contributes to our want-fulfillment. Perhaps he should have said “we ought not to want to be happy slaves.”

Click to return to Table of Contents B OOK REVIEWS PDR 26(1) 159

Sen argues that human development cannot be judged only by end- states, and that the freedom to choose between different options is an important component of wellbeing. A given commodity bundle has a different significance to the consumer according to whether she has other options, though she does not exercise them, or whether that same bundle is the only one available. There is a difference between a starving pauper, a fast- ing monk, and Gandhi on a hunger strike, a difference not reflected in the low calorie intake of all three. Only the starving pauper lacks capability. Sen’s capabilities, though more important than achievements, cannot be observed or readily measured, while achievements can. If failure of achievement is voluntary, it is acceptable. But one could argue that it is better to separate freedom of choice, and to look at poverty or deprivation in terms of observable achievements. In this sense, all three individuals are deprived, though clearly not equally poor. Sen puts achievement and freedom of choice together under “capability.” Happiness, as experienced by the individual, is not what human development can aim at or is mainly about. Not only can the government not deliver happiness;6 people may be miserably poor and yet be contented. Indian women much less frequently report being ill than Indian men. On the other hand, in re-surveying villages in Gujarat after 20 years, Jodha (1988) found that the households whose real income per head had declined by more than 5 percent felt that they were, on average, better off on most of their criteria of wellbeing.7 Besides income and consumption, they were concerned with independence (especially from patrons), mobility, security, and self-respect. This should serve as a warning against attempting to simplify measures of poverty into single indicators, especially those relying on income and consumption, and against relying solely on quantitative indicators. In a chapter on “Women’s agency and social change” Sen discusses the notion of “cooperative conflict,” noting that women and men have both congruent and conflicting interests. One often hears the argument that to leave the talents and contributions of half of humanity underused amounts to enormous waste, the elimination of which would benefit both men and women. On the other hand, there are privileges and advantages enjoyed by men (not only within the family) whose surrender for the sake of equality, empowerment, and the independent agency of women would be contrary to the selfish interests of men. An interesting finding reported from India is the statistically significant relation between the female–male ratio in the population and the scarcity of violent crimes. Different explanations of the causal processes have been offered. Sen writes: “For efficient provision of public goods, not only do we have to consider the possibility of state action and social provisioning, we also have to examine the part that can be played by the development of social values . . . ” (p. 269). But public goods are not necessarily state-pro-

Click to return to Table of Contents 160 PDR 26(1) B OOK REVIEWS duced. The distinction between public and private goods is not useful in deciding what should be done by the state and what by the private sector. Many public goods are most efficiently supplied by small groups of people (e.g., farmers who exercise common control over common grazing rights) or by members of the private sector under contract to the state, both acting in ways quite apart from the role of social responsibility that Sen rightly emphasizes. In the penultimate section of the final chapter Sen discusses the relations between two approaches: that of human capital and that of human capability. He uses education as an example: “If education makes a person more efficient in commodity production, then this is clearly an enhance- ment of human capital. This can add to the value of production in the economy and also to the income of the person who has been educated. But even with the same level of income, a person may benefit from education— in reading, communicating, arguing, in being able to choose in a more informed way, in being taken more seriously by others and so on” (pp. 293–294). At first blush, there indeed appears to be a unity of interest between the human capital focus (the human capital proponents take an even narrower view than the human resource proponents) and the concentration on human capabilities. Although the motives of the two approaches are different, both have the same cause at heart, and their advocates should embrace each other, for example, when it comes to promoting education. Means are means toward ends, which presumably are the same ends in both camps. This harmony of interests is reinforced by the widespread notion that “all good things go together.”8 This unity of interests would exist if there were rigid links between economic production (as measured by income per head) and capability development (reflected by human indicators such as life expectancy or literacy, or by achievements such as self-respect, not easily measured). But, as Sen points out repeatedly in various contexts, these two sets of indicators are not very closely related. A lot of these discrepancies are, of course, the result of different income distributions. A high average income per head can conceal great inequalities. But there are other reasons too. The content of and access to social services vary, particularly in poor countries, and different ratios devoted to primary education and basic health services are also important. Nor is there agreement on policies between the human capital and the human capabilities proponents. Means have a way of acquiring the characteristics of ends to which those who sponsored the initial ends do not sub- scribe.9 The following seven points of difference are not based on the unity of logic but are comparable to the unity of psychological traits in different personalities. First, the capabilities advocates are concerned also with the unpro- ductive, the lame ducks, the unemployables: the old, infirm, disabled, chroni-

Click to return to Table of Contents B OOK REVIEWS PDR 26(1) 161 cally sick. As Sen points out, these people suffer from a double disadvan- tage: they face greater difficulties in earning income and in converting income into wellbeing. There may be a bonus for the community in looking after them, if not for higher productivity, at least for reduced reproductivity. If parents know that the community will care for them if they become disabled or infirm, an important cause of the desire for large families and particularly for many sons disappears. This will also reduce discrimination against females. But these benefits are incidental. Second, the ability to convert means into ends, resources into worthwhile or satisfying activities, varies widely between different people. Even such a basic good as food meets the needs of nutrition differently according to the rate of metabolism, the sex, age, and workload of the individual, the climate, whether an individual female is pregnant or lactating, and whether she is ill, has parasites in her stomach, or needs the food for other uses than her own consumption, such as entertainment or ceremonies. Third, the human capital approach lends itself to treating individuals as passive “targets”; the approach that sees them as ends regards them as active, engaged agents. Adherents of the latter approach would appeal more to people’s full, active participation. Fourth, the content of the educational curriculum (and health program) of the human capital and the human capabilities proponents is different, as is implied in Sen’s example. Proponents of human capital will aim at general education and learning for its own sake, and for understanding the world, while proponents of human capability will be more vocational, aiming at training (including training for flexibility) rather than education. Fifth, their views on the role of women will differ, human capital proponents advocating access to the labor market and human capability advocates also stressing the nurturing functions: breastfeeding, preparing nutri- tious meals, and looking after the family. This raises the question whether there is a separate women’s sphere or whether freedom and autonomy are to be aimed at. Those who advocate women’s freedom on grounds of efficiency will welcome the benefits for men also, because they see men and women as engaged in a positive-sum game. On the other hand, those who are concerned with women’s rights as an end will advocate policies that reduce the benefits to men and involve sacrifices by them. Sixth, their sectoral priorities will be different, housing being least connected with raising production, education most connected, with nutrition and health somewhere in the middle. Seventh, the constituencies to which they appeal for applause and support will also be different. The human capital advocates appeal to mainstream economists, bankers, including the World Bank, and technocrats;

Click to return to Table of Contents

162 PDR 26(1) B OOK REVIEWS

the human capability advocates to the churches, NGOs, action groups, ide- alists, and moral and political philosophers. The approach that sees nutrition, education, and health as ends in themselves rather than primarily as means to higher productivity will argue for projects and programs that enhance these ends, even when conventionally measured rates of return on these investments turn out to be zero. It amounts to standing the conventional approach on its head, or rather back on its feet again.10 The occasional minor disagreements with Amartya Sen’s argument that I have alluded to in this essay only add to my pleasure and admiration for his mind and his work.

Notes

1 For an elegant elaboration of this argu- mists, who are the trustees, not of civilization, ment see Hirschman (1991). but of the possibility of civilization.” 2 I have benefited from very helpful com- 7 Jodha (1988). ments by Wilfred Beckerman, Sudhir Anand, 8 Thus Behrman (1993) writes about bet- and Paul Demeny. ter nutrition among poorer members of soci- 3 See Beckerman (1999: 71–92). ety: “That productivity and equity concerns are in harmony is an important plus.” 4 Sinha (1998). 9 Some of the differences are attributable 5 In Russia male life expectancy is less to the fact that one group attaches end values than female life expectancy by 12.2 years. to what, for the other group, are pure means. There are many missing men! The capability advocates attach such value to 6 Keynes proposed the toast to the Royal participation. Economic Society: “to economics and econo- 10 See Streeten (1994).

References

Beckerman, Wilfred. 1999. “Sustainable development and our obligations to future generations,” in Andrew Dobson (ed.), Fairness and Futurity: Essays on Environmental Sustainability and Social Justice. Oxford: Oxford University Press. Behrman, Jere. 1993. “The economic rationale for investing in nutrition in developing countries,” World Development 21(11): 1749. Hirschman, Albert O. 1991. The Rhetoric of Reaction: Perversity, Futility, Jeopardy. Cambridge, MA: The Belknap Press of Harvard University Press. Jodha, N. S. 1988. “Poverty debate in India: A minority view,” Economic and Political Weekly, Special Number, November. Sinha, Rajesh. 1998. “Sen, Teresa Nobel Prize, a Christian plot: Singhal,” Indian Express, 28 De- cember. Streeten, Paul. 1994. “Human development: Means and ends,” The American Economic Review, 84(2): 232–237.

Click to print book review Click to return to Table of Contents

B OOK REVIEWS PDR 26(1) 163

Rocky Mountain Visions: A Review Essay*

VACLAV SMIL

Why should I review a book whose American edition sold out before its publication date, a book called by its dust-jacket boosters a “conceptual mile- stone,” “the design manual for the 21st century,” the wisdom of “three of the world’s best brains,” and “the bible” for the next industrial revolution?1 I am not a keen book reviewer but such an astonishing lack of historical perspective and such marketing hubris devoid of any sense of proportion, emblem- atic as they may be of the times we live in, impel me to pen a modest essay recalling useful lessons from the past and offering cautionary observations about the present and the future. The book introduces four central strategies of natural, as opposed to conventional, capitalism: radically increased resource productivity; biomimicry (reducing waste through recycling); service and flow economy (a shift from purchasing goods to renting services); and investing in natural capital, that is in ecosystems, in order not just to ensure the supplies of valuable goods but to sustain the provision of irreplaceable environmental services (“capitalism as if living systems mattered”). What is so stunning about these max- ims? What in all of this is so paradigmatically pathbreaking, so breathtak- ing, that a comparison with Adam Smith came immediately to the mind of another reviewer; what is so irresistibly biblical about these visions that “come not a moment too soon” to tell us all how to live? Those readers familiar with writings by Amory and L. Hunter Lovins will find very little here that they have not seen before. Although Paul Hawken, a businessman and the author of the “classic book” (the dust jacket, again) and the eponymous PBS series (Growing a Business), is listed as the first author, the book is unmistakeably the latest entry in the line of the Lovinses’ grand prescriptions for paradisiacal futures. Actually, large parts of the book are simply recycled, with only slight alteration, from their re-

*Review of Paul Hawken, Amory Lovins, and L. Hunter Lovins, Natural Capitalism: The Next Industrial Revolution. Boston: Little, Brown and Company, 1999. xix + 396 p. $26.95.

Click to print book review Click to return to Table of Contents 164 PDR 26(1) B OOK REVIEWS cent writings, particularly from Factor Four, which the co-CEOs of the Rocky Mountain Institute (RMI) coauthored with Ernst von Weizsäcker.2 As a witness to Amory Lovins’s transformations during the past three decades—from zealous critic of nuclear energy to militant advocate of “soft- path” energies to his most recent incarnation as a guru of corporate consulting coming to save the planet through natural capitalism—I have felt like a tourist trudging along a particularly well-worn path. That is because all the outward changes have been much less important than the underlying constants in his writings. He always writes well, and he steadfastedly advocates smarter techniques, higher efficiency, cleaner environments, and less pollution: who could disagree? In addition, there is the money—and nothing sells better to a population addicted to double-digit increases of stock market indexes than the promise to “make a lot of money on the side” while doing socially and environmentally admirable deeds.3 My attitude toward these visions and promises has not changed either: it has always been a mixture of strong approval on one hand and re- current exasperation and resolute disagreement on the other—reactions potentiated by this book. Both the Lovinses and Hawken say many things with which I wholeheartedly agree; and parts of their and my writings and speeches, although informed by different backgrounds, are readily inter- changeable. I share their calls for technical rationality, higher efficiency, lower environmental impacts, and more considerate farming—but I cannot foresee such easy walks and such stunning rewards in so short a time as they claim or imply. My quarrel is not with their goals but with the excessive promises, repeated exaggerations, wishful thinking, and righteous insistence that theirs is the only enlightened way. Our fundamental differences lie in the awareness of history and in the readiness to prescribe grand solutions. I am constantly aware of the presence and importance of the past (natural for someone coming from Kafka’s city), and hence I marvel how even the most “revolutionary” ideas grow so unmistakeably from tangled thickets of old thought, and how little our individual contributions matter when one looks at incremental and cumulative progress in science, engineering, and human wellbeing. And having lived for 27 years in a vassal state of the Soviet Empire has made me dis- trustful of any normative solutions revealed in point-by-point instruction sheets by “best brains.” (Comrade Stalin, as I well recall from my grade- school days, had one too!) De omnibus dubitandum: I may have strong preferences based on what I think is fairly good evidence—but inevitable uncertainties and ignorance make me anxious about what might go wrong and hence cautious about the course I would advocate for others to follow. I know that there are no actions without unintended consequences, that the propensity for catastro- phe is conserved in every civilization, and that both as individuals and as

Click to return to Table of Contents B OOK REVIEWS PDR 26(1) 165 nations we have immense capacity for self-delusion, denial, and irrational choices in the name of rationality. And I much prefer gradual workings of evolutionary change to sweeping calls for new revolutions, no matter how sleekly merchandised they may be.4 In contrast, the Lovinses have never shown any qualms about pre- scribing for all of humanity. Theirs is a linear world where, after the break with the unbearably irrational past, everything will get better if only we march under their supremely rational techno-banner, which is now not merely soft and green but has a “lot of money” on the side. Just read the scenario opening the book:

[C]ities have become peaceful and serene. . . . Living standards for all people have dramatically improved, particularly for the poor and those in developing countries. Involuntary unemployment no longer exists, and income taxes have largely been eliminated. . . . there are few if any active landfills; worldwide forest cover is increasing. . . . The frayed social nets of Western countries have been repaired. . . . A progressive and active union movement has taken the lead to work with business, environmentalists, and government.... Is this the vision of a utopia? In fact, the changes described here could come about in the decades to come. . . . (pp. 1–2)

Here we have no cautious reformers: our trio has perfect solutions for everything, and the cures are nearly instant. After millennia of miseries and troubles, misguided humanity has finally found its true saviors who could stick its dumb muzzle into the trough of perpetual happiness in a matter of mere decades. Of course, there is nothing new about the idea of social salvation through a technical fix, only the ingredients keep changing: in this case fuel cells and Hypercars are to become the guarantors of human con- tentment. Nor is it the first time energy conversions have been singled out to play such an omni-catalytic role: during the 1950s it was to be the too- cheap-to-be-metered nuclear electricity, and after the OPEC oil shocks of the 1970s we were promised rapid energy self-sufficiency and untold social rewards through wind generators on rooftops and biogas digesters in back- yards.5 The precept of natural capitalism also resembles those earlier grand schemes because of its supposed infallibility: in the current American parlance, it has no downside. No Murphy’s law stalks the trio’s super-efficient nirvana, no unintended consequences will follow if the world does as they say. While many doubting, fallible agnostics wait for the accumulated evidence before venturing a verdict on complex affairs, the authors are true, unwavering believers-with-a-mission who know they are right. Often they are—often they are not, and, given the deplorable state of scientific literacy, too many people could be easily misled by their facile affirmations.

Click to return to Table of Contents 166 PDR 26(1) B OOK REVIEWS

And so, leaving their opening fairy tale aside, when the authors state that “One of the keys to the most beneficial employment of people, money and the environment is radical increases in resource productivity” (p. 9), I can only respond: of course, well said—and I could have yanked out scores of such citations and displayed them approvingly. But when the book claims, citing the declaration of the Factor Ten Club, that “within one generation, nations can achieve a ten-fold increase in the efficiency with which they use energy, natural resources and other materials” (p. 11),6 then I must respond: what a ridiculous exaggeration. Every scientist or engineer should be a professional skeptic and doubt unprecedented claims: the authors should have questioned the Factor Ten statement—not called it “prophetic.” If it were to happen, the factor ten improvement in a single generation would mean a magical disappearance of population size as a factor in economic development and in global environmental degradation. The only interesting matter would then be how to organize this miracle. Five billion people (more than four-fifths of humanity) in Asia, Africa, and Latin America now claim only about one-third of the world’s resources—but wringing ten times as many useful services out of their current fuel, electricity, mineral, wood, and water consumption would make them developed by any definition: China’s standard of living would be lifted above that of today’s Japan; India would be better off than Argentina. North America and the European Union, content with maintaining today’s high standard of living, would have no need for 90 percent of their current resource use, a shift resulting in plummeting global commodity prices. Poor countries could then snap up these give-away commodities (capacity to produce them will not disappear) and, using them with ten times higher efficiency than they do today, they could support another 7 billion people enjoying a high standard of living—and all of this with no increase in the global use of energy and materials! Who would care if the global population total were to grow by another 2 or 3 billion before stabilizing? The United States would consume less energy in the year 2020 than India uses today.7 China, with nearly 1.5 billion people in 2020, could reap the world’s largest food harvest by applying less fertilizer than does France today.8 Conversely, with a constant rate of energy and material consumption, a country would have at its disposal ten times as much light, heating, cooling, and transport capacity, or it could grow ten times as much food with the same amount of fertilizers or enclose ten times as much living space with the same amount of materials.9 The authors must know that none of this can happen in one generation. But I am not sure whether they are as ignorant of the precedents of their “four central strategies of natural capitalism” as their ahistorically exuberant “introduction” of these great principles suggests—or whether they are just resorting to a marketing hyper- bole that seems to be the norm these days.

Click to return to Table of Contents B OOK REVIEWS PDR 26(1) 167

In any case, a few reminders are in order. Doing more with less is an admonition that resonates through millennia of both Western and Eastern thought. Medieval monastic orders were excellent embodiments of nearly all of the tenets of natural capitalism. All energies were soft (monks’ muscles, donkeys, oxen, here and there a small waterwheel); production of all food and often also of all cloth and all wooden and metallic necessities or adorn- ments of life was local, artisanal, and utterly decentralized. Nothing was packaged, everything was recycled, and monks owned nothing; they were merely renting the order’s bleak cells, hard beds, and coarse cloth in return for copying missals, singing a cappella, curdling cheese, and weeding vegetables.10 And if they were Franciscans, then, following the precepts of their sainted founder from Assisi, they were the protectors of that most valuable of all natural capital, the biodiversity bequeathed to us by evolution: “brother wolf,” ilex groves, soaring falcons, wild flowers, and “little brethren the birds” were as dear to their hearts as were lepers and beggars.11 Ancient Romans would not be surprised: where human mores are concerned, nihil novum sub sole. More recently—once the great machine of Western modernization began spinning its innovations—generations of engineers spent their professional lives wringing every bit of efficiency from their contraptions and mim- icking nature in their devices and structures.12 Entire nations, guided by the invisible hand rather than by any bibles, have been relentlessly following the path of reduced waste and higher productivity. Given the ubiqui- tous waste that accompanies common fuel and electricity conversions, the quest for higher energy efficiency has been the very essence of secular technical progress. Quantitative historical perspectives are sorely needed here, and I will offer just a few examples concerning energy and materials. The best charcoal-fueled late-nineteenth-century iron blast furnaces needed only about one-tenth as much fuel as their medieval counterparts: a factor ten efficiency gain, but it took about 500 years.13 Today’s best commercial ammonia synthesis needs less than one-quarter of the energy required by BASF’s first Ludwigshafen plant in 1913: factor four in less than a century.14 While Edison’s carbonized paper loops converted a mere 0.2 percent of electricity into light in his first bulbs, fluorescent tubes introduced in the late 1930s were about 7 percent efficient: factor 35 in just two generations.15 While energy efficiencies of animate prime movers and biomass combustion improved only slowly during antiquity and throughout the Middle Ages, impressive examples of cutting down on the use of materials and doing more with less go back more than two millennia. The Parthenon’s ar- chitraves (447 BC) weigh about 2.7 t/m3 and their free span was less than 2.5 m; the Pantheon’s bricks (27 BC) and light concrete average just 1.7 t/m3

Click to return to Table of Contents 168 PDR 26(1) B OOK REVIEWS but the building’s coffered vault spans almost 44 m: a nearly 30-fold increase in mass/space spanning factor in about 400 years.16 Again, the modern systematized quest for better performance speeded up many such rates. For example, in 1750 the iron and wood of Newcomen’s steam engines weighed well above 500 grams per watt of their capacity; by 1900 iron, steel, and aluminum making up the best internal combustion engines weighed just 1 g/W: a mass decline factor of 600 in 150 years.17 Declines in the use of materials have been historically so large that they can be easily seen even in aggregate national statistics: when measured as mass per unit of GDP, the use of water, timber, steel, or copper in the US economy has been dropping steadily since before the Lovinses were born.18 And what has been no less noteworthy about these trends is that the rates of decline often appear to be entirely autonomous. Perhaps most remarkably, the energy intensity of the US economy—how much energy is used per inflation-adjusted dollar of national product—fell at exactly the same compounded rate during the crisis-fraught 1930s as it did during the e.com-bewitched 1990s (about –1.2 percent per year).19 A book filled with such examples could be as long as Natural Capital- ism because that potent mixture of systematized curiosity- and profit-motivated innovation has been driving human advancement particularly rapidly for the past 200–300 years. Cumulative technical and managerial innovations, incremental learning, and diffusion brought most of these improvements regardless of prevailing commodity prices, government interventions, or the presence or absence of consulting gurus—or of any epochal “introductions” by best brains. A “revolutionary strategy” of “radically increased resource productivity” formulated by the trio is in fact a key evolutionary trend of human history. That some gains within this multifaceted process are now realized faster than ever is only as expected: huge accumu- lating investment in our scientific understanding, instant communication of new findings, and globally fungible capital had better make some difference! But what about the premise that traditional capitalism does not value the largest stock of the Earth’s capital locked in ecosystems and in numerous services they provide? Should not this omission be fixed, as the book advocates, by taking a proper account of these costs, so that production will eventually generate little or no waste and investment will not merely sustain but also expand those precious stores of natural capital? Once again, labels may be new—common use of the term environmental services dates only to the 1970s—but there is nothing new about advocating new accounting schemes for a better valuation of resources or about reinvesting repeatedly in natural capital. Supposedly superior accounting systems proposed during the twentieth century include Frederick Soddy’s new economics; the US Technocracy of the 1930s, which was to run the world by an energy, rather than a money, standard; and Howard Odum’s

Click to return to Table of Contents B OOK REVIEWS PDR 26(1) 169

EMERGY.20 Much like natural capitalism, the first two schemes were to bring great social transformations by doing away with poverty and unemployment and by spreading affluence, while Odum’s questionable concept of accounting for every good and service in terms of embodied solar energy was explicitly designed to save the biosphere. And centuries, even millennia, of continuous production from crop fields would have been impossible without appropriate reinvestment into natural capital: every farmer who ever rotated crops or plowed straw and manure into soil to replenish its organic content and to improve its tilth and water-holding capacity has been such an investor.21 And while tropical deforestation gets the headlines, extensive increase of forest cover in some temperate regions of the world, including eastern North America, has been curiously unreported: European increases have been particularly rapid.22 Since the 1950s some industrial activities have internalized their environmental burdens to a surprisingly high degree. US coal, one of the Lovinses’ great villains, is a good example. Its price now includes the producers’ contributions to a fund for black lung disease as well as the cost of land reclamation after the mining is over; all large power plants have elec- trostatic precipitators (eliminating more than 99.5 percent of particulate matter), most of them have much costlier flue gas desulfurization (SO2 emissions have been declining since the 1970s, and those from large stations shrank by another 30 percent during the 1990s); and reductions of NOx emissions and further controls of smaller particulates are underway.23 Of course I, too, would like to see thousands more environmental burdens appropriately internalized, but the task is not one of simple bureau- cratic fiats: these are fascinating but exceedingly tricky matters that have no easy resolution. If, obviously, “there are some resources that no amount of money can buy” (p. 147)—and if these include nearly all of the essential functions evolved during the nearly 4 billion years of the biospheric complexi- fication—how are we to value them in order to make businesses focus on sustaining and renewing this invaluable capital? Even much simpler challenges are exceedingly hard. If we are to in- ternalize the true cost of gasoline-run cars (my fervent wish), what price are we to put on additional photochemical smog–induced asthma attacks: $50 or $500 per event, the price of an emergency check-up or of a child’s and parents’ momentary (or even anticipated) anguish?24 Indeed, how do we decide in the first place which shares of new attacks in particular cities are due to the smog and which to the mysteriously global higher incidence of the disease?25 If we decide to include in gasoline’s price the cost of the 1991 Gulf War and the share of the Pentagon’s budget that has been spent since then on aircraft carriers and prepositioned planes and troops in Saudi Arabia and Kuwait, we must remember that the American ships, jets, and troops were, and are, in the Gulf also because Iraq was, and is, bent on

Click to return to Table of Contents 170 PDR 26(1) B OOK REVIEWS developing nuclear weapons, and that most of the Middle Eastern oil is actually bought by Europe and Japan; how do we split all that? And if we want to burden the combustion of fossil fuels with the future possible cost of fossil fuel–induced global warming, we would face a very different discounting choice with a slow warming, say an average of 1.5 oC over a century, than with a rapid one (perhaps in excess of 3 oC in 50 years).26 This list could go on and on—and there is no indisputable science to inform such decisions, no correct way to apportion, or to discount, the costs. Hence, do not look for a biospherically integrated price sheet stuck on the window of a new car for sale in the year 2005. Natural Capitalism skates over such matters. Its prescribed solution is a massive taxation of everything the authors consider a burden on the environment: all electricity nonrenewably generated, all fossil fuels, chlorine, air traffic, all roads, pesticides, synthetic fertilizers, piped-in water, timber, all wastes. “Thermal insulation and superwindows in such a world will have a bigger payout than Microsoft stock. You will be able to make Warren Buffet[t] returns by simple investment in hardware-store technologies” (p. 167). The book says nothing about who will decide what the levels of these myriads of taxes will be, and how one can impose them globally to save the biosphere. If China, the world’s largest CO2 emitter from coal combustion, will not tax its coal why should Japan tax its much cleaner natural gas; if Indo- nesia will burn all of its immense and rich biodiversity, then how futile to protect a tiny patch of poor Mediterranean scrubland. And what about those people who cannot do without such “evils” as piped water and synthetic fertilizers? The book advocates gathering rain water from roofs to giant vats (as a certain doctor does in Austin, Texas), and it implies that we can be fed from organic farming. How does one replicate that gathering feat as a tenant on the fifteenth floor of a Hong Kong or Osaka housing estate? And, yes, heavy taxes on nitrogen fertilizers may dramatically reduce their application in Iowa—and overweight Americans may simply have less corn to feed their steer and pigs and hence, perhaps, a less meaty diet and actually be better for it. But without synthetic nitrogen fertilizers China could feed only half its current 1.25 billion people, and while savings in nitrogen applications are both possible and highly desirable, we can never make their use 90 percent efficient. No factor of ten here. Contrary to what the authors imply, biospheric cycles operate with considerable unavoidable inefficiencies. The bacteria-run nitrogen cycle is particularly leaky because of natural nitrifica- tion, leaching, denitrification, and volatilization, and were the Chinese to boost the element’s average uptake efficiency by 50 percent (from almost 50 percent to 75 percent) they would do exceedingly well.27 Natural Capitalism also propagates two standard modes of Lovinsian discourse: toss out a quotable, captivating sentence, deduce an epochal con-

Click to return to Table of Contents B OOK REVIEWS PDR 26(1) 171 sequence from it, and move on to another sound bite; or make a sweeping proposal and declare that it can be turned easily into reality, without acknowledging much evidence to the contrary. Over the years many engineers, scientists, and economists have been exasperated by this peculiar kind of discourse-by-declaration. Two glaring examples from the book suffice. “Some 15 percent of global food is already grown in cities. In China, urban farming in back gardens, on little plots, and on rooftops provides 85- plus percent of urban vegetables—more in Beijing and Shanghai—plus large amounts of meat and treecrops” (p. 200). This is patent nonsense. Growing 15 percent of the world’s food in cities would be equivalent to harvesting about 400 million tons of grain, or almost 20 percent more than the entire US cereal crop.28 Even assuming, wrongly, that all of the world’s vegetables would be grown in suburban fields (obviously not the same as “in cities”), the proportion would be a mere 2 percent.29 Moreover, by saying “already” the authors seem to imply that this preposterously high share might grow, while in reality the periurban yields are increasingly endangered by high levels of ozone generated by photochemical smog.30 “By now, most readers are probably wondering why, if such big energy savings are both feasible and profitable, they haven’t all been exploited” (p. 254). Readers would wonder less if they were told about the extensive economic and technical literature which shows that people are often not interested in mundane energy conservation because they have an excessive demand on the rate of return,31 and which also demonstrates that achieved savings are commonly much more modest than initially believed. For example, the true costs of utility-subsidized electricity conservation are often significantly higher than the costs reported by utilities and these costs, in turn, are much higher than those suggested by the technical potential totals cited by the Lovinses.32 Moreover, a great deal of evidence shows that energy efficiency does not necessarily save energy.33 But this has long been the RMI practice: while they keep preaching the gospel of the super-efficient world, many devotees of efficiency and op- timization know that in many cases gaps between technical potential and actually accepted, durable, and reliable everyday performance will not be maximally narrowed because of often-unanticipated technical glitches, social inertia, basic human inconsistency, and personal priorities and preoc- cupations. And if all of this is about “making sense, making money” (to use the title of Norman Myers’s review of Natural Capitalism in Nature34), why are people not making sense and money right now? They are surely free to do so, for example, without having to wait for the appearance of customized Hypercars ordered by e-mail (so-called ultra-light hybrids, promised for 2005 or so). They could have been making money by virtue of the much lower capital, fuel, and maintenance costs of driving Honda Civics, rather than

Click to return to Table of Contents 172 PDR 26(1) B OOK REVIEWS

Chevy Suburbans, ever since those little, frugal, nimble vehicles appeared in the mid-1970s. Instead, people vastly prefer those monstrous trucks and vans and “sport-utility” machines costing commonly twice as much but being less than half as fuel-efficient as Civics: here we have the factor four in reverse, alive and well at every intersection in North America. Opportunities for savings are always there, always potentially large, but they are not taken for any number of reasons, not least because people are neither predictably nor consistently rational. How should I close this essay? Nothing is so salutary when looking ahead as taking a look back. In 1976 Amory Lovins published his signature piece in Foreign Affairs, in which he argued for an accelerated demise of centralized energy production and for the takeover of the world energy supply by soft—that is, decentralized, small, and renewable—conversions.35 Two years later, in a follow-up review, he plotted two very similar graph lines predicting fractional market penetration of these soft techniques in the US commercial energy market and concluded that their slope “does not appear unreasonably steep in view of the simplicity, short lead times, and diversity” of soft conversions.36 Those lines predicted that in the year 2000 the United States would draw about 35 percent of its commercial energy from soft energies. An- other two years later, in 1980, I listened, both bemused and irked, when at an international energy conference the Lovinses concluded their sales sermon about the inevitability of soft energies during the coming generation by saying: “It must be that way.”37 The year 2000 is here, and renewable energies supply about 7.5 percent of US primary energy. But half of this is hydroelectricity generated by water compounded by large dams, definitely no species of small, decentralized soft-energy production.38 Biomass (mostly combustion of wood waste by lumber and pulp and paper industries and some fuel alcohol), wind, and solar energy provide just over 3 percent of all US energy use, or about one-tenth of Amory Lovins’s “not unreasonable” projection. Being off by more than 90 percent—factor ten in reverse—hardly qualifies as a triumphant vision of an inevitable future. The world does not run by blueprints, not even if they come with sooth- ing qualifiers of “soft” or “natural.” The world does not follow any prescribed linear path, and taxation of everything is surely not salvation of everything. The world of impressively higher efficiency may actually be, as is ours today compared to the one of a millennium or a century ago, one of much greater waste: unintended consequences are not eliminated by wishful thinking. The US economy, we are told, should simply rid itself of $2 trillion of waste that is “a built-in feature of an outmoded industrial system.” The trio provides a long list of things to be eliminated, ranging from hidden social

Click to return to Table of Contents B OOK REVIEWS PDR 26(1) 173 costs of driving, guarding sea lanes bringing oil, and bad dietary choices, to current insurance systems, illegal drugs, lawsuits, complex tax codes, and all government waste. The list reads like part of a naive sermon: humanity shedding its cherished modern vices just because the three supremely rational utopians tell it to do so. Too bad that we are not told how the trio will make society virtuous within a generation when countless religious visionaries, social and economic reformers, and political dictators (from be- nevolent to brutal) have failed to prevent or eliminate even lesser vices throughout recorded history. Unrevealed modalities of this great purge aside, would not ridding the country of nearly a third of its gross domestic product in a matter of years have some very far-reaching side effects, including many unforeseeable and unpleasant consequences? And so to you, younger readers—who, after finishing Natural Capital- ism, are eagerly anticipating a world filled with gas-reforming plants at every wellhead so the continent could be flooded with hydrogen-feeding myriads of silent fuel cells, where only healthy food choices come from bucolic roof patches that use no synthetic fertilizer, where nobody sues anybody else for the sake of maximized efficiency, where factories that produce Wal- Mart junk ooze no waste, where government does not waste a penny, and where super-sleek F-16-like indestructible composite-fiber hypercars (sorry, HypercarsTM, or I’d get sued) that weigh less than you do, create no environmental burdens and act like power plants while you sleep—my only counsel is: beware! Turning once more to those durable Romans, I will quote Pliny the Younger who, writing nineteen centuries ago, put it best: “It is better to believe the world than individuals. For individuals can deceive and be deceived, whereas no one ever fooled the world.”

Notes

1 All quotes from the book’s dust jacket. a lot of money on the side by getting into al- 2 Ernst von Weizsäcker, Amory B. Lovins, ternative sources of energy and energy con- and L. Hunter Lovins, Factor Four: Doubling servation. This is a huge deal.” Cited in: “Nat Wealth, Halving Resource Use (London: Earth- Cap Buzz,” Rocky Mountain Institute Newsletter 15 scan, 1997). This collection of examples of (Fall 1999): 3. more-efficient economic activities, although 4 On closer examination it turns out that still too wide-eyed, is much more realistic than few technical revolutions are worthy of that the new book. For more on Factor Four, see term. Certainly not the Neolithic Revolution, Vaclav Smil, “Nature’s services, human follies: although school children will probably be A review essay,” Population and Development Re- brainwashed forever with Gordon Childe’s view 24 (1998): 613–623. mistaken idea that our species transformed it- 3 The quote comes from President Clin- self in a matter of a few generations from gath- ton’s speech to a Democratic National Commit- erers to farmers. Overwhelming archaeologi- tee dinner: “. . . if you read it, you will be con- cal evidence reveals instead the evolutionary vinced that whatever you’re doing and character of the change and a prolonged co- however well you’re doing it, you could make existence of the two modes of subsistence.

Click to return to Table of Contents 174 PDR 26(1) B OOK REVIEWS

And definitely not the Industrial Revolu- 9 Of course, a brand new super-efficient tion, which, too, was a complex, protracted af- building can easily use less than one-tenth the fair rather than a rapid shift of economic forces. energy of an old, shoddily built structure—but For example, almost all of England’s coalfields a country’s entire building inventory will not operating in 1900 were opened up between be replaced or retrofitted, or plastered over 1540 and 1640; and the country still had more with photovoltaic collectors, in two decades. shoemakers than coal miners, and more black- Similar limits to structural shifts are true for smiths than ironworkers in 1851: R. Cameron, transportation and, given the fact that energy “A new view of European industrialization,” use in many processes is nearing thermody- The Economic History Review 38 (1985): 1–23. namic limits, also for some sectors of indus- 5 Hopes pinned on decentralized, small- trial production. scale, renewable energy conversions during the 10 And medieval monasticism’s quest for 1970s owed a great deal to an uncritical self-sufficiency made important contributions countercultural embrace of Schumacher’s du- to innovations that eventually resulted in bious advocacy of worth measured by size: E. Western technical supremacy: Lynn Townsend F. Schumacher, Small Is Beautiful: Economics as White, Medieval Religion and Technology: Collected if People Mattered (New York: Harper & Row, Essays (Berkeley, CA: University of California 1973). Amory Lovins’s labeling of such con- Press, 1978); George Basalla, The Evolution of versions as soft, a thoroughly misleading but Technology (Cambridge: Cambridge University very captivating term (after all, it is the mar- Press, 1988). keting, not the substance that sells!), and his 11 Roger D. Sorrell, St. Francis of Assisi and vastly exaggerated claims for the speed of their Nature: Tradition and Innovation in Western Chris- commercial penetration, only raised more false tian Attitudes toward the Environment (New York: hopes. For a contemporaneous critique of this Oxford University Press, 1988). small-is-beautiful renewable “solution” see Vaclav Smil, “Renewable energies: How much 12 Among many fascinating accounts of and how renewable?” The Bulletin of the Atomic this quest in different industries I would par- Scientists 35(1979): 12–19. For a recent review ticularly recommend those dealing with the of renewable energy performance in the early history of commercial electricity: David United States, see Robert L. Bradley, Jr., Re- E. Nye, Electrifying America: Social Meanings of a newable Energy: Not Cheap, Not “Green” (Wash- New Technology, 1880–1940 (Cambridge, MA: ington, DC: Cato Institute, 1997). The MIT Press, 1990); Margaret Cheney, Tesla, Man Out of Time (Englewood Cliffs, NJ: 6 A group of 16 scientists, economists, Prentice-Hall, 1981); Matthew Josephson, government officials, and business people from Edison: A Biography (New York: McGraw-Hill, Europe, the United States, Japan, Canada, and 1959). India convened in Carnoules, France in 1994 “to discuss their belief that human activities 13 And coke-fueled blast furnaces have were at risk from the ecological and social im- become even more efficient even faster: their pact of materials and energy use” (p. 11). The energy consumption is now about 1/15 of Factor Ten Club, as the participants called what it was in 1800: Vaclav Smil, Energy in themselves, issued a declaration that contained World History (Boulder, CO: Westview Press, the claim just quoted. 1994), p. 180. 7 The United States consumed energy 14 Vaclav Smil, Greening the Earth (Cam- equal to about 2.3 billion tons of oil in 1999, bridge, MA: The MIT Press, 2000). or 8.5 t/capita; India used 300 million tons: 15 See Smil, cited in note 13, p. 231. British Petroleum Amoco, Statistical Review of 16 The Parthenon is marble from Penteli- World Energy 1999 (London: BP Amoco, 2000) kon. The Pantheon is a bold, beautiful, and >http://www.bpamoco.com/worldenergy< clever combination of bricks and concrete (per- 8 In 1997 China applied 23.6 million tons haps the greatest of Roman material inven- of nitrogen in synthetic compounds, France 2.3 tions)—and the concrete gets lighter (made million tons: Food and Agriculture Organiza- with tufa, travertine limestone, and pumice) tion, FAOSTAT Agriculture Data (Rome: FAO, as the coffered vault converges on the stun- 2000) >http://apps.fao.org< ning central eye. Too few people realize that

Click to return to Table of Contents B OOK REVIEWS PDR 26(1) 175 this span was never surpassed by preindustrial tality and morbidity benefits from meeting the builders: Michelangelo’s double dome of San air quality standards in the Los Angeles Basin Pietro spans 41.75 m, Brunelleschi’s Santa ended up with totals ranging from as little as Maria del Fiore just a smidgen over 42 m. $2.4 billion to as much $20 billion: J. V. Hall 17 And today’s best gas turbines have car- et al., “Valuing the health benefits of clean air,” ried this declining mass/power trend by yet Science 255 (1992): 812–816. Clearly, an order- another order of magnitude since the late of-magnitude difference provides a poor guide 1930s: Smil, cited in note 13, pp. 177 and 231. for good public policy. 18 Iddo K. Wernick and Jesse H. Ausu- 25 Among recent reports see W. O. C. M. bel, “National materials flows and the environ- Cookson and M. F. Moffatt, “Asthma: An epi- ment,” Annual Review of Energy and the Environ- demic in the absence of infection?” Science 275 ment 20 (1995): 463–492. Of course, the use (1997): 41–42. of silicon, aluminum, titanium, and plastics has 26 For the latest published consensus been rising. range offered by the Intergovernmental Panel 19 Calculated from data in Sam H. Schurr on Climate Change see Climate Change 1995: and Bruce C. Netschert, Energy in the American Impacts, Adaptations and Mitigation of Climate Economy 1850–1975: An Economic Study of Its His- Change (New York: Cambridge University tory and Prospects (Baltimore, MD: Johns Press, 1996). Hopkins University Press, 1960); Angus 27 The authors appear to be entirely ig- Maddison, Monitoring the World Economy, 1820– norant of the fundamental role of synthetic ni- 1992 (Paris: OECD, 1995); US Energy Infor- trogen fertilizers in the very existence of at least mation Administration, Annual Energy Review one-third of humanity: see Smil, cited in note 1999 (Washington, DC: EIA, 2000). That rate 14, and for a shorter exposé: Vaclav Smil, of sustained decline would save about 20 per- “Population growth and nitrogen: An explo- cent of energy use in one generation, far less ration of a critical existential link,” Population than the “prophetic” 90 percent drop endorsed and Development Review 17 (1991): 569–601. by the trio. Nor do the authors appreciate a large land- 20 Juan Martínez-Alier, Ecological Econom- sparing effect of synthetic fertilizers: without ics: Energy, Environment, and Society (Oxford: Ba- them all densely populated countries would sil Blackwell, 1987); Frederick Soddy, Wealth, have to cultivate much larger areas of land, Virtual Wealth and Debt: The Solution of the Eco- further reducing their biodiversity: Dennis T. nomic Paradox (London: Allen & Unwin, 1933); Avery, “Saving nature’s legacy through better Technocracy, Inc., The Energy Certificate (Savan- farming,” Issues in Science and Technology 14 nah, OH: Technocracy, Inc., 1938); Howard T. (1997): 59–64. Odum, Environmental Accounting: EMERGY and Environmental Decision Making (New York: John 28 Total global food harvest is now about Wiley, 1996). 2.8 billion tons of dry matter in cereals, legu- minous and sugar crops, tubers, fruits, and veg- 21 Continuing importance of this practice etables; cereals account for about two-thirds: is described in Vaclav Smil, “Crop residues: FAO, cited in note 8. Agriculture’s largest harvest,” BioScience 49 (1999): 299–308. 29 Global vegetable harvest is now about 22 Since 1960 the area of European for- 60 million (dry matter) tons a year: FAO, cited ests grew by about 12 percent, and the French in note 8. Next time I am in Beijing I must total was up by almost 30 percent, impressive correct my oversight of many trips over the gains in just two generations: FAO, cited in past 20 years and find those rooftop gardens note 8. producing plenty of food. 23 Robert L. Bradley, Jr., The Increasing 30 William L. Chameides et al., “Growth Sustainability of Conventional Energy (Washing- of continental-scale metro-agro-plexes, re- ton, DC: Cato Institute, 1999); US Energy In- gional ozone pollution, and world food pro- formation Administration, cited in note 19. duction,” Science 264 (1994): 74–77. 24 One example of this challenge is that 31 “Money to burn,” The Economist, 6 various assumptions in estimating annual mor- January 1990, p. 65.

Click to return to Table of Contents

176 PDR 26(1) B OOK REVIEWS

32 Paul L. Joskow and Donald B. Marron, 36 Amory B. Lovins, “Soft energy tech- “What does a negawatt really cost? Evidence nologies,” Annual Review of Energy 3 (1978): from utility conservation programs,” The En- 477–517. ergy Journal 13 (1992): 41–74. 37 For a printed version of that presenta- 33 One of the best recent reviews of this tion see Amory B. Lovins and L. Hunter Lovins, topic is Horace Herring, Does Energy Efficiency “If I had a hammer,” in Robert A. Bohm et al. Save Energy? The Economists Debate (Milton (eds.), World Energy Production and Productivity: Keynes: The Open University, 1998). Also avail- Proceedings of the International Energy Symposium able at >http://www-tec.open.ac.uk/eeru< (Cambridge, MA: Ballinger Publishing, 1981), 34 Norman Myers, “Making sense, mak- pp. 123–146. For my brief response in the same ing money,” Nature 402 (1999): 13–14. volume see pp. 151–152. 35 Amory B. Lovins, “Energy strategy: The 38 US Energy Information Administra- road not taken?” Foreign Affairs 55 (1976): 65– tion, cited in note 19. 96.

Click to print book review Click to return to Table of Contents

B OOK REVIEWS PDR 26(1) 177

GEORGE J. BORJAS Heaven’s Door: Immigration Policy and the American Economy Princeton, NJ: Princeton University Press, 1999. xvii + 263 p. $27.95. This book represents a lucid and cogent amassing of the arguments that George Borjas, one of the leading immigration economists in the United States, has been making for more than a decade about the economic effects of the post-1965 waves of US immigrants. He stresses the declining skills of immigrants relative to those of natives, shows the distributional consequences of these changes for low-skilled native workers, and argues for a new approach to immigration policy that puts a premium on immigrants’ skills. The core of the book summarizes much of Borjas’s previous work in this area, with appropriate updating to the 1996–98 Current Population Surveys (CPSs) and occasional backdating to include the 1970 census. The stylized facts are presented as “symptoms” of US immigration policy during the past 40 or 50 years. These include the following: (1) the volume of immigration to the United States is at an all-time high when one considers both legal and unauthorized immigrants; (2) the skills and economic performance of immigrants have declined relative to those of natives, not only because successive waves of immigrants lag farther and farther behind natives but also because there is little catching up to natives associated with added years of US labor-market experience; (3) the declining relative economic performance of immigrants can be explained by a shift in their countries of origin—away from industrial countries and toward developing ones—that was induced by the 1965 amendments to the Immigration and Nationality Act; (4) the presence of immigrants produces little increase in the average income of native workers, but immigrants do redistribute income away from less-skilled natives to those who are better off; (5) states and localities with relatively large numbers of poor immigrants are experiencing higher tax bills for schools and other immigrant-related services; (6) the skill gap between immigrants and natives has a “half-life” of about one immigrant generation, which means that ethnic skill differences may persist for three or more generations; and (7) a faster integration of immigrants into the economic mainstream is hampered both by “the culture, attitudes, and economic opportunities that permeate the ethnic networks” (p. 14)— what Borjas calls “ethnic capital”—and by a tendency for ethnic ghettos to insu- late immigrants from the dominant culture. Borjas also seeks to “reframe the immigration debate,” moving from arguments about symptoms to a discussion of the optimum objectives of US immigration policy. He believes that immigration policy should be used primarily to maximize the economic wellbeing of the native population, where this population is defined as all persons currently residing in the United States and where economic wellbeing de- pends not only on per capita income but also on no further worsening of income inequality. Borjas then argues that US immigration policy should try to increase substantially the average education and skill level of legal immigrants, that a point system similar to ones that have been used successfully in Canada and Australia is a useful policy tool to select individuals who possess favorable socioeconomic characteristics, and that a cap of roughly 500,000 legal immigrants per year might be about the right number of migrants to admit.

Click to print book reviews Click to return to Table of Contents 178 PDR 26(1) B OOK REVIEWS

Perhaps no one has contributed more to our understanding of the economic consequences of immigration to the United States than Borjas, and much of his work is now embodied in conventional wisdom as illustrated by the National Acad- emy of Sciences’ comprehensive study of the economic and demographic impacts of US immigration (Smith and Edmonston 1997). In early debates, Borjas showed that new immigrants did not make rapid wage progress after they entered the United States—that what seemed in a single census like rapid wage mobility with added years of US experience was instead the result of declining skills of immigrants relative to natives across successive immigrant cohorts. He traced these declining skills not to what is happening with immigrants from a particular country, but rather to the fact that source countries are increasingly those in which average levels of education are lower. His arguments also played an influential role in the decision by Congress in 1990 to increase the number of visas for skilled immigrants. Borjas’s policy prescriptions are likely to be more controversial than his scholarly findings and interpretation. For one thing, Congress considered the idea of a point system during the 1980s and quickly abandoned it on political grounds. More- over, restricting legal immigration only to those with a high level of talent will greatly expand the incentives for undocumented migration by persons with insufficient skills to obtain a “passing grade” under the point system. Borjas suggests that undocumented immigration might be taken more seriously if his proposed cap of 500,000 permanent resident visas each year were reduced by the size of the annual net undocumented flow, which is now estimated to be between 200,000 and 300,000. But this measure, if enacted, would be likely to reduce the allowable number of legal immigrants to close to zero while not having the desired effect on the average skill level among new migrants. If legal routes of entry are foreclosed, many individuals with fewer skills will still be determined to work in the United States. More fundamentally, however, Borjas pays too little attention to undocumented immigration both in the interpretation of his empirical results and in his policy prescriptions. Most of his analyses rely on census data that include natives, natu- ralized citizens, legal aliens, and undocumented migrants. Borjas seems to want to implicate the 1965 changes in US immigration law, amendments that abolished national-origin quotas and substituted a preference system based on having close relatives in the United States, as being responsible for the subsequent decline in immigrants’ relative skills. But it is also true that the ending of the Bracero program in the 1960s gave rise to a sharp increase in the flow of unauthorized immigrants into the United States. This too, in all probability, has contributed significantly to a decline in immigrants’ relative education and skills, especially because half of the undocumented residents in the United States are from Mexico. In short, it is not at all clear how much of the post-1965 decline in the skills of new immigrants relative to natives is attributable to legal immigration reforms in the 1960s and how much to a surge in undocumented immigration that began at the same time and that shows no signs of abating. As long as the data that Borjas uses contain representative amounts of undocumented migrants, one may wonder whether reform of America’s legal immigration system to increase the skill levels among migrants is the whole answer.

Click to return to Table of Contents B OOK REVIEWS PDR 26(1) 179

The data problem, of course, is not of Borjas’s making. The Census Bureau does not ask about one’s legal status in the decennial census or the monthly CPS. Nev- ertheless, one might ask how different Borjas’s findings would be if undocumented migrants could somehow be removed from the data and his analyses redone based just on US citizens and legal immigrants. His policy prescriptions would have a firmer basis if the same amount of relative decline in migrants’ skills were detected. In recent work, Passel and Clark (1998) have developed an imputation scheme to assign legal status to individuals in the census and CPS. It could be instructive to replicate Borjas’s analysis using only a sample of “legal” immigrants. A final point is that strengthening American resolve to curb undocumented migration would have many of the same consequences Borjas expects from his proposals for legal immigration reform. Undocumented immigration may account for roughly one-third of net total US immigration; undocumented migrants are typically less skilled than legal immigrants; and most unauthorized immigrants tend to originate in a handful of countries. Consequently, more effective controls over unauthorized migrants would increase the average level of education and skill among the new immigrant streams, permit more diversity in terms of sending countries, and reduce the overall volume of immigration. Each of these is an important objective that Borjas hopes will be accomplished by his proposed point system and related reforms on total numbers. Moreover, evidence from “successful” immigration states such as New Jersey suggests that a well-educated foreign- born population that draws from many sending countries with no single country predominating and that resides in the United States in legal status not only eases some of the potentially adverse economic and fiscal effects of immigration, but also contributes to a smoother integration of immigrants into the social, political, and economic mainstream and to a more tolerant climate of public opinion toward immigrants (Espenshade 1997). None of these comments should be interpreted, however, as undercutting Borjas’s achievement in increasing our understanding of the impact of immigration on the economy and residents of the United States. Heaven’s Door is rigorous and packed with solid, empirically based findings; it crowns a period of remark- able scholarly productivity for its author; and it sets the benchmark for what we know about immigration to the United States and its economic effects.

Office of Population Research THOMAS J. ESPENSHADE Princeton University

References

Espenshade, Thomas J. 1997. Keys to Successful Immigration: Implications of the New Jersey Ex- perience. Washington, DC: The Urban Institute Press. Passel, Jeffrey S. and Rebecca L. Clark. 1998. Immigrants in New York: Their Legal Status, In- comes, and Taxes. Washington, DC: The Urban Institute. Smith, James P. and Barry Edmonston. 1997. The New Americans: Economic, Demographic, and Fiscal Effects of Immigration. Washington, DC: National Academy Press.

Click to return to Table of Contents 180 PDR 26(1) B OOK REVIEWS

DOUGLAS S. MASSEY, JOAQUÍN ARANGO, GRAEME HUGO, ALI KOUAOUCI, ADELA PELLEGRINO, AND J. EDWARD TAYLOR Worlds in Motion: Understanding International Migration at the End of the Millennium Oxford: Clarendon Press, 1998. xiv + 362 p. $85.00. Worlds in Motion is one of several recent attempts at theorizing about international migration. The authors are mostly well-known scholars and the volume was produced under the auspices of the International Union for the Scientific Study of Population. The ambitiousness of the undertaking is clearly if immodestly stated in the volume’s preface: “ . . . to develop a comprehensive theoretical understanding of the social and economic forces responsible for international migration . . . in the late twentieth century.” The goal is largely met. I say “largely,” because it is in synthesizing familiar theories, rather than in theorizing per se, that the volume’s strength lies. Although the volume also assesses the applicability of the theories in several world regions, this part of the effort is much more mechanically done and thus less successful. The discussion of four major theoretical traditions associated with the initiation of a migration flow—(a) neoclassical economics, (b) the “new economics” of labor migration, (c) world systems theory, and (d) labor market segmentation— and of the two more recent theories associated with the continuation of such flows—social capital and cumulative causation—is intelligent and persuasive. Of course, there is little that is new in this regard, at least for readers of this Review, for the book’s two theoretical chapters have already appeared in their totality in Volumes 19 and 20. The added value lies instead with attempts to apply these theories in five world regions—North America, Europe (really, the European Union space), the Gulf States, Asia and the Pacific, and South America—as a means of testing the applicability of each theory in each region. (The absence of Africa is a major gap; the mandate of the Asia/Pacific chapter is overly broad.) The analytical treatment of the selected regions is useful and generally en- hances one’s understanding of the migration processes in each. However, the attempt to impose a single template across all five chapters is forced—even artificial—and in some ways it may even detract from the analytically interesting treatments. In fact, the exercise of simply rehearsing each of the major theories, seriatim, and normatively “testing” each one has a perfunctory character to it. This contrasts sharply with the real value of the substantive chapters, namely, the way in which several of the authors distill often large bodies of literature on vast geographic regions into observations that can actually become the basis for a more systematic comparative inquiry. For instance, the Asia/Pacific chapter usefully classifies states into three geographical “nodes” and cogently discusses their four distinctive features: (a) the important role a variety of societal institutions play in having created and managing a migration “industry”; (b) the almost unique role governments play in the entire migration process; (c) the fact that, increasingly, most states from the region are simultaneously importing and exporting immigrants; and (d) the region’s enormous potential for additional emigration. The chapter then identifies and discusses the critical migration-specific forces that will continue to be relevant in the years ahead. These are (a) the region’s demography (enormous labor surpluses in some countries together

Click to return to Table of Contents B OOK REVIEWS PDR 26(1) 181 with severe labor shortfalls in others), (b) the growing tradition of emigration in the region, and (c) unresolved within- and across-borders ethno-religious conflicts. Of particular interest in the Gulf States chapter is the discussion of these states’ attempts to maximize the economic benefits of migration by minimizing the migrants’ participation in society and associated costs. The “demonstration effect” of this approach, particularly in Japan, should represent one of the more interesting developments in the migration field, as Japan and possibly other countries attempt to reduce their “vulnerability” to immigration’s “downsides.” The authors’ treatment of two forces that have not received as rigorous a treatment as other variables in the migration field, population growth and the role of the state, is also intelligently done—albeit given short shrift. The authors correctly observe that despite the vast demographic literature, the relationship between population growth and international migration has been insufficiently theorized, and they speculate about some of these connections (pp. 286–292). Their speculations are refreshingly forward-looking and worth reflecting upon regardless of one’s pre- dispositions toward theory. The two chapters on the relationship between migration and development—at the national and the community level—are good summaries of the vast literature on this topic. The chapter focusing on the relationship between migration and community development seems to be inching its way toward a useful recasting of how one might look at this elusive connection, but it ultimately disappoints the reader. On balance this is a well-written book: the language is direct, the theoretical discussion eschews unnecessary jargon, neologisms are kept to a minimum, and the extensive summaries of the relevant bodies of literature are cogent. It is puz- zling, then, that the authors of the regional chapters chose to rely on statistical information that predates by several years the volume’s publication date of 1998. The fact that the volume’s explicit aim is to make a theoretical contribution does not make this criticism any less relevant. As the preface makes clear, each of the regional studies went through several iterations, yet the authors apparently chose to retain the data they used during those chapters’ earliest versions. Considering the time (and, presumably, “resources”) expended in the preparation of this IUSSP report, the dated statistical information, even if not crucial to the volume’s theoretical bent, detracts from the book’s usefulness. Furthermore, for a volume that offers many important insights, it is highly pedestrian in some notable respects; the stylized chapters on Europe and North America are two such instances. Theoreticians should find the volume useful as a single source where many of the most influential migration theories are treated with clarity, making the book the most valuable synthesis on these matters of which I am aware. The effort to test the general applicability of these theories in a variety of regions, if only by using a common template for all chapters, is also more successful than most similar attempts. And those interested in having a well-organized and analytically valuable thumbnail sketch of the migration subsystems in five world regions will find plenty to chew on.

International Migration Policy Program DEMETRIOS G. PAPADEMETRIOU Carnegie Endowment for International Peace

Click to return to Table of Contents 182 PDR 26(1) B OOK REVIEWS

ALISON GAMES Migration and the Origins of the English Atlantic World Cambridge, MA and London: Harvard University Press, 1999. xiii + 322 p. $45.00. In this notable contribution to the history of the English Atlantic world, Alison Games describes the experiences of thousands of settlers and travelers who left the port of London in 1635 for continental Europe and the far-flung colonies of England’s emerging American empire. The significance of her approach lies in its breadth. Unlike most historians of early modern migration to the New World who have tracked the paths of particular groups of migrants to specific American locales, Games takes as her canvas the entire English Atlantic—British settlements in the New World from Surinam to the Bahamas and from the Chesapeake Bay to northern Massachusetts. Above all, the book is a study of people on the move. Although the regional origins of migrants and their subsequent destinations are given careful consideration, much of the narrative emphasizes the pivotal role of migration itself in the formation of new Atlantic societies. The 1630s witnessed the beginnings of a mass movement across the Atlantic; the decade was the turning point after which England’s colonial project became self-sustaining. Existing colonies rapidly spawned new settlements, and as colonies multiplied so English America developed, not as geographically dispersed fragments isolated from one another but as a collection of highly diverse settlements connected in varying degrees by trade and migration. Intended as a measure to ensure that passengers embarking overseas had first taken the oath of allegiance and supremacy to the king and were in conformity with the Church of England, a decree of December 1634, issued by order of the Commission for Foreign Plantations, is one of the rare instances of government- sponsored registration of emigrants in the seventeenth century. Information com- piled by clerks at the port of London during 1635 provides a unique opportunity to investigate the social character of migration to America at a formative moment. Names and ages of some 4,900 men and women who set sail for the colonies are listed (together with details of 1,600 soldiers and 1,000 travelers bound for the continent), representing at least one-half of all emigrants who took ship for America from Britain’s ports during the year. They fall into two main categories: those bound for the plantation colonies of Virginia and the West Indies (3,415) and those whose religious convictions took them to the Puritan settlements of New England (1,169). Games confirms the quite different experiences of people going to the northern colonies compared to those of migrants bound for the Chesapeake and islands. Nearly four-fifths of migrants who ended up in the plantations during 1635 left London as indentured servants, bound to serve in the tobacco fields of Virginia, Barbados, and St. Kitts for typically between four and seven years. They conform to the general profile of indentured servants in this period: the great majority (86 percent to 95 percent) were male, aged between 15 and 24, unmarried, unskilled, and traveled alone. By contrast, New England migration was characterized by the movement of family and kin groups, the presence of men of substance, and a more balanced sex ratio. Although historians will not be surprised by her results, Games succeeds not only in providing an authoritative analysis of the data but also in bringing her findings to life through the use of examples of individual experiences. Leaving

Click to return to Table of Contents B OOK REVIEWS PDR 26(1) 183

London in the summer of 1635 to set up in trade, Nathaniel Braddock died within six months of arriving in the Chesapeake (pp. 42–45). Thousands of young men— petty merchants, factors, or servants—who took ship for Virginia in these years suffered the same fate. Edward Rainsford, on the other hand, arrived in Massa- chusetts as a young man of 26 and after a long and successful mercantile career died in 1680 at the age of 71 (pp. 185, 234). Taken together the experiences of travelers underline the great diversity of motives for leaving England and the mul- titude of trajectories their lives took in America. Migrants’ fortunes in the colonies are the focus of the second part of the book. Understandably, given the patchy evidence, Games’s treatment of their collective and individual experiences is somewhat uneven; far more information exists for New England settlers than for those who went to the Chesapeake and Caribbean. Nevertheless, she is able to track down sufficient numbers to convey an impression of the precariousness of life in the plantations and to emphasize significant differences between the colonies. Economic opportunities for men of humble status, for example, were much brighter in Barbados than in Bermuda or Virginia in the second half of the 1630s. In the case of New England, she is able to assess the impact of the 1635 arrivals on religious disputes in Massachusetts and to assemble impressive evidence about the propensity of migrants to continue moving within and beyond the colony. If Games’s description of migrants’ backgrounds and lives in individual colonies reveals few surprises, the advantage of her approach becomes evident in those chapters where she adopts a broad comparative perspective. Her study of different forms of Puritanism in New England, Bermuda, and Providence, Rhode Island demonstrates both the range of nonconformist beliefs carried across the Atlantic and the extensiveness of the Puritan diaspora. Similarly, her consideration of settlers’ movements from one colony to another and of their travels back and forth between America and England—sometimes on business, sometimes for pleasure— underscores the vital interconnections between colonial societies and with the mother country that “made it impossible for any part of the colonial world to evolve in isolation” (p. 191). This is an important book—written with verve, imagina- tion, and clarity—that should quickly establish itself as a standard text in the growing literature on Atlantic history.

International Center for Jefferson Studies JAMES HORN Monticello, Virginia

JOSEPH P. FERRIE Yankeys Now: Immigrants in the Antebellum United States 1840–1860 New York: Oxford University Press, 1999. xii + 223 p. $49.95. Joseph Ferrie has meticulously sampled data from shipboard passenger lists in the 1840s to begin his story of the first mass migration to the United States after colonial times. Most significantly, he followed specific individuals from the ships, linking them to the 1850 and 1860 censuses. His intent was to develop panel data that

Click to return to Table of Contents 184 PDR 26(1) B OOK REVIEWS demonstrate how the immigrants’ lives changed between departure from Europe and the census dates. In large part, this book is about the data set and what it tells about occupational status, residential shifts, and the accumulation of personal wealth following immigration. Ferrie asks such questions as, Did the skilled Ger- man immigrant in Milwaukee come with or without skills? What skills did the Irish laborer bring from Europe to New York? Was there a relation between these skills and the work immigrants actually performed in New York in 1850? To provide context, data for native-born Americans are also developed from standard PUMS (Public Use Microdata Series) samples of the 1850 and 1860 censuses. Attention to this wave of American immigration is overdue. Between 1847 and 1860, well over 100,000 immigrants per year arrived in the new country, filling labor needs in the industrializing cities of the eastern seaboard and fueling the westward expansion. The 370,000 arrivals in 1850 pushed the immigration rate above 15 per thousand total population per year for the first (and last) time in the country’s history. Ferrie’s focus is on immigrants from Britain, Ireland, and the German States (who together constituted 93 percent of immigrants between 1840 and 1860). As he points out, despite the fact that reception policies developed in the 1850s facili- tated later American receptivity to immigration, it is in effect the “forgotten” migration. The reason, Ferrie writes, is that this wave of immigrants quickly and thoroughly assimilated into nineteenth-century labor markets. It was at this time that the United States first made decisions permitting the quick and ready assimilation of immigrants into the economy as free labor. As with later immigrations, this early wave was accompanied by a backlash (focused by the briefly influential “Know-Nothing” party), as native-born artisans experienced new competition in local labor markets. For the immigrants themselves it was a time of downward, upward, and sideways social mobility. The patterns in this mobility are the focus of Yankeys Now. (The alternative spelling of Yankees—originally the nickname of those who lived in New England—is taken from correspondence by an English couple who had settled in upstate New York. In 1872 one of them wrote to friends in Yorkshire, “I don’t think we could live in England. . . . We are yankeys now.”) The chapter of greatest general interest is the penultimate, which asks what effect immigration had on the larger labor market in the period 1850–60. As in studies of late-twentieth-century US immigration, Ferrie finds that the immediate effects varied for different portions of the labor market. On the East Coast, wages, particularly for native-born skilled artisans, fell, as machine production using unskilled Irish labor revolutionized manufacturing. Likewise, effects were most severe in areas where immigrant populations concentrated. For example, in the textile manufacturing center of Lowell, Massachusetts, the proportion of the population that was native born declined from 90 percent in 1849 to only 35 percent some six years later. Particularly affected in Lowell were artisans working in guild trades such as iron casting, carpentry, shoemaking, tailoring, and cabinetmaking. Wages fell in the face of increased competition from semiskilled, but quick-learning immigrants. This initiated a wave of strikes throughout the Northeast and a surge in strength for the nativistic “Know-Nothings” in the 1854 election, before they passed from the political scene two years later. The parallels to more modern examples

Click to return to Table of Contents B OOK REVIEWS PDR 26(1) 185 around the world are obvious. Recently both in the United States and in Europe, labor migration has produced similar short-term effects. Ferrie is an economic historian, and his analysis is focused by econometric detail. From this context he has done the yeoman’s work of sampling shipboard manifests, combing census schedules, and converting both into data appropriate for hazard modeling. Finally, he makes calculations needed to exploit the data. This is important work and, one hopes, will be a basis for a broader understanding of this critical era in American immigration history. Generalists will find, however, that much of Ferrie’s narrative drifts off into technical discussions of the compromises and assumptions necessary to exploit the data set. Because of this, the book does not elaborate the broader context relevant to students of migration, a task for which the data analyzed are otherwise well suited. At points Ferrie briefly alludes to the many qualitative sources of data (e.g., diaries and letters) that could have fleshed out his statistical analysis. Despite this acknowledgment, his impulse is always to rush to the next variable and squeeze the next inference out of his data set. As a result, he does not dally with the color of the human lives behind the statistics that qualitative sources offer. Thus, while the book may be important for understanding the narrower topic of immigrants in pre–Civil War labor markets, the broader question of how or why America’s first waves of immigrants assimilated socially is not addressed.

Department of Sociology and Social Work TONY WATERS California State University, Chico

FRANCINE M. DEUTSCH Halving It All: How Equally Shared Parenting Works Cambridge, MA: Harvard University Press, 1999. 327 p. $24.95. Francine Deutsch’s book picks up where Arlie Hochschild’s pathbreaking 1989 book The Second Shift left off, by looking at the struggles, successes, and failures of a sample of 300 dual-earner families with young children in America in the 1990s. Deutsch begins with the premise that gender equality, while not impossible, is negotiated in the face of strong ideological and institutional barriers that must be overcome for couples to share parenting and breadwinning. The book illustrates the different paths that lead some couples to discover equally shared parenting as the workable solution to their work–family dilemmas, often after more traditional arrangements failed. Deutsch also sensitively portrays the couples who began parenting with egalitarian ideals, only to later abandon their principled belief in gender equality in the face of deeper personal and cultural resistance. The couples’ experiences provide a fascinating glimpse into the worlds of modern dual-earner families across the class spectrum (unfortunately there is not much variation by race among the surveyed couples). Deutsch’s data provide several surprising insights into the dynamics that inform couples’ negotiations about the division of labor in the family. First, in con-

Click to return to Table of Contents 186 PDR 26(1) B OOK REVIEWS trast to structuralists’ emphasis on how gender inequality at work encourages men to specialize in breadwinning (a sentiment echoed by many of her respondents), she finds that the men in her sample earn more than their wives only because they work longer hours. In those families in which husbands were primary bread- winners, wives on average earned an hourly wage equal to that of their husbands but worked fewer hours. Deutsch argues that men’s greater earning power is mostly the result of couples’ decisions about who should allocate more time to the market, rather than the cause of such time allocations. In contrast to rational-choice theories about specialization, Deutsch finds that men whose hourly wages were clearly higher than those of their wives had been the beneficiaries of an earlier joint decision to emphasize the husband’s career at the expense of the wife’s, despite their equal earning potential at that earlier time. In explaining these paradoxical findings, Deutsch emphasizes the cultural or ideological construction of gendered parenting that leads so many women to feel that child care is their special obligation and responsibility and so many men to feel that their wives are more competent and capable than they are at nurturance. She shows how the decision to emphasize men’s economic contributions stems more from deeply rooted feelings about domesticity than from rational calculations about earnings. Yet couples routinely attribute their choices to rational calculations, despite the evidence, sometimes because they are uncomfortable with a discourse of sexual difference and want to support gender equality in principle. Some of the most poignant episodes in the book come from Deutsch’s descriptions of the losses of marital intimacy and career aspirations among the women in un- equally sharing arrangements, and how the women resist seeing their disadvan- tage or resign themselves to it. In contrast to her analysis of the unequal couples, Deutsch relies more heavily on a materialist or structural explanation of why couples “halve it all.” She portrays a picture of rational decisionmaking in the stories of couples who share breadwinning and parenting equally. These are the couples who, faced with the irrationality of women’s greater responsibility for child care while still working for pay, consciously or unconsciously created a system of equal sharing. Some began this way and “stayed the course” as their children were born and placed greater demands on the household. Other couples became equal sharers after periods of specialization that left one or both partners dissatisfied with the marital relationship or with their material standard of living. Again, the most surprising finding is that many equal sharing couples began parenthood with a period of intense maternal specialization in child care, but shifted successfully in spite of the strong mother–child bond and father’s relative lack of early involvement. Deutsch emphasizes that these couples span the class spectrum and are not always situated in social networks that support nontraditional understandings of masculine and feminine roles. Yet these couples recognized the advantages of equal sharing for both themselves and their children and persisted even when employers, coworkers, or relatives questioned their arrangements. The book will be of interest to scholars involved in the study of contemporary families and gender differentiation in the intersection of work and family life in developed countries. Because it is an essentially descriptive study, it suggests rather than confirms theories about the prominence of economic versus ideological fac-

Click to return to Table of Contents B OOK REVIEWS PDR 26(1) 187 tors in the construction of modern family life. It is rich with insights about how contemporary parents view childbearing and childrearing, how they negotiate family responsibilities in an economic context that makes wives’ employment commonplace, and how they ultimately feel about their responsibilities. While Deutsch champions the wisdom and greater wellbeing of the equally sharing couples, her approach is neither dogmatic nor myopic about the dilemmas and sacrifices these couples often confront. Nevertheless, my strongest criticism of the book is the ease with which the author accepts the greater career sacrifices that equal sharers had to make relative to their peers unencumbered with childrearing responsibilities. She accepts, with some degree of equanimity, the fact that equally sharing mothers and fathers will both be penalized in the workforce. To me, this acceptance is disconcerting and deserves greater critical discussion. But this reservation does not detract overall from the quality of the work or the scholarly findings therein.

Department of Sociology JENNIFER GLASS University of Iowa

SHORT REVIEWS

by John Bongaarts, Martin Brockerhoff, Susan Greenhalgh, Baochang Gu, Geoffrey McNicoll

ORLEY ASHENFELTER AND DAVID CARD (EDS.) Handbook of Labor Economics, Volume 3 Amsterdam and New York: Elsevier, 1999. 1746 p. in 3 parts. $295.00. The Elsevier Handbooks in Economics aspire to be “a definitive source, reference, and teaching supplement for use by professional researchers and advanced gradu- ate students.” Each handbook comprises commissioned survey articles that together span a particular branch of economics. The Labor Economics handbook, edited by Orley Ashenfelter and Richard Layard, initially appeared in 1986. It then contained 22 chapters in two substantial volumes. The present set, described as Volume 3, consists of three books (3A, 3B, and 3C) that add a further 31 chapters. They include George J. Borjas, “The economic analysis of immigration”; Jere R. Behrman, “Labor markets in developing countries”; Stephen Nickell and Richard Layard, “La- bor market institutions and economic performance”; and Joseph G. Altoni and Rebecca Blank, “Race and gender in the labor market.” The editors see Volume 3

Click to return to Table of Contents 188 PDR 26(1) B OOK REVIEWS as representing a much-expanded field of research and one with much greater empirical content. Holding to the idea that it is a continuation of the earlier volumes, the pagination follows on: Volume 3A starts at page 1277. The indexes, however, apply just to these three parts, with the same index pages printed in each. Another handbook in the series, the Handbook of Population and Family Eco- nomics, Volume 1, was reviewed by Mark Perlman in PDR 24(4).

PAUL BOYLE, KEITH HALFACREE, AND VAUGHAN ROBINSON Exploring Contemporary Migration Essex, UK: Addison Wesley Longman, 1998. xiv + 282 p. $41.00. Exploring Contemporary Migration is overwhelming in the aspects of spatial mobility covered—theory, data sources and methods of measurement, historical and contemporary trends, and specialized issues such as forced migration, changes in labor migration in the 1990s, and migration policy, the last discussed in terms of “social engineering.” Such breadth of scope inevitably sacrifices depth; the book is an ideal tool for teaching students but will offer few new insights to, and provoke little argument from, seasoned scholars. The authors do not, however, assume a neutral stance in their perspectives on migration. For example, they contend that economic, rational choice, and deterministic theories of migration—which predict movement by measuring characteristics of individuals or of different places—have received undue attention as compared to cultural theories and humanist accounts of migration, which acknowledge that such characteristics never bear a tight relationship with the likelihood of migrating. Political power, for instance, motivates or restricts the movement of individuals and social groups differentially. The authors advocate, instead, a pluralist approach to understanding migration not unlike Giddens’s theory of structuration, attaching equal importance to human agency and social structure. Elsewhere, the conventional life-cycle model, which identifies migration in strictly bounded stages in people’s lives—schooling, marriage, childrearing, retirement—is rejected in favor of a life-course approach, which acknowledges the diversity behind people’s changing lives. Arguing against a life-cycle perspective, the authors note that “social conventions and norms are now changing so rapidly and frequently that almost every cohort faces a different range of choices at successive life transitions than those offered its predecessor.” These and numerous other examples indicate the authors’ disinclination to accept generalizations regarding migration, and their preference for qualitative research approaches that, perhaps impractically, often require extensive and detailed data collection. A chapter on “cultures of migration” is particularly ambitious and makes for lively reading. Here, the discussion veers from Woody Guthrie’s “escapist” lifestyle to the nomadism glorified by Jack Kerouac and practiced, in different form, by Australian aborigines, to the diasporas of Gypsies, Jews, and the Irish. A chapter on migration and “social engineering” keenly observes how governments—in Eng- land, Germany, Israel, South Africa, the United States, and many developing countries—have used migration policy, including in some instances forced emigration, more effectively than manipulation of fertility and mortality to shape national char-

Click to return to Table of Contents B OOK REVIEWS PDR 26(1) 189 acteristics. This is evidenced by the 1994 crisis in Rwanda, which produced an enormous toll of deaths but several times as many departures from the country. For the many skeptics who doubt the general effectiveness of migration policies, this book may offer thought-provoking counter-evidence.—M.B.

WEIXING CHEN The Political Economy of Rural Development in China, 1978–1999 Westport, CT: Praeger Publishers, 1999. xiv + 173 p. $55.00. How could the Chinese Communist Party overcome the ideological and political constraints on China’s economic development without changing the framework of one-party rule? Chen, a political scientist at East Tennessee State University, analyzes the political economy of rural development in China over the last two decades. Development in China, he argues, can be viewed as a search by the state for a balance between the dual goals of improving people’s welfare and strengthening the state, between ideological purity and pragmatic measures fostering development. The Chinese Communist Party, Chen proposes, is being transformed from a party of politics engaged in “class struggle” into a party of economics primarily preoccupied with economic growth. “Deng Xiaoping Theory” has been established as the new official ideological discourse; the village conglomerate has emerged as a new form of political-economic unit in rural China; and village elections have become a means for the state to reorganize China’s peasant majority. Chapter notes, bibliography, index.—B.G.

DAVID A. CROCKER AND TOBY LINDEN (EDS.) Ethics of Consumption: The Good Life, Justice, and Global Stewardship Lanham, MD: Rowman & Littlefield, 1998. xviii + 585 p. $89.00; $27.95 (pbk.). This collection begins promisingly with an essay by Herman Daly on consumption and the environment, and another by Mark Sagoff taking issue with many of Daly’s arguments—and more generally with the propositions of ecological economists. On the widely cited study by Vitousek and colleagues that found 40 percent of net primary production being used in human-dominated ecosystems, Sagoff writes: “This argument rests on two premises: first, that total net primary production is fixed or limited in nature; and second, that economies, in order to grow, must co- opt correspondingly more organic matter. Both premises are false.” The rest of the collection, aside from a few empirical pieces setting out facts about consumption, presents almost a surfeit of worthy thinking about consumption: among other things, essayists call for developing habits of “morally constrained behavior,” for eating less meat and more grains, for job-sharing and more leisure time, for lives that are “less pressured and more centered on friends, family, and activities of inherent value and fuller dignity,” for frugality, and for imposing a global resources dividend—a resource tax to help eliminate poverty worldwide. An essay on consumption as culture urges protecting the “huge number of very poor peasants in

Click to return to Table of Contents 190 PDR 26(1) B OOK REVIEWS

Western Europe” from succumbing to “the convenience of American fast food and cheap agriculture.” The papers emerge from a project of the Institute for Philoso- phy and Public Policy, University of Maryland. One-third of them have previously been published.—G.McN.

JAMES Z. LEE AND WANG FENG One Quarter of Humanity: Malthusian Mythology and Chinese Realities, 1700–2000 Cambridge, MA: Harvard University Press, 1999. xiii + 248 p. $47.50. The myth of the title of this study is the view of China as an impoverished subsistence economy, held at a low-level Malthusian equilibrium, its performance and prospects always dimmed by overpopulation—a view that some observers have barely revised even after several decades of fertility decline and rapid economic growth. The myth is seen as having justified the government’s intensive family planning program, with its success credited to “the autocratic legacy of the Chi- nese state.” The realities, according to the authors, include centuries of slow gains in food production, height increases and mortality improvements at least through the twentieth century, and, within marriage, a widespread pattern of “late starting, early stopping, and long spacing” sufficient to yield 2–3 fewer births than was typical for preindustrial European couples. The Chinese demographic regime is depicted as one of universal female marriage but low marital fertility, moderate overall mortality but (in the past) high female infanticide, and frequent recourse to both male and female adoption. The authors stress the continuity of collective influence on fertility decisions, translated since the 1970s into state policy. “For the Chinese public,” they write, “the broadly accepted and deeply believed goals of the family planning program legitimate the use of state and public coercion.” The sections of the book discussing China’s demographic system made up the core of the article by the authors that appeared in PDR 25(1). Appendix on Chinese population sources, bibliography, index.

JONATHAN M. MANN AND DANIEL J. M. TARANTOLA (EDS.) AIDS in the World II: Global Dimensions, Social Roots, and Responses New York: Oxford University Press, 1996. xxxiv + 616 p. $34.50. By the close of the twentieth century the HIV/AIDS epidemic had reached all cor- ners of the globe, and infections continued to spread rapidly despite substantial preventive efforts. In 1999, the number of deaths from AIDS reached an estimated 2.6 million (5 percent of the total global death toll) and 5.6 million new HIV infections occurred. By the end of 1999, 33.6 million individuals were living with HIV/ AIDS, the large majority of whom are expected to die from the disease. The situation is far worse in sub-Saharan Africa than in other continents. In a few countries in South and East Africa, more than one in five adults are infected. In other continents infection levels among adults are typically still below one percent, but

Click to return to Table of Contents B OOK REVIEWS PDR 26(1) 191 the upward trend in some countries, particularly in India, is a cause of major concern. Although somewhat dated, this volume, a follow-up to the first edition published in 1992, gives a comprehensive overview of the various dimensions of the epidemic. The 41 chapters, written by 125 authors, cover a wide range of topics, including the status of the epidemic, its socioeconomic impact, new knowledge on the causes of the disease and its treatment, and the individual and institutional responses. The main message from the volume is that the epidemic represents an unprecedented threat to the health and development of many societies and that the national and international responses to this threat remain highly inadequate. One of the most interesting issues raised in this edition is the possibility of a new and potentially larger second wave of the epidemic in regions such as North America, Latin America, and Western Europe, where the prevalence of HIV infection seems to have stabilized at relatively low levels. It is argued that after having rapidly infected the groups at highest risk, the epidemic is now poised to spread at a slower pace into the much larger general population of low-risk individuals. Al- though largely based on computer simulations and not widely accepted, this wor- risome scenario should be considered a possibility. It bolsters an already powerful argument for strong efforts to halt the further spread of the epidemic in all regions of the world. Several appendixes provide detailed statistical information on past trends and the status of the epidemic up to 1995.—J.B.

JULIAN L. SIMON The Economic Consequences of Immigration, Second Edition Ann Arbor: University of Michigan Press, 1999. xxxiv + 434 p. $59.50; $24.95 (pbk.). The second edition of the late Julian Simon’s treatise on immigration, originally published by Blackwell in 1989, appears under a new imprint. The text is essentially the same, but material from a 1995 Cato Institute report by Simon is ap- pended to several chapters and a scattering of 1990s references is added. The book’s conclusions can be read as supporting much higher rates of immigration to the United States and more attention to economic characteristics than to family reunion. He wrote: “The negative consequences of any level of immigration which is politically imaginable at present are at most speculative, rather than documented. Therefore, a policy which is both prudent and also consistent with these observations would be to increase immigration quotas in a series of increments of significant size—perhaps half a percent, or one percent, of total population at each step— to check on any unexpected negative consequences, and to determine whether demand for admission ever exceeds the supply of places.” Given recent net annual immigration rates of around 0.3 percent of total US population, these experimen- tal “significant increments” correspond to a doubling or quadrupling of admission numbers. (“Literary prudence,” he said, dictated that he not write about unlimited immigration.) There was a case, however, for selecting migrants by education and assets, for running a guestworker program, and, better still, for selling foreigners the right to immigrate. Simon as scholar and as libertarian provocateur become entangled.

Click to return to Table of Contents 192 PDR 26(1) B OOK REVIEWS

Research on the economics of immigration expanded considerably in the 1990s, with results much more qualified than Simon’s. In particular, it finds small effects of immigration on per capita GDP and appreciable negative effects on lower-skilled workers. Much of this research is brought together in the recent National Research Council’s report, The New Americans, and in the background papers prepared for it (see PDR 24, no. 1, pp. 166–169 and 24, no. 3, pp. 636–637). See also George J. Borjas’s Heaven’s Door, reviewed in this issue. Bibliography, index.—G.McN.

SOPEMI Trends in International Migration: Continuous Reporting System on Migration. Annual Report, 1999 Edition Paris: OECD, 1999. 328 p. $56.00. Many aspects of international migration in OECD countries have changed markedly in the 1990s, as Trends in International Migration describes in rich detail. Fol- lowing a decline in immigration flows since 1993, immigration levels began to rise in several member countries in 1997. The recent increase in immigration is unre- lated to Asia’s financial crisis—which has not led to sizable outmigration from that region—but does stem in part from the diversification of countries of origin elsewhere, with migration occurring over longer distances. Whereas most inflows are still motivated mainly by family reunion—particularly in the United States, Canada, and France—increasing proportions of entrants are highly skilled workers filling specialized employment niches, or are temporary laborers. Several countries of the European Union (EU) experienced an upsurge of asylum seekers in 1998, linked to ethnic conflict in Kosovo, economic collapse in Albania, and an outflow of Kurds from Iraq and Turkey. And in contrast to the United States and Canada, population growth in the EU is increasingly attributable to immigration rather than natural increase (the relative contribution of these demographic components of growth has remained stable in North America since the 1960s). The increasing number of asylum seekers in the 1990s has prompted the EU to speed up the processing of applications (hence denials) and to extend restrictive visa requirements to cover a larger number of countries. The potential of immigration to dampen population aging in much of Europe, however, has not yet generated much deliberation by policymakers, nor has the OECD issued recommendations on this subject, despite extensive coverage of the issue in the media and in United Nations reports. This book rightly notes that increased inflows of foreigners and births to immigrants will not necessarily act as a brake on population aging in low-fertility countries. Such an effect would require a continuing succession of migration waves and substantially higher fertility among immigrants than nationals. Yet, persistent high immigration in the future should not be presumed, given the abysmal unemployment levels among immigrants—three times higher, for instance, than among nationals in Denmark and the Netherlands— that may discourage uprooting; nor is comparatively high fertility among immigrants likely to persist, given the typical assimilation of newcomers to local reproductive norms within a few years of arrival. At present, there is a preference in

Click to return to Table of Contents B OOK REVIEWS PDR 26(1) 193 most countries for mechanisms that cater to demand for temporary labor and not for policies calling for permanent immigration. In light of these circumstances, one may conclude from this book that the prospects are not good that immigration from less developed countries will sustain population sizes in countries with below-replacement fertility.—M.B.

JACQUELINE STEVENS Reproducing the State Princeton: Princeton University Press, 1999. xv + 307 p. $49.50; $18.95 (pbk.). Political science manages to have virtually nothing to say about fertility. Jacqueline Stevens terms it Locke’s “stunning rhetorical victory,” in his Two Treatises of Civil Government, to have convinced people of all stripes that membership of political societies, in contrast to membership of families, requires consent, and thus calls for a distinct analytical strategy making no use of family analogies. The reality, she argues, is just the opposite: political societies acquire members chiefly by birth. Reproduction, it might be said, is the state’s way of reproducing itself—in the author’s words, “political societies develop birth practices so as to provide a connection between the current population and those of the past and future.” In this densely argued, provocative essay, Stevens unravels the implications of making birth the criterion for membership of a society and challenges those theorists who depict the modern state as distanced from primordial views of a nation based on ancestral ties. She examines the ways that states organize themselves around the ostensibly natural categories of gender, ethnicity, generation, and kinship, drawing instances from a wide array of historical and contemporary societies. Her own radical stance is one that would do away with “political societies based on kinship forms,” or, phrased slightly more modestly in the conclusion, that would seek to find ways of reproducing political societies in which “kinship principles would play a diminished role.” Her proposals are “elimination of any state involvement in marriage and the curtailment of citizenship requirements based on birth or ances- try.” The author is a political scientist at the University of Michigan. Bibliography, index.—G.McN.

UNITED NATIONS DEVELOPMENT PROGRAMME, CHINA The China Human Development Report New York: Oxford University Press, 1999. vi + 135 p. $45.00; $24.95 (pbk.). Although the United Nations Development Programme (UNDP) has been publishing its Human Development Report annually since 1990, this volume is the first such endeavor to review and measure human and social development focusing specifically on the People’s Republic of China. As a follow-up to the Copenhagen Decla- ration and Programme of Action promulgated at the World Summit on Social De- velopment in 1995, UNDP’s China Country Office organized the production of this Report. It was prepared by a consultant group of 11 international and Chinese

Click to return to Table of Contents 194 PDR 26(1) B OOK REVIEWS specialists, most of them with background in economics. The primary focus is on poverty alleviation, but the Report also touches on such human/social development issues as income, health care, education and nutrition, population and migration, gender equality, social security, and natural environment. For the first time human and social development is quantified and ranked according to UNDP’s Human Development Index (HDI) at the provincial level in China. Population issues are briefly discussed as background, for the Report is mainly intended to show the correlations of human/social development with economic growth rather than with demographic changes. But with more than 50 tables and graphs, the Report is informative and should be viewed as a rich resource for readers interested in measures of human and social development in China. Bibliography, statistical appendix.—B.G.

UNITED NATIONS, ECONOMIC COMMISSION FOR EUROPE Economic Survey of Europe 1999, No. 1 Geneva, 1999. xi + 222 p. $70.00. Although largely concerned with Europe’s economies—particularly, in recent years, with the problems of the “economies in transition”—this ECE series occasionally looks at demographic matters. Issue Number 1 of 1999 contains a chapter entitled “Fertility decline in the transition economies, 1982–1997: Political, economic and social factors.” The explicandum is the “grossly depressed fertility” of these countries, with 1997 TFRs averaging 1.35 (down from 2.07 in 1982). Bulgaria (1.09) and Latvia (1.11) have become the lowest-fertility countries, below Spain (1.15)— though not below the former East Germany (0.95; it is described as having “re- bounded” from a trough of 0.76 in 1993). The causes of the decline are discussed under three headings: political change, associated with heightened anxiety about the future; economic downturn, eroding living standards and social support systems; and the spread of Western family patterns and reproductive behavior, as seen especially in the fall-off in marriage and rise in age at first birth. It is argued that these factors operate to different degrees across the region, with Central Eu- rope and the Baltic states more affected by pressures for convergence with the West, the European CIS countries by economic deterioration. An annual publication until 1997, the Economic Survey of Europe now appears three times per year. Statistical appendix.—G.McN.

KATHLEEN S. UNO Passages to Modernity: Motherhood, Childhood, and Social Reform in Early Twentieth Century Japan Honolulu: University of Hawaii Press, 1999. x + 237 p. $24.95. In this slim study, Kathleen Uno, a historian at Temple University, explores the growth and expansion of daycare centers in pre–World War II Japan. Given the foreign origins of such institutions, their growth in Japan seemed paradoxical in

Click to return to Table of Contents

B OOK REVIEWS PDR 26(1) 195

an intensely nationalistic time. And they presented a striking contrast to the “education mother” of today’s Japan, who shuns work outside the home to devote full time to ensuring the academic success of her child. To Uno, the daycare centers presented puzzles worth historical attention, in particular because unraveling them promised to shed light on changes in Japanese attitudes toward motherhood, childhood, and childrearing. Her research in the archives of two major centers high- lighted the importance of two factors in their rise. First, the style of childcare implied by the centers was congruent with nineteenth-century childrearing patterns, which assigned daily care of the young not only to the birth mother, but also to kin and nonkin members of the household. Second and more important, the expansion of the daycare centers was linked to a deep-seated, “almost desperate” drive for national progress. Japanese leaders and social reformers realized that the construction of a strong nation required the creation of citizen-subjects who could meet the challenges of modernity as patriotic colonists, conscripts, and workers. National salvation required both the molding of a new citizenry suited to the challenges of industrial development and imperial expansion, and the amelioration of social problems stemming from urban and industrial development. The daycare centers were ideally suited to these tasks. In this context, the Western idea that women could contribute to national wellbeing through their labors not in the fields, but in the home, had strong appeal. In creating a national citizenry, the Japanese state and social reformers also remade conceptions of male and female, old and young. Just as women were turned from producer-reproducers into full-time reproducers, children were transformed from laborers for the household and local community into dependents oriented toward the nation-state. While Uno’s intriguing findings and interpretations will be of interest to students of Japanese social history, the repetitiousness of the text and the absence of comparative perspective may well limit the book’s appeal to other readers. Bibliography, index.— S.G.

Click to print book reviews Click to return to Table of Contents Click to return to Table of Contents

DOCUMENTS

The Census Bureau on Prospects for US Population Growth in the Twenty-First Century

The U.S. Census Bureau periodically releases projections of the US resident population, detailed by age, sex, race, and Hispanic origin. The most recent of these, issued 13 January 2000, for the first time extend to the year 2100 and also include information on the foreign- born population. (Earlier projections were carried up to 2080.) The extensive tabulations presenting the new set, and detailed explanation of the methodology and the assumptions underlying the projections, are accessible at the Census Bureau’s web site: http://www.census.gov. A brief summary of some of the main results of these projections is reproduced below from United States Department of Commerce News, Washington, DC 20230. (The Census Bureau is an agency of the Department of Commerce.) Uncertainties as to future trends in fertility, mortality, and net migration over a period of some 100 years are very great, as is illustrated by the massive difference in the projected size of the population for 2100 in the three variants produced. The “middle” projected population figure of 571 million (which represents a growth of some 109 percent over its current level) is bracketed by “lowest” and “highest” alternative projections of 283 million and 1.18 billion, respectively. With somewhat lesser force, the point also applies to the 50- year time span considered in the well-known country-by-country projections of the United Nations. These projections are also detailed in three variants: low, middle, and high. The UN projections (last revised in 1998) envisage less rapid growth in the United States during the first part of the twenty-first century than do the Census Bureau’s. The projected population figures for 2050 in the three variants (low, middle, and high) are as follows (in millions):

U.S. Census Bureau 313.5 403.7 552.8 United Nations 292.8 349.3 419.0

Since the initial age and sex distributions from which the two sets of population projections start are essentially identical, these differences reflect assumptions by the Census Bu- reau with respect to the three factors affecting population dynamics in the next 50 years. In the middle series, each of these assumptions is more growth-producing in the Census Bureau’s set than in that of the United Nations. Thus, in the middle of the twenty-first century the Census Bureau anticipates male and female life expectancies of 81.2 and 86.7 years; the corresponding figures according to the UN are 78.8 and 84.4 years. Net immigration to the United States per 1000 population at midcentury is assumed to be 2.2 by the United Nations

POPULATION AND DEVELOPMENT REVIEW 26(1) (MARCH 2000) 197

Click to print article Click to return to Table of Contents 198 D OCUMENTS and somewhat above 2.4 according to the Census Bureau. The factor most affecting the difference between the projected population sizes, however, is the differing assumptions with respect to fertility. The middle UN series anticipates a midcentury US total fertility rate of 1.9 children per woman; the Census Bureau’s assumption is slightly above 2.2. A notable feature of the Census Bureau’s projection methodology in comparison to that of the UN is the recognition of differences in mortality and fertility, and also in immigration, with respect to race and Hispanic origin. Thus, at midcentury the white non-His- panic population is assumed to have a total fertility rate of 2.03; the corresponding figure for the population of Hispanic origin is 2.56. (Fertility in other population subgroups is expected to lie between these values, although closer to the fertility of non-Hispanic whites.) And Hispanic immigration, currently the major component within total immigration, is assumed to remain significant throughout the next five decades (although by midcentury it is expected to be far exceeded by immigration of non-Hispanic Asians). As a result, the structure of the US population by race and Hispanic origin is expected to shift markedly. To the extent that fertility and mortality differentials persist, such a shift also affects the mean fertility and mortality figures of the total population.

The nation’s resident population, 273 million The data also show lowest and highest on July 1, 1999, is projected to reach 404 alternative projections. The lowest series million in 2050 and 571 million in 2100. projects population growth to 314 million in These results are based on middle-level as- 2050 and then a decline to 283 million in sumptions regarding population growth dur- 2100. The highest projects 553 million people ing the century. in 2050 and 1.2 billion in 2100. Even though childbearing levels in the The projections do not take into account United States remain quite close to the level possible future changes in the way people needed only to replace the population, the report their race and ethnicity and, because increasing number of potential parents and of the length of time covered and other un- continued migration from abroad would be certainties, they are considered less reliable sufficient to add nearly 300 million people for the latter part of the century. during the next century. Because the His- According to the middle series projec- panic and Asian and Pacific Islander populations, the Hispanic population (of any race) tions in the U.S. are younger than the na- would triple from 31.4 million in 1999 to tion as a whole and because they continue 98.2 million in 2050. By 2005, Hispanics may to receive international migrants, these popu- become the nation’s largest minority group. lations will become increasingly prominent. The percentage of Hispanics in the total population could rise from 12 percent in 1999 to 24 percent in 2050. The Asian and Pacific Islander popula- TABLE 1 Total U.S. resident population (millions): Middle, lowest tion, meanwhile, would more than triple, and highest series, 1999 to 2100 from 10.9 million in 1999 to 37.6 million in 2050. Its percentage of the total population Year Middle Lowest Highest would rise from 4 percent now to 9 percent 1999 272.8 272.7 273.0 in 2050. 2025 337.8 308.2 380.4 According to the projections, the non- 2050 403.7 313.5 552.8 Hispanic White and African American populations would increase more slowly than the 2075 480.5 304.0 809.2 other groups. The non-Hispanic White popu- 2100 571.0 282.7 1,182.4 lation would rise from 196.1 million in 1999

Click to return to Table of Contents D OCUMENTS 199

TABLE 2 U.S. resident total population (millions), percent distribution, and change in absolute and percentage terms, by race and Hispanic origin, middle series, 1999 and 2050 Race and Hispanic origin 1999 2050 1999 to 2050

Total population 272.8 100.0 403.7 100.0 130.9 48.0 White 224.7 82.3 302.5 74.9 77.8 34.6 Black 34.9 12.8 59.2 14.7 24.4 69.8 American Indian, Eskimo and Aleut 2.4 0.9 4.4 1.1 2.0 83.8 Asian and Pacific Islanders 10.9 4.0 37.6 9.3 26.7 245.2

Total population 272.8 100.0 403.7 100.0 130.9 48.0 Hispanic (of any race) 31.4 11.5 98.2 24.3 66.9 213.3 Non-Hispanic White 196.1 71.9 213.0 52.8 16.9 8.6 Black 33.1 12.1 53.5 13.2 20.4 61.5 American Indian, Eskimo and Aleut 2.0 0.7 3.2 0.8 1.2 60.2 Asian and Pacific Islanders 10.3 3.8 35.8 8.9 25.5 248.8

to 213.0 million in 2050, a 9 percent increase. The population age 65 and over would Its share of the total population would de- grow from 34.6 million in 1999 to 82.0 mil- cline, however, from 72 percent in 1999 to lion in 2050, a 137 percent increase. The pro- 53 percent in 2050. jections also show an especially rapid surge in The African American population, ac- the elderly population as the surviving “baby cording to the projections, would rise from boomers” pass age 65; in the year 2011, baby 34.9 million in 1999 to 59.2 million in 2050 boomers (those born between 1946 and 1964) a 70-percent increase; under this scenario, will begin turning 65. Between 2011 and 2030, the African American share of the total popu- the number of elderly would rise from 40.4 lation would increase slightly, from 13 per- million (13 percent of the population) to 70.3 cent to 15 percent. million (20 percent of the population). Between 1999 and 2050, the total num- The projections show that the number of ber of foreign-born would more than double, children under 18 would increase from 70.2 increasing from 26.0 million to 53.8 million. million in 1999 to 95.7 million in 2050. How- The proportion of the nation’s population ever, their share of the nation’s population that is foreign-born may rise from 10 percent would decline slowly, falling from 26 percent in 1999 to 13 percent in 2050. in 1999 to 24 percent in 2050.

TABLE 3 U.S. resident total population (millions), percent distribution, and change in absolute and percentage terms, by age, middle series, 1999 and 2050 Age (years) 1999 2050 1999 to 2050

Total population 272.8 100.0 403.7 100.0 130.9 48.0 Under 18 70.2 25.7 95.7 23.7 25.6 36.5 18–64 168.1 61.6 225.9 56.0 57.8 34.4 65 and over 34.6 12.7 82.0 20.3 47.4 137.2 Median age 35.5 (X) 38.8 (X) (X) (X)

(X) Not applicable. SOURCE: For each of the tables: Population Projections Program, Population Division, U.S. Census Bureau, Washington, D.C. 20233.

Click to return to Table of Contents

200 D OCUMENTS

The projections are based on assumptions migration is assumed to vary over time and about future childbearing, mortality and mi- decrease generally relative to the size of the gration. The level of childbearing among population. Assumptions for the lowest and women for the middle series is assumed to highest series are summarized in a working remain close to present levels, with differ- paper, titled Methodology and Assumptions for ences by race and Hispanic origin diminish- the Population Projections of the United States: ing over time. Mortality is assumed to de- 1999 to 2100. [Population Division Working cline gradually with less variation by race and Paper No. 38, U. S. Census Bureau, 13 Janu- Hispanic origin than at present. International ary 2000.]

Click to print article Click to return to Table of Contents ABSTRACTS

The US Decennial Census: Political that these neo-Malthusian deforestation nar- Questions, Scientific Answers ratives badly misrepresent people–forest relationships. They obscure important nonlin- KENNETH PREWITT ear dynamics, as well as widespread anthropogenic forest expansion and landscape The US decennial census was initiated in enrichment. These processes are better cap- 1790 to facilitate nation-building tasks, es- tured, in broad terms, by a neo-Boserupian pecially that of reconfiguring political rep- perspective on population–forest dynamics. resentation as the population grew and However, comprehending variations in lo- settled new territories. To this basic task of cale-specific trajectories of change requires power distribution have been added other fuller appreciation of social differences in key governmental functions, such as the use environmental and resource values, of how of census data in guiding revenue sharing diverse institutions shape resource access and and in the enforcement of nondiscriminatory control, and of ecological variability and path policies. Throughout its history the census dependency in how landscapes respond to has been the focus of partisan clashes. Fol- use. The second half of the article presents lowing the identification of the “differential and illustrates such a “landscape structura- undercount”—a measure of how census cov- tion” perspective through case studies from erage differs across demographic groups and the forest–savanna transition zones of Ghana geographic areas—the partisan battles inten- and Guinea. sified, and in recent decades have come to focus not just on how the census counts are used but how the census data are collected. It has been argued that census methodology Reproduction, Compositional could be designed to predetermine given par- Demography, and Economic tisan outcomes, and for the 2000 census this Growth: Family Planning in charge shifted from “could be” to “is.” The England Long Before the Fertility Census Bureau has taken extraordinary steps Decline to demonstrate that no partisan considerations have affected the design or implemen- SIMON SZRETER tation of the census, and that its decisions EILIDH GARRETT are based solely on the best technical judgment available. This article offers a radical reinterpretation of the chronology of control over reproduction in England’s history. It argues that, as a result of post–World War II policy preoccu- Challenging Neo-Malthusian pations, there has been too narrow a focus Deforestation Analyses in West in the literature on the significance of reduc- Africa’s Dynamic Forest Landscapes tions in marital fertility. In England’s case this is conventionally dated to have occurred from 1876, long after the industrial revolu- MELISSA LEACH tion. With a wider-angle focus on “reproduc- JAMES FAIRHEAD tion,” the historical evidence for England in- Many influential analyses of West Africa take dicates that family planning began much it for granted that “original” forest cover has earlier in the process of economic growth. progressively been converted and savannized Using a “compositional demography” ap- during the twentieth century by growing proach, a novel social pattern of highly pru- populations. By testing these assumptions dential, late marriage can be seen emerging against historical evidence, exemplified for among the bourgeoisie in the course of the Ghana and Ivory Coast, this article shows eighteenth and early nineteenth centuries.

POPULATION AND DEVELOPMENT REVIEW 26(1) (MARCH 2000) 201

Click to return to Table of Contents 202 A BSTRACTS

There is also evidence for a more widespread 1997 and over 25 percent in two countries resort to such prudential marriage through- of Southern Africa. HIV/AIDS infection is not out the population after 1816. When placed the result of ignorance, as nearly everyone in this context, the reduction in national fer- has sufficient knowledge about AIDS and tility indexes visible from 1876 can be seen how it is transmitted. The high levels of AIDS as only a further phase, not a revolution, in arise from the failure of African political and the population’s management of its repro- religious leaders to recognize social and duction. sexual reality. The means for containing and conquering the epidemic are already known, and could prove effective if the leadership Absent and Problematic Men: could be induced to adopt them. The lack of individual behavioral change and of the Demographic Accounts of Male implementation of effective government Reproductive Roles policy has roots in attitudes to death and a silence about the epidemic arising from be- MARGARET E. GREENE liefs about its nature and the timing of death. ANN E. BIDDLECOM International responsibility may have to be taken before the needed effective policies are Both men and women are important actors put in place. in bringing children into life, yet demographic studies on reproduction have tended to focus on women alone. The aims of this article are: 1) to describe why men have at- Long-Term Population Projections tracted limited interest as subjects of such and the US Social Security System research; 2) to evaluate existing research on men’s roles in developing countries; and 3) RONALD LEE to suggest directions for future research on male reproductive roles. Men, once ne- According to a report recently issued by the glected, are now included in research on fer- Technical Panel for the US Social Security tility but from a narrow, overly problem-ori- Administration, the long-term financial out- ented perspective. A review of the literature, look for the system is worse than previously however, raises questions about the ad- thought. The worsening projected by the equacy of a problem-oriented approach. The panel in the long-run funding imbalance of authors argue that demography should fo- the Social Security System is mostly due to cus on men not only as women’s partners, the recommendation by the panel to add an but also as individuals with distinct repro- extra four years to the currently projected ductive histories. In situations, now increas- increase in life expectancy by 2075: from ingly common, where the links between 81.8 years to 85.9 years. The panel recom- marriage and childbearing erode, the differ- mended no change in the current interme- ences in men’s and women’s reproductive diate projected long-run TFR of 1.9 and net experiences and the costs and benefits of immigration of 900,000 persons per year. parenting will become more salient for fu- The recommendation to increase the pro- ture research. jected gains in life expectancy was based on international trends as well as on historical trends in the United States and the absence of biological evidence ruling out such gains. Rethinking the African AIDS Industrial countries have a history of under- Epidemic predicting the growth of their elderly population, and it is expected that large mortal- JOHN C. CALDWELL ity adjustments may be needed in the projections for public pension programs also Half the AIDS victims in the world are in East in industrial countries other than the United and Southern Africa, where adult HIV sero- States. prevalence was 11.4 percent by the end of

Click to return to Table of Contents A BSTRACTS / FRENCH 203

Le recensement décennal aux États- « originale » a progressivement été conver- Unis : réponses scientifiques à des tie en savane au cours du vingtième siècle questions politiques par les populations en état de croissance. En vérifiant ces hypothèses en regard des preu- ves antérieures recueillies au Ghana et en KENNETH PREWITT Côte d’Ivoire, le présent article indique que Aux États-Unis, le recensement décennal a ces exposés de faits néomalthusiens présen- été adopté en 1790 pour favoriser les tâches tent de manière inexacte la relation entre la nécessaires au développement du pays, no- population et la forêt. Ils masquent l’impor- tamment celle de reconfigurer la représen- tante dynamique non linéaire ainsi que tation politique au fur et à mesure de la crois- l’élargissement généralisé de la forêt et l’en- sance démographique et du peuplement de richissement du paysage d’origines anthro- nouveaux territoires. À cette tâche fonda- piques. En termes génériques, ces méthodes mentale de distribution du pouvoir, on a sont mieux captées par une perspective ajouté d’autres fonctions gouvernementales néobosérupienne sur la dynamique popula- clés, comme l’utilisation des résultats de re- tion–forêt. Toutefois, la compréhension des censement pour guider le partage des recet- variations sur les trajectoires locaux-spécifi- tes et appliquer les politiques non discrimi- ques du changement appelle une meilleure natoires. Depuis son adoption, le recense- appréciation des différences sociales dans les ment décennal a fait l’objet de batailles valeurs environnementales et les ressources: partisanes. Suite à l’identification du « sous- comment les diverses institutions façonnent- dénombrement différentiel »—élément qui elles l’accès aux ressources et leur contrôle ? sert à mesurer de quelle façon le champ d’ob- comment les paysages réagissent-ils à l’usage servation du recensement diffère selon les dans les contextes de variabilité écologique groupes démographiques et les aires géogra- et de dépendance de parcours ? La deuxième phiques—les batailles partisanes se sont in- partie du présent article présente et illustre tensifiées et, au cours des récentes décennies, cette perspective de «structuration de pay- ont été axées non seulement sur la façon sage» au moyen d’études de cas tirées des dont les chiffres du recensement sont utilisés zones de transition forêt–savanne du Ghana mais aussi sur la façon dont les données de et de la Guinée. recensement sont recueillies. Il a été allégué que la méthodologie de recensement pourrait être conçue de sorte que des résultats partisans soient décidés à l’avance; lors du recensement de l’an 2000, cette allégation a Reproduction, démographie passé de « pourrait » à « était ». Le Census compositionnelle et croissance Bureau a pris des mesures extraordinaires pour démontrer qu’aucune considération économique : planification familiale partisane n’avait influé sur la conception ou en Angleterre bien avant la baisse la mise en œuvre du recensement et que ses de la fécondité décisions n’étaient fondées que sur le meilleur jugement technique disponible. SIMON SZRETER EILIDH GARRETT Le présent article offre une réinterprétation Remise en question des analyses radicale de la chronologie du contrôle sur la néomalthusiennes sur le reproduction dans l’histoire de l’Angleterre. déboisement des forêts dynamiques Les auteurs allèguent que, en raison des pré- d’Afrique occidentale occupations liées aux politiques d’après- guerre, la documentation a eu un regard très étroit sur la signification de la réduction de MELISSA LEACH la fécondité des mariages. En ce qui concerne JAMES FAIRHEAD l’Angleterre, cette réduction a vraisemblable- En Afrique occidentale, de nombreuses ana- ment eu lieu à partir de 1876, bien après la lyses importantes présument que la forêt révolution industrielle. Si l’on porte un re-

Click to return to Table of Contents 204 A BSTRACTS / FRENCH gard plus large sur la «reproduction», l’his- mes, ainsi que les coûts et les avantages du toire démontre qu’en Angleterre la planifi- parentage deviendront de plus en plus ap- cation familiale a débuté beaucoup plus tôt parents. dans le processus de croissance économique. À l’aide de la méthode de «démographie compositionnelle», une nouvelle structure sociale d’un mariage tardif et prudentiel est Repenser l’épidémie du sida en observée chez la bourgeoisie au cours du dix- Afrique huitième siècle et au début du dix-neuvième. On observe également un recours plus ré- JOHN C. CALDWELL pandu au mariage prudentiel dans toute la population après 1816. Dans ce contexte, la La moitié des victimes du sida dans le monde réduction du taux de fécondité à l’échelon résident en Afrique orientale et australe. À national à partir de 1876 ne peut être inter- la fin de 1997, la séroprévalence chez les prétée que comme une autre phase—et non adultes se chiffrait à 11,4 % et à 25 % dans une révolution—dans la gestion de la popu- deux pays d’Afrique australe. L’infection du lation sur sa reproduction. VIH/sida n’est pas due à l’ignorance car à peu près tout à chacun possède des connaissan- ces suffisantes sur le sida et ses modes de transmission. Les taux élevés du sida sont imputables aux dirigeants religieux et poli- Problématique de l’homme absent : tiques africains qui n’ont pas su reconnaître analyse démographique du rôle de la réalité sociale et sexuelle. Les moyens de l’homme dans la reproduction contenir et d’enrayer l’épidémie sont déjà connus et pourraient s’avérer efficaces si les MARGARET E. GREENE dirigeants s’appliquaient à les adopter. L’ab- ANN E. BIDDLECOM sence du changement de comportement in- dividuel et la non-application d’une politi- Les hommes aussi bien que les femmes ont que gouvernementale efficace sont le reflet un rôle important à jouer dans la procréa- des attitudes qui prévalent sur la mort et du tion. Or, les études démographiques sur la silence sur la nature de l’épidémie et le mo- reproduction ont eu tendance à ne s’intéres- ment de la mort. La responsabilité de met- ser qu’aux femmes. Les objectifs de la pré- tre en œuvre les politiques efficaces néces- sente étude sont : 1) de décrire les raisons saires pourrait fort bien devoir provenir des pour lesquelles les hommes ne font l’objet instances internationales. que d’un intérêt limité dans ce genre de recherche; 2) d’évaluer la recherche actuelle sur les rôles des hommes dans les pays en voie de développement; et 3) de suggérer des directions pour la recherche future sur les Prévisions démographiques à long rôles des hommes dans la reproduction. terme et le système de sécurité Ignorés auparavant, les hommes sont main- sociale aux États-Unis tenant considérés dans la recherche sur la fécondité, quoique d’un point de vue étroit RONALD LEE et trop utilitaire. Toutefois, une analyse bi- bliographique soulève des questions sur la Selon un rapport récent émis par le Comité pertinence d’une approche utilitaire. Les consultatif technique de la US Social Security auteurs allèguent que la démographie de- Administration, les perspectives financières à vrait être axée sur les hommes non seule- long terme pour le système sont pires que ment en tant que partenaires de la femme, prévues. L’aggravation prévue par le Comité mais aussi en tant qu’individus possédant des dans le déséquilibre du financement à long antécédents reproductifs distincts. Dans les terme du système de sécurité sociale est situations de plus en plus fréquentes où les attribuable surtout à la recommandation liens entre le mariage et la procréation se faite par le Comité d’ajouter une période ad- dégradent, les différences dans les antécé- ditionnelle de quatre ans à l’augmentation dents reproductifs des hommes et des fem- actuelle prévue de l’espérance de vie d’ici à

Click to return to Table of Contents A BSTRACTS / SPANISH 205

2075 : de 81,8 ans à 85,9 ans. Le Comité n’a ainsi que sur l’absence de preuve biologique recommandé aucun changement dans l’in- réfutant de tels gains. Les pays industrialisés dice synthétique de fécondité actuel de 1,9 sont connus pour «sous-prédire» la crois- prévu à long terme et à l’immigration nette sance de leur population âgée et l’on s’at- de 900,000 personnes par année. Cette re- tend à ce que des ajustements importants des commandation d’accroître les gains projetés taux de mortalité soient également nécessai- dans l’espérance de vie a été fondée sur les res dans les prévisions de programmes de tendances internationales et les tendances pension gouvernementale dans les pays in- enregistrées antérieurement aux États-Unis dustrialisés autres que les États-Unis.

El Censo Decenal de los Estados Un reto al análisis neo-maltusiano Unidos: Asuntos políticos, de deforestación en el dinámico respuestas científicas paisaje selvático del Africa Occidental KENNETH PREWITT MELISSA LEACH El censo decenal de los Estados Unidos se ini- JAMES FAIRHEAD ció en 1790 para facilitar las tareas de esta- blecer la nación, especialmente la de la Muchos análisis influyentes de Africa Occi- reconfiguración de la representación políti- dental dan por supuesto que la selva “origi- ca al aumentar la población y poblarse nue- nal” ha sido progresivamente convertida y vos territorios. A esta tarea básica de distri- transformada en sabana durante el siglo bución del poder se han agregado otras veinte por el incremento de la población. Al funciones gubernamentales claves, tales examinar estas suposiciones frente a la evi- como la utilización de datos de censo como dencia histórica, ilustrada por Ghana y la guía para la distribución de ingresos y en la Costa de Marfil, este artículo demuestra que aplicación de políticas no discriminatorias. El estas narrativas neo-maltusianas de censo a través de su historia ha sido foco de deforestación representan muy falsamente conflictos partidistas. Luego de identificarse las relaciones de población–selva. Ellas el “subrecuento diferencial”—una medida de obscurecen importantes dinámicas no cómo difiere la cobertura del censo a través lineares como también la extensa expansión de grupos demográficos y áreas geográficas selvática antropogénica y el enriquecimien- —las batallas partidistas se intensificaron y to del paisaje. En amplios términos, estos en las décadas recientes no solamente se han procesos pueden ser mejor captados por una centrado en cómo los recuentos censales son perspectiva neo-Boserupiana sobre las diná- usados pero en cómo se han recolectado los micas de población–selva. Sin embargo, com- datos de censo. Se ha sostenido que la me- prendiendo las variaciones en las trayecto- todología censal podría ser diseñada para rias de cambio específicos al lugar requiere pre-decidir resultados partidistas dados y, una apreciación más amplia de las diferen- para el censo del año 2000, esta acusación cias sociales en cuanto a los valores de me- mudó de un “podría ser” a un “fue”. La Di- dio ambiente y de recursos; de cómo diver- rección del Censo ha tomado medidas ex- sas instituciones determinan el acceso a y traordinarias para demostrar que ninguna control de los recursos; y cómo los paisajes consideración partidista ha afectado el dise- responden al uso según de la variabilidad ño o la ejecución del censo, y que sus deci- ecológica y la dependencia de ruta. La se- siones se han basado únicamente en las me- gunda parte del artículo presenta e ilustra tal jores opiniones técnicas existentes. perspectiva de “estructuración del paisaje” a

Click to return to Table of Contents 206 A BSTRACTS / SPANISH través de estudios de caso de las zonas de embargo la tendencia en los estudios demo- transición de selva–sabana de Ghana y Gui- gráficos ha sido enfocar a la mujer solamen- nea. te. Los objetivos de este artículo son: 1) des- cribir por qué los hombres han atraído un interés limitado como sujetos de tal investi- Reproducción, demografía de gación; 2) evaluar las investigaciones que existen sobre el papel de los hombres en los composición, y crecimiento países en desarrollo; y 3) sugerir vías para económico: Planificación familiar en futuras investigaciones sobre el papel Inglaterra mucho antes de la reproductivo masculino. Los hombres, antes disminución de la fecundidad omitidos, son ahora incluidos en las investigaciones sobre fecundidad pero desde una SIMON SZRETER perspectiva estrecha orientada específica- EILIDH GARRETT mente hacia problemas. Un examen de la li- teratura en este campo sin embargo suscita Este artículo ofrece una re-interpretación ra- interrogantes acerca de cuán adecuado es el dical de la cronología del control sobre la re- enfoque orientado específicamente hacia producción en la historia de Inglaterra. Se problemas. Las autoras sostienen que la de- sostiene que, como resultado de preocupa- mografía debería enfocar al hombre no so- ciones de políticas de posguerra, las publica- lamente como pareja de la mujer pero tam- ciones han sido demasiado restringidas en su bién como individuo con una historia enfoque resaltando solamente el descenso de reproductiva propia bien determinada. En la fecundidad matrimonial. En el caso de In- situaciones, ahora cada vez más comunes, en glaterra, esto se data convencionalmente que los lazos entre el matrimonio y la pro- como habiendo ocurrido desde 1876, mucho creación se han erosionado, las diferencias más tarde que la revolución industrial. Con en las experiencias reproductivas de los hom- un ángulo de enfoque más amplio sobre “la bres y de las mujeres y los costos y benefi- reproducción”, la evidencia histórica para cios de la paternidad cobrarán cada vez ma- Inglaterra indica que la planificación fami- yor importancia en investigaciones futuras. liar comenzó mucho antes en el proceso de crecimiento económico. Usando un criterio basado en “la demografía de composición”, se vislumbra un patrón social ingenioso al- Reconsiderando la epidemia tamente prudente de matrimonios tardíos africana de SIDA que emerge entre la burguesía en el curso del siglo dieciocho y principios del diecinue- JOHN C. CALDWELL ve. Hay evidencia también de que se recu- La mitad de las víctimas de SIDA en el mun- rre en forma más extensa a tales matrimo- do se encuentran en el Africa Oriental y Me- nios prudentes a través de la población ridional donde la sero-prevalencia adulta de después de 1816. Cuando se coloca dentro VIH fue de un 11.4 por ciento a fines de 1997 de este contexto, la reducción en los índices y sobre un 25 por ciento en dos países del nacionales de fecundidad visibles desde 1876 Africa Meridional. Las infecciones de VIH/ pueden verse sólo como otra fase, no una SIDA no son resultado de ignorancia ya que revolución, en la manera que la población casi todas las personas tienen suficientes co- gestionaba su reproducción. nocimientos sobre el SIDA y como se trans- mite. Los altos niveles de SIDA han resultado más bien de la falta por parte de los líderes Hombres ausentes y problemáticos: africanos, tanto políticos como religiosos, de Recuentos demográficos del papel reconocer la realidad social y sexual. Los reproductivo masculino medios para contener y conquistar la epidemia ya son conocidos y podrían ser eficaces MARGARET E. GREENE si el liderazgo pudiera ser persuadido de su ANN E. BIDDLECOM adopción. La falta de cambios en el compor- tamiento individual y de la implantación de Tanto los hombres como las mujeres son ac- políticas gubernamentales eficaces tienen raí- tores importantes en dar vida a los hijos, sin ces en las actitudes hacia la muerte y un si-

Click to return to Table of Contents A BSTRACTS / SPANISH 207 lencio sobre la epidemia que surge de creen- recomendación hecha por el Grupo de agre- cias sobre su carácter y el momento en que gar cuatro años adicionales al aumento de se anticipa la muerte. Quizás habrá que asu- la esperanza de vida proyectado para el año mir una responsabilidad internacional antes 2075: esto es de 81.8 años a 85.9 años. El que las políticas eficaces que se requieren Grupo recomendó no hacer ningún cambio sean implantadas. en la proyección actual intermedia a largo plazo de la tasa de fecundidad global de 1.9 y una inmigración neta de 900,000 personas Proyecciones a largo plazo de la por año. La recomendación de aumentar las ganancias proyectadas en la esperanza de vida población y el Sistema de Seguro se basó en tendencias internacionales como Social de los Estados Unidos también en tendencias históricas en los Es- tados Unidos, y la ausencia de evidencias bio- RONALD LEE lógicas que descarten la posibilidad de tal ga- nancia. Los países industriales tienen De acuerdo a un informe recientemente he- antecedentes de sub-predecir el crecimiento cho público por el Grupo Consultor Técnico de su población anciana, y se estima que será de la Administración de Seguro Social de los necesario hacer grandes ajustes de la morta- Estados Unidos, la perspectiva financiera a lidad en las proyecciones para los programas largo plazo para el sistema es peor de lo pen- públicos de jubilación también en países in- sado anteriormente. El empeoramiento, pro- dustriales otros que los Estados Unidos. yectado por el Grupo, del desequilibrio a largo plazo en el financiamiento del sistema de Seguro Social se debe principalmente a la

Click to return to Table of Contents AUTHORS FOR THIS ISSUE

ANN E. BIDDLECOM is Research Investigator, Institute for Social Research, University of Michigan.

JOHN C. CALDWELL is Emeritus Professor of Demography, Australian National University, Canberra.

JAMES FAIRHEAD is Reader, Department of Anthropology, School of Oriental and African Studies, University of London.

EILIDH GARRETT is Senior Research Associate, Cambridge Group for the History of Population and Social Structure.

MARGARET E. GREENE is Senior Program Associate, Center for Health and Gender Equity, Takoma Park, Maryland.

MELISSA LEACH is Fellow, Environment Group, Institute of Development Studies, University of Sussex.

RONALD LEE is Professor of Demography and Economics, University of California at Berkeley.

KENNETH PREWITT is Director, U.S. Census Bureau.

VACLAV SMIL is Distinguished Professor, University of Manitoba, Winnipeg, Canada.

PAUL STREETEN is Professor Emeritus of Boston University.

SIMON SZRETER is University Lecturer in History and Fellow, St. John’s College, Cambridge. He is currently a Research Fellow of the UK Economic and Social Research Council.

208 POPULATION AND DEVELOPMENT REVIEW 26(1) (MARCH 2000)

Click to return to Table of Contents