THE WACO, TEXAS, ATF RAID and CHALLENGER LAUNCH DECISION Management, Judgment, and the Knowledge Analytic

Home , Rocco Petrone

ARPAGarrett / / March ATF RAID 2001 AND CHALLENGER DECISION THE WACO, TEXAS, ATF RAID AND CHALLENGER LAUNCH DECISION Management, Judgment, and the Knowledge Analytic

TERENCE M. GARRETT University of Texas Pan American

The author argues that the Challenger space shuttle launch disaster and the Bureau of Alcohol, Tobacco and Firearms (ATF) raid on the Branch Davidian compound both offer insights for managers and organization theorists as to how managers make judgments concerning their employees based on conceptions of how the employees ought to do their work. Managers with a knowledge of “management as science” objectify the work of employees under them. Workers know their work as craft based on firsthand experience. The author argues that traditional management practice results in decision making that does not take into account the knowledge of all organizational participants, and this leads to catastrophe. “Worker” knowledge and “management” knowledge, as well as other kinds of knowledge in organizations, are frequently incompatible. This aspect is characteristic of modern organizations but tends to be accentuated during times of organizational crisis. These two cases illustrate well the problems involved in decision making within complex organizations.

The Challenger space shuttle launch of January 28, 1986, and the Bureau of Alcohol, Tobacco and Firearms (ATF) raid in Waco, Texas, on February 28, 1993, are two infamous cases where managers made decisions that resulted in death and destruction in an arguably unnecessary fashion. To come to an understanding concerning the management practices of the Waco/ATF raid and Challenger launch decision case studies, I will apply Carnevale and Hummel’s “knowledge analytic.”1 In short, the knowledge analytic is concerned with multiple knowledges in modern organizations:

1. Idealism, two types: (a) the highly abstract and numerical knowledge of high-level managers and executives (pure reason), and (b) the idealism affiliated with investors and citizens;

AUTHOR’S NOTE: This article was originally presented to the Issues in Policy Development and Administration Panel at the annual meeting of the Southwestern Political Science Association, San Antonio, Texas, 1999.

Initial Submission: May 25, 1999 Accepted: June 25, 2000 AMERICAN REVIEW OF PUBLIC ADMINISTRATION, Vol. 31 No. 1, March 2001 66-86 © 2001 Sage Publications, Inc. 66 Garrett / ATF RAID AND CHALLENGER DECISION 67

2. The scientific knowledge of managers (application of executive directives to workers, or “public management” as it is now constituted); and 3. Realism, two types: (a) the knowledge of workers (experiential), and (b) the knowledge of the consumer (recipients who use the goods and service, etc., provided by the organization).

The primary tension within most organizations exists within the incompatibility of knowledges based on management science and worker realism.2 Management science as idealism needs constant revision to overcome its tendency to capture itself within its own techniques, which are insensitive to human demands. The lived experience of workers frequently defies the rational constraints placed by managers on their work. When workers present work that is at odds with the “perfect” plans put forward by managers, they tend to be ignored, at best, and/or blamed, at worst, if managements’ plans go awry.3 No total knowledge system yet conceived by idealism can ever be perfect (Carnevale & Hummel, 1996, pp. 69- 70). Imperfection is an ever-present fact of human existence. With the Waco/ATF raid and Challenger launch incidents, I will be comparing two well-known examples of management failure to illuminate in a stark fashion the incompatibility of worker and management knowledges. The two cases were selected because they highlight the crucial nexus between “managers” and the “managed” (i.e., workers). Managers cannot isolate themselves from their work and their workers.4 Worker-manager separation has been promoted in both theory and practice by classic renditions of social science, especially through many analyses of the Challenger launch decision. The knowledge analytic transcends these two cases and is useful in beginning to understand the complexity of decision making in human organizations. The knowledge analytic is not the theory of human organizations, but is useful for practitioners engaged in the art of management to recognize the shortcomings of management science. Both cases share important similarities that provide us (public administrators and the public) with knowledge concerning decision making that is useful in future situations. I believe that the two cases illustrate problems that are faced in many modern organizations, even though the agencies under consideration here deal with missions that involve life- and-death situations, and most other public organizations do not share the same conditions. In the analysis of both cases, I will render a historical judgment of events leading to the ill-fated launch and raid. I examine the Challenger launch decision first. I then follow with the Waco/ATF raid case and conclude with a discussion of the implications of the knowledge analytic.

BACKGROUND OF THE CHALLENGER ACCIDENT

Numerous interpretations and explanations of the causes of the ill-fated Chal- lenger launch decision have been offered by NASA administrators, the Rogers 68 ARPA / March 2001

Commission, Morton Thiokol engineers, and academicians. All assessments of the launch point to the fact that the launch took place when it was too cold for the O-rings to seat properly, thus allowing for hot gases to escape and ignite the solid rocket boosters (SRBs). The primary question surrounding the event for government investigators and scholars alike is: Why did the launch have to take place when there was ample evidence for postponement in order not to endanger the lives of the astronauts?5 Briefly and simply, explanations for the “pressure to launch” entail the following:

1. Political pressure. Two types: (a) The Reagan administration wanted to have a public relations coup for the president while the “teacher in space” (Christa McAuliffe) was orbiting, or (b) the Soviet Union might be perceived of as gaining a military advan- tage over the United States. 2. Economic pressure. Two types: (a) The launch delays were costing money and the goodwill of budget committee chairs in Congress, or (b) competition with scientists who believed that the space shuttle program was a waste of scarce resources that would be better spent on sending unmanned space vehicles into orbit. 3. Psychological pressure. The senior NASA executives and managers were either narcissistic or arrogant-vindictive, identifying their personal needs with those of the organization where pride and hubris interfered with proper judgment.6

Other theorists, notably risk management theorists,7 claim that the accident was inevitable and part of the problem of human beings operating in the intricate complexities of technology. The key events leading to the ill-fated launch decision include the activities between NASA and Morton Thiokol management and Morton Thiokol engineers. There was a series of meetings between the engineers and management to deter- mine whether the Challenger should be launched under adverse weather conditions occurring at the time. The “penultimate,” or fifth, meeting, which began at 8:45 p.m. (EST) on the evening before the launch, resulted in the presentation of numerous charts from Roger Boisjoly, a senior scientist and member of the Seal Task Force at Thiokol, consisting of information about the history of O-ring blow-by and erosion in the SRBs of previous flights and subscale testing an static tests on the O-rings (Charles, 1996, p. 116). Boisjoly opposed the launch, and most of the data he presented supported a no-launch decision. In particular, Boisjoly, an engineer with the Structures Section of Morton Thiokol, gave testimony to the Rogers Com- mission about the launch deliberation process on February 14, 1986:

Mr. Boisjoly: I first heard of the cold temperatures prior to launch at 1:00 o’clock on the day before launch, and from past experience, namely the SRM-15 launch, of which I was on the inspection team at the Cape, it just concerned me terribly. And so we started in motion to question the feasibility of launching at such a low temperature, especially when it was going to be predicted to be colder than the SRM-15 [previous cold weather shuttle launch]. Garrett / ATF RAID AND CHALLENGER DECISION 69

So we spent the rest of the day raising these questions. ...Ifelt we were very successful up until early evening, because it culminated in the recommendation not to fly, and that was the initial conclusion. I was quite please with that. ...Iwasbasically concerned with how temperature, low temperature, affects the timing function and the ability of the seal to seal. Low temperature—and I stated this for over a year—is away from the direction of goodness. I cannot quantify it, but I know that it is away from the direction of goodness. During the course of the evening, I also produced photos of the SRM-15, and my colleague produced photos of SRM-22. And you could visually see the difference in the amount of soot, as characterized past the O-ring seal. I was asked then on the net to support my position with data, and I couldn’t support my position with data. I had been trying to get data since October on O-ring resiliency, and I did not have it in my hand. We have had tremendous problems in trying to get a function generator and a machine to actually operate and characterize this particular pressurization function rate. . . . So the formal part of the presentation was finished. ...Ibasically had no direct input into the final recommendation to launch and I was not polled. I think Astronaut Crippen hit the tone of the meeting exactly right on the head when he said that, the opposite was true of the way the meetings were normally conducted. We normally have to absolutely prove beyond a shadow of a doubt that we have the ability to fly, and it seemed like we were trying to prove, have proved that we had data to prove that we couldn’t fly at this time, instead of the reverse. That was the tone of the meeting in my opinion. (Presidential Commission on the Space Shuttle Challenger Accident, 1986, Vol.IV,pp. 674-676, italics added for emphasis)

Boisjoly could not prove quantitatively to all the participants’ satisfaction that he had the data to support his conclusions.8 It was pointed out to Mr. Boisjoly in the meeting that previous missions (61-A, for example) had flown at 75 degrees Fahrenheit and that that flight had soot blow-by as well. The implication was that the temperature of the O-rings was not a salient factor, but “Mr. Boisjoly did, however, point out to the members of the teleconference that by far the worst blow-by to occur in the Shuttle’s history was on flight 51-C, and that this was an indication that temperature was a factor of O-ring resilience” (Charles, 1996, p. 117). Boisjoly was not alone in voicing his concerns regarding the effect on the O-rings, as he was supported by Robert K. Lund (vice president for Engi- neering at Thiokol). Mr. Lund recommended with the concurrence of the other engineers that the launch should not proceed until the O-ring temperature reached 53 degrees Fahrenheit. He also indicated that the launch was being considered at the lowest temperature of any previous flight (p. 117). Resistance to the no-launch decision was raised primarily from the managers from Marshall Space Flight Center. NASA/Marshall manager Stanley R. Reinartz was “appalled” by the recommendation of Morton Thiokol. And NASA/Marshall manager Lawrence Mulloy made the statement that “my God, Thiokol, when do you want me to launch, next April?” NASA managers Mulloy and Hardy were “not 70 ARPA / March 2001

convinced that the cold weather would result in a slowed O-ring, blow-by, or ultimate disaster of the Challenger” (Charles, 1996, p. 118). The Thiokol-Wasatch caucus went offline to come up with a “management” decision. Mulloy’s and Hardy’s comments added to the overall confusion. Joe Kilminster of Thiokol and others were confused further by Mr. McDonald’s statements regarding the O-rings: “in the proper position to seal if blow-by of the primary O-ring occurred.” “While many members of the teleconference perceived Mr. McDonald’s comment as a supporting statement for a launch go-ahead, Mr. McDonald did not intend to communicate that message [and] was totally opposed to a launch at such cold temperatures” (p. 118). For the most part, the engineers were ignored in the Thiokol-Wasatch caucus as not one engineer in a nonmanagement position made any positive statement supporting a launch (Charles, 1996, p. 188). After the Thiokol-Wasatch meeting was concluded, the entire teleconference reassembled to conclude discussions involving the SRBs, O-rings, and the decision as to whether to launch STS-51L. The conference ended with the final management decision to recommend proceeding with the launch, despite noted concerns about SRB O-ring blow-by (Charles, 1996, pp. 118-119). Even with the problems involving the SRBs, other problems on the day of the launch occurred that, perhaps, should have had an impact on preventing the launch of 51-L on January 28, 1986.9 The fateful launch of 51-L was thus scheduled for liftoff at 11:38 a.m. (EST). The result was the death of seven astronauts.

FINDING THE TRUTH ABOUT CHALLENGER AND FEYNMAN’S “ORGCOM” THEORY

Nobel prize-winning physicist and member of the Rogers Commission Richard P.Feynman was one of the few members of the commission who did not have direct ties to the NASA organization prior to the ill-fated Challenger launch decision. Organization communications theorist Philip K. Tompkins10 (1993) writes that Feynman is credited with having established a theory of organization communication in which he defined the differences between engineers and management. Feynman was attempting to get to the truth of what had transpired prior to the Chal- lenger launch decision. He was trained as a physicist but had an intuitive grasp of the differences between management and engineers, which led him to the Marshall Space Flight Center. Feynman11 explained how his theory of organization communication works:

Suddenly I got an idea. I said, “All right, I’ll tell you what. In order to save time, the main question I want to know is this: Is there the same understanding, or difference of understanding, between the engineers and the management associated with the rocket engines as we have discovered associated with the solid rocket boosters?” Mr. Lovingood says, “No, of course not. Although I’m now a manager, I was trained as an engineer.” Garrett / ATF RAID AND CHALLENGER DECISION 71

I gave each person a piece of paper. I said, “Now, each of you please write down what you think the probability of failure for a flight is, due to a failure in the engines.” I got four answers—three from the engineers and one from Mr. Lovingood, the manager. The answers from the engineers all said, in one form or another (the usual way engineers write—“reliability limit,” or “confidence sub so-on”), almost exactly the same thing 1 in about 200. Mr. Lovingood’s answer said, “Cannot quantify. Reli- ability is determined by studies of this, checks on that, experience here”—blah, blah, blah, blah. “Well,” I said. “I’ve got four answers. One of them weaseled.” I turned to Mr. Lovingood and said, “I think you weaseled.” He says, “I don’t think I weaseled.” “Well, look,” I said. “You didn’t tell me what your confidence was; you told me how you determined it. What I want to know is: After you determined it, what was it?” He says, “100 percent.” The engineers’ jaws drop. My jaw drops. I look at him, everybody looks at him—and he says, “Uh . . . uh, minus epsilon?” “OK. Now the only problem left is, what is epsilon?” He says, “1 in 100,000.” So I showed Mr. Lovingood the other answers and said, “I see there is a difference between engineers and management in their information and knowledge here, just as there was in the case of the rocket, but let me not bother you about it; let’s continue with the engine.” (Feynman, 1988a, p. 34)

In this story, Feynman was able to convey to his readers the problem of percep- tion between members of the same organization. Feynman’s theory is based on the knowledge of engineers and their immediate and day-to-day work in contrast with the more distant and abstract knowledge of managers. In addition to the differences in knowledge between the managers and engineers, Feynman also modified his “orgcom” theory to accommodate the apparent organizational decay at NASA.12 He wrote: “My theory is that the loss of common interest— between the engineers and scientists on the one hand and management on the other—is the cause of the deterioration in cooperation, which, as you’ve seen, produced a calamity” (Feynman, 1988a, p. 37). Another problem that came to the attention of Feynman on his trip to the Mar- shall Space Flight Center is the differences in types of knowledge between executives and managers. In this case, he is referring to a discrepancy between a range safety officer and “the big cheeses at NASA”:

We finally divided into working groups [between commission hearing meetings], and I went to Marshall with General Kutyna’s group. The first thing that happened there was, a range safety officer by the name of Ulian came to tell us about a discussion he had had with NASA higher-ups about safety. Mr. Ulian had to decide whether to put explosive charges on the side so ground control could destroy the shuttle in case it was falling onto a city. The big cheeses at NASA said, “Don’t put any explosives on, because the shuttle is so safe. It’ll never fall onto a city.” Mr. Ulian tried to argue that there was danger. One out of every 25 rockets had failed previously, so Mr. Ulian estimated the probability of danger to be about 1 in 100—enough to justify the explosive charges. But the higher-ups at NASA said that the probability of failure was 1 in 100,000. That means if you flew the shuttle every day, the average time before your first accident would be 300 years—every day, one 72 ARPA / March 2001

flight for 300 years—which is obviously crazy! Mr. Ulian also told us about the problems he had with the big cheeses—how they didn’t come to the meetings sometimes and all kinds of other details. (Feynman, 1988a, pp. 33-34)

Feynman’s theory, devised from a single case and his own life experience, provides an insight into how various levels of an organization, that is, executive, manager, and worker, can become uncooperative in joint endeavors. The NASA organization was obviously suffering from a lack of communication. Tompkins (1993) offers his interpretation of Feynman’s orgcom theory as “the failure of communication in the decision to launch Challenger was the failure to exercise automatic responsibility—to solve the problem or see that it was communicated up the line, rather than encouraging Morton Thiokol to recommend the flight” (p. 150). Tompkins demonstrated that a communication problem existed between workers and management through an examination of the Marshall leadership over the years of the space center’s existence. Von Braun was a “charismatic” leader, and Tompkins asked the rhetorical question: “The sociologist Max Weber expressed the crisis for charisma in the question: How to avoid mere routinization after the person of the organization is gone?” (Tompkins, 1993, p. 159). Tompkins then traced the Marshall leadership subsequent to Lucas. Prior to Lucas’s arrival as chief of the Marshall Space Flight Center in 1974, Rocco Petrone was sent in 1973 to Marshall where he had a brief tenure. Tompkins described unsolicited characterizations of Petrone by interviewees “as a ‘hatchet man’ who was determined to ‘weed out’ all of the Germans, to ‘cut out the fat’. . . . Petrone’s methods created a ‘persecution complex’ at the Marshall Center” (p. 160). Tompkins also confirmed Malcolm McConnell’s (1987, p. 161) harsh assessment of Dr. Lucas.13 Employees inter- viewed by Tompkins at Marshall had made numerous scathing remarks about Lucas’s leadership at the space center to the effect that “concerns from top and mid- dle managers establish something close to a consensus that Lucas’s communication style produced an ineffective system of organizational communication” (Tompkins, 1993, p. 165). In his study, Tompkins demonstrated that the Marshall Space Flight Center had some serious communications problems. Tompkins provided the cover of anonym- ity to the Marshall employees to get them to express their true feelings concerning the management style of Dr. Lucas. The Marshall Space Flight Center organization had a pattern of hostility and intimidation that had been built into the organization’s culture during Lucas’s tenure. Managers Mulloy, Kingsbury, and Reinartz were also a critical part of the decision to launch the Challenger and, according to organizational participants, were important to maintaining the management style of the senior managers at Marshall. Tompkins’ analysis of the dysfunctional managerial style of communication at the Marshall Space Flight Center, which was influenced by Feynman’s orgcom theory, provides us with an insight into some of the management practices at NASA Garrett / ATF RAID AND CHALLENGER DECISION 73

prior to the Challenger launch decision. Above all else, Tompkins and Feynman show that there is a major discrepancy between what managers and workers know concerning their work and their organization.

BACKGROUND OF THE WACO/ATF RAID DECISION

Like the Challenger launch decision, the tragedy at Waco, Texas, on February 28, 1993, could have been avoided if the dominant management ideology were not socially constructed to restrict the commonsense experiential knowledge possessed by workers (in this case, undercover agents) on the scene of the raid. Agents and managers within the Bureau of Alcohol, Tobacco and Firearms (ATF)properly proceeded with an arrest warrant for Vernon Wayne Howell (also known as David Koresh) and other Branch Davidians for possession of illegal firearms and explosive devices at the Mount Carmel compound outside of Waco, Texas.14 ATF agents having jurisdiction over the case devised three alternative plans for arresting David Koresh. The first proposal involved capturing him outside the compound, thus avoiding a direct confrontation with Koresh and the other heavily armed Branch Davidians.15 The second proposal entailed conducting a siege of the compound in which the ATF would surround the area and wait for Koresh’s surrender.16 The third proposal (the one adopted after ruling out the first two options) was the raid, which eventually resulted in the deaths of over 80 people after the crisis was finally resolved.17 One thing is certain concerning the ATF raid: The execution of the plan to arrest Koresh by raiding the Davidian compound went awry. Disaster resulted from the initial raid, as four agents and at least one Branch Davidian died from gun- shot wounds.18 Much like the Challenger launch decision, the Waco/ATF raid had a major discrepancy between rational management “thinking” and worker “experience” as to what decision ought to be made at critical times. This aspect of the knowledge analytic, the problem between management and worker knowledge, is represented here. The key problem is illustrated by the organizational interaction between the Special Response Team (SRT) raid managers Assistant Special Agent in Charge (ASAC) Chuck Sarabyn and Special Agent in Charge (SAC) Phillip Chojnacki and Undercover Special Agent Robert Rodriguez. The essence of the problem in this case is that Rodriguez, after having infiltrated the Branch Davidian compound on the morning of the raid, escaped from the scene and told the raid managers that Koresh knew that the ATF and the National Guard were coming at least 2 hours prior to the actual event scheduled for 10:00 a.m. February 28, 1993 (Report of the Treasury, 1993, p. 88). Immediately after leaving the Branch Davidian Mount Carmel Compound, a visibly upset Rodriguez reported to special agent Cavanaugh at a nearby undercover house about Koresh’s foreknowledge of the impending raid by the ATF and 74 ARPA / March 2001 the National Guard. Cavanaugh was primarily concerned as to whether “he had seen any guns, had heard anyone talking about guns, or had seen anyone hurrying around. Rodriguez responded in the negative to all three questions. Cavanaugh then told Rodriguez to report his observations to Sarabyn” (Report of the Treasury, 1993, p. 89). The events transpired in the following manner:

Rodriguez called Sarabyn at the command post and told him that Koresh was upset, that Koresh had said ATF and the National Guard were coming, and that as Rodriguez left Koresh was shaking and reading the Bible. Sarabyn asked Rodri- guez a series of questions from a prepared list provided by the tactical planners: Did you see any weapons? Was there a call to arms? Did you see them make any preparations? Robert responded in the negative to each question. Then, Sarabyn asked what the people in the Compound were doing when Rodriguez left. Rodri- guez answered that they were praying. Next, Sarabyn called Cavanaugh, who reported that there was no observable activity in the Compound. Initial accounts by the participants in and witnesses to Rodriguez’s conversations with Cavanaugh and Sarabyn differed significantly with respect to whether Rodri- guez clearly communicated that Koresh knew the raid was imminent. Although there remains some variance with respect to Rodriguez’s actual words, all key participants now agree that Rodriguez communicated, and they understood, that Koresh had said the ATF and National Guard were coming. (Report of the Treasury, 1993, pp. 89-90, italics added for emphasis)

It was clear to all the managers and agents that the raid was going forward, even though the recipient of the arrest warrant, David Koresh, was aware of the impending ATF action. This knowledge was apparent to all the rank-and-file agents, especially because the continuance of the raid would possibly result in tragedy:

Sarabyn arrived at the staging area at 9:10 a.m. Witnesses report that he was excited and obviously in a hurry. Agents in the parking lot when Sarabyn arrived recall that he ran to them and told them that they had to hurry, making statements such as, “Get ready to go, they know we are coming” and “They know ATF and the National Guard are coming. We’re going to hit them now.” Similarly, agents inside the civic center recall Sarabyn running in and calling for their attention. He announced, “Robert has just come out. Koresh knows that ATF and the National Guard are coming.” Sarabyn told the agents they would proceed immediately. Sarabyn exhorted the agents to move quickly, repeatedly telling them to hurry, to get their gear because Koresh knew they were coming. There was no formal briefing, discussion or evaluation of Rodriguez’s information. Several agents report having had qualms about going forward, especially since Koresh had mentioned the National Guard, yet they also felt questioning the decision would be inappropriate.(Report of the Treasury, 1993, pp. 90-91, italics added for emphasis)

Of course, Koresh and his followers had prepared for the imminent arrival of the ATF agents. The result was one of the worst incidents in federal law enforcement history, involving the ATF and subsequent to the raid, the arrival of Federal Garrett / ATF RAID AND CHALLENGER DECISION 75

Bureau of Investigation agents. The press that had been alerted to the event reported that ATF agents could be heard preparing for the raid:

According to some of the former cult members in the Compound at the time, preparations were being made in the Compound, although not detectable by Cavanaugh and the forward observers. Even as Rodriguez was departing, Perry Jones and the female members of the Compound had gathered in the chapel, thinking that they had been [asked to come] for a church service. They had been waiting almost an hour when Koresh came in and ordered them back to their rooms. The older women and children went to the second floor and began to lay on the floor in the hallway, away from the outer walls of the Compound. Many of the cult members began to arm themselves, some with 9mm pistols and rifles....Some donned bulletproof vests, others put on ammunition vests....Ammunition was distributed. The Com- pound members assumed stations at the windows, waiting for the ATF agents to arrive. Media personnel used radio and cellular telephones to communicate with one another and used scanners to monitor law enforcement frequencies during the hour before the raid. Several members of the press heard on scanners “no guns in the win- dow,” and “it’s a go” moments before ATF raid trucks entered the Compound’s drive- way. (Report of the Treasury, 1993, pp. 92-95)

We see that there is plenty of evidence for the managers not to go through with the raid. There was advance warning from Rodriguez to the raid commanders that Koresh knew of the impending raid, yet it took place anyway. At this point, one might conclude that the Waco/ATF raid was simply an example of incompetent management at the most basic level, that is, the raid managers Sarabyn and Chojnacki were unable to correctly read the situation and pull back from the abyss of the raid disaster.19 Clearly, the agents on the scene using commonsense judgment recognized relatively early into the exercise that they were being placed in jeopardy by the ATF raid managers Chojnacki and Sarabyn. Large amounts of men and resources were brought into the raid decision as well as months of planning. The pressure to issue the warrant for Koresh had risen to the point were the SACs believed they had to proceed. However, the roots of the incident go much deeper. The effect of executive knowledge on the ATF organization is demonstrated in this case by the following verbal exchange during congressional hearings between Representative Peter J. Visclosky of Indiana and Ronald K. Noble, Assis- tant Secretary for Enforcement, Treasury Department, on June 9, 1993:

Mr. Visclosky: You...talked about the proportionality of law enforcement responses, and law enforcement service of search and arrest warrants. Generically, when you examine if proportionaly [sic] what types of issues do you concern yourself with? Mr. Noble: I am not sure that I completely understand the question. Mr. Visclosky: For example, you mentioned that there were a large number of machine guns and a large case of ammunition in this case. Is the issue of propensity to violence in the past you serve a warrant or is it just perspective [sic] propensity taken into consideration when to violence that is considered when you decide how to approach it? 76 ARPA / March 2001

Mr. Noble: Sir, I would answer that by saying that as a manager, what I would do is I would talk to the experts. I know, since I was a prosecutor and since I teach evidence, I know how to ask questions. I would ask the theoretical kinds of questions of why are you using this option versus another option? What are the costs? What are the benefits? What is the likelihood of success? Of failure? If it goes wrong, how wrong might it go? And get answers. And then based on that, I would still, still—I want to highlight this—still give the operational Bureau head a range of discretion. I don’t know how wide or how narrow it is, but I recognize that I can’t make the operation decision. I have to make sure that I ask the kinds of questions that the policies that need to be in effect are in effect, that people are properly trained, equipped, supported, and then there are people in positions who are the tactical experts who will make recommendations to their chain of command. Ultimately a plan will come to me. If there are two ways of doing it, A and B, and I would prefer A, but in terms of the likelihood of success, either is the same, I wouldn’t say, well, I prefer A. Go ahead. I would say, you are the manager, you are the person with operational responsibility, you are the person with people on-site, you make that determination. Here are what my concerns are. (Treasury, Postal Service, and Gen- eral Government Appropriations for Fiscal Year 1994, 1993, p. 27, italics added for emphasis)

This exchange between the congressional representative and the Treasury Department executive indicates one of the central problems of modern organizations: Executives know their work primarily in terms of abstract quantities, cost-benefit analysis, legal-rationalistic rules, and a strict adherence to the organizational scalar chain of command. Managers automatically assume responsibility for applying abstract policies in a rational manner from executives to workers. Separation of the executive level of knowledge from the managerial level of knowledge is complete. The managers of the Waco/ATF raid similarly demeaned the useful knowledge of agent Rodriguez when he presented information that showed that Koresh knew the raid was imminent. Chojnacki and Sarabyn clearly demonstrated by their actions that their knowledge was superior to that of Rodriguez. The raid took place even though the managers knew that Koresh was aware of the raid plans. The decision to go through with the raid was predetermined by the management ideology (see below) prevalent in the ATF and other modern organizations. We are now at the core of the problem with modern public agencies and organizations such as the ATF. Executives such as Assistant Secretary Noble have highly abstract knowledge based on highly mathematical and rationalistic criteria. Street-level managers, in this case the SRT agents Chojnacki and Sarabyn, have to apply (make decisions based on probabilities through scientific training) com- mands generated by executives and apply them to a real-world work situation by managing employees. Workers, craftsmen in organizations with hands-on direct, experiential knowledge (in this case exemplified by agent Rodriguez), have to do Garrett / ATF RAID AND CHALLENGER DECISION 77 the actual work. This separation of knowledges between executives, managers, and workers deserves a closer examination in the following section.

THE CHALLENGER LAUNCH DECISION, THE WACO/ATF RAID, AND THE KNOWLEDGE ANALYTIC

We have now arrived at the point where multiple conceptions of knowledge may be considered when analyzing the Challenger launch decision and the Waco/ATF raid case studies. We have seen in the interpretive orgcom theoretical frameworks provided by Feynman and Tompkins that there is a discrepancy in assessing reality by executives, managers, and workers. Traditional and unconventional interpretations offered by theorists analyzing the launch decision represent a largely mutually exclusive and one-dimensional image of what transpired in this complex human event. There is a tendency by traditional theorists to perceive the reality of the event from either the management or the worker perspective. When theorists engage in an explanation of an actual event and pursue it from a single image, something of the complexity of the event is lost in translation. Feynman and Tompkins were intrigued by the evidence they had gathered from individuals about the phenome- non of perceptual differences between members of the organization. Feynman pointed out the wild discrepancy regarding the probability of a mishap based on the SRB technology between the Marshall manager, Judson Lovingood, and the engineers. Feynman also was bewildered by the safety concerns expressed by Mr. Ulian, that is, the safety factor of putting explosive charges on the space shuttle launch system to prevent a mishap that would affect population centers near the launch site, in that the “big cheeses” had an unrealistic assessment about the chances of a catastrophe occurring. Tompkins uncovered the veil of secrecy about the excessive absurdities through which the Marshall Space Flight Center had oper- ated under Dr. Lucas. A culture of intimidation was built to control information received from lower level participants in the organization. Information, or more specifically, knowledge about the work that was presented by members of the Mar- shall Space Flight Center, was suppressed and devalued by the omnipresent management ideology. Feynman and Tompkins indicate and anticipate in their analyses that there are at least two kinds of knowledge in organizations: executive/managerial and worker. There is a conceptual framework available for interpreting differences between knowledges (plural) provided by Carnevale and Hummel (1996, p. 9). Based largely on the work of the philosophers Immanuel Kant, Edmund Husserl, and Martin Heidegger, Carnevale and Hummel (p. 19) developed the knowledge analytic20 as follows:

Once we...raise the question of how we know and how different people at different places in organizations know, problems arise, for us and for them: How do we know what we want? 78 ARPA / March 2001

How do those whom we instruct know what we want? How do they know what it takes to work out what we want? The modern organization structures the answers to these questions because it structures knowledge: We, we executives, know what we want by the numbers. We, we managers, know what we see and want as objects; these we define and manipulate: organizational structures, the work as an object, the workers as objects; all these get named by us and moved around for maximum economy and efficiency like [a] piece on a chessboard. “They”...well, they, the workers, are not assumed to know much of anything. They simply do. They work. They carry motions of working according to our plans. The resulting knowledge/work pyramid . . . Executives know the ideal product; Managers know the means as objects. Workers work.

The assumption, however, that executives and managers have knowledge and workers don’t is simply not borne out by further analysis. It is a management ideology: a way of thinking about knowledge that furthers management interests and prevents profound questions from arising, not only about work, not just about respect and reward, but about the necessity of having workers who know what they are doing. The underlying principle here is that organizations are not simply divided into those who have knowledge and the rest is simply working. Instead: both the organiz- ing and the working require knowledges, plural, and these knowledges are not only different but incompatible in the sense of one not comprising the other. [The first principle of the modern organization] is that modern rational and scientific knowledge of work processes must be balanced by opportunities to translate it back into actual working moves. Knowledge of pure ideas, even the detached objective knowledge of mid-managers, must be translated into the less pure but engaged knowledge of what to do. Since modern organizations ordinarily value ideas more than reality, reforms tend to tighten the rule of ideas and subvert the reality of work. In short the answer to idealism’s move toward perfection is: the protection of imperfection. For all of these questions, however, it is necessary to inquire into the nature of the knowledges involved: to engage in a knowledge analytic. The knowledge analytic asks: What are the kinds of knowledges in modern organizations, how do they work, how can they be brought to work together? The approach reexamines management’s claim to possessing a monopoly of knowledge...[though it] does not share the conventional modern assumption that there is only one kind of ultimate knowledge in organizations. The analytic asks a sim- ple question: How do people in organizations know things? [There is not simply just the knowledge of pure reason.] Questioning the monopoly of reason also calls in question the possibility of a single knowledge elite . . . the knowledge analytic asks whether there is not a prejudg- ment or bias in the way the modern way of thinking ties effective work to ever greater perfection in one, single kind of knowing which we are used to calling by the singular term “knowledge”. . . . When it was tacitly assumed that there was only one kind of knowledge, all reforms could be viewed in terms of perfecting that kind of knowledge. The history of reforms could be read as gradual but solid progress toward a perfect state when no more adjustments and adaptations would be needed. The tolerances Garrett / ATF RAID AND CHALLENGER DECISION 79

could be tightened and closed . . . the knowledge analytic begins by reopening the question of a plurality of valid knowledges. (Carnevale & Hummel, 1996, pp. 2-9, italics added for emphasis)

In the Challenger case study, the Thiokol engineers were used as instruments for their knowledge by managers, both NASA and Morton Thiokol, to legitimate the management decision to launch Challenger. When the engineers, who were inti- mate with the working knowledge of the space shuttle system, refused to cooperate with the management decision, they were harassed, ignored, and later fired, as in the case of Roger Boisjoly. Psychoanalytical organization theorist Howard Schwartz (1990) made the argument that the NASA organization managers charged with making the decision to launch were narcissistic and had an idealized conception of themselves to the point that common sense was lost. Sound judgment was altered by a psychosis that was prevalent in the NASA organization culture. NASA was thus the “perfect” organization captured by its idealism. NASA managers were the organization. The Waco/ATF raid similarly demonstrated that the ATF special agents were (mis)used as instruments in the course of the raid. In particular, agent Rodriguez, who had been working on the case for several months, was ignored, and this action came at the expense of other agents who were killed or wounded, as were eventually several of the Branch Davidians after the siege failed under the FBI. Chojnacki and Sarabyn were both part of a cover-up in which they altered documents and attempted to get other agents to lie concerning the raid. The tendency to think about the case study in a preferred way of thinking, or under the rubric of a single knowledge, is powerful and difficult to overcome. To get beyond the traditional focus of seeing events from the view of either just the management perspective or just the worker perspective, we must be prepared to understand that there are differences in how members in modern organizations perceive reality. The knowledge analytic

outlines the differences between how managers know and how workers know as the paradoxical key problem to be overcome by modern organizations—paradoxical because what makes modern organizations so powerful is that they have solved the problem of translating working knowledge into management knowledge but not the problem of translating management knowledge back into working knowledge. (Carnevale & Hummel, 1996, p. 11)

Following Carnevale and Hummel (1996, pp. 17-18), we can readily see in the Challenger and Waco/ATF case studies the problem of different knowledges and how this can affect the way decisions are made. This may be seen in Figure 1 and Figure 2. Thiokol engineer Roger Boisjoly, involved in worker realism, could not quantify his rationale for not allowing the shuttle to launch, even though he believed, based on his experience, that the O-rings would not be able to respond properly to the cold temperatures at the time of the launch. 80 ARPA / March 2001

A. Structural Outline of the 1986 Space Shuttle Accident Normal, everyday activity Crisis Challenger explosion, January 28, 1986 Decision options Critical factors Rationalization and the considered affecting decision-making (by managers) decision considerations process (deliberation) (a) Launch/no (a) 36 degrees Fahrenheit ambient (a) Relatively little emphasis was given launch/delay temperature at launch; too cold, to workers (engineers) involved in previous coldest temperature was the work; they were ignored 53 degrees at critical times in the deliberative process Launch option (b) O-ring deterioration; sheets of ice (b) Belief that odds were in favor of selected threatened shuttle orbiter; naval success, based on mathematical rescue operations within models and science; decision to acceptable tolerance launch was “managerial” and not based on worker experience; time limitations and scarce resources emphasized in decision B. Phenomenological Outline Based on Carnevale and Hummel’s Knowledge Analytic Applied to the Challenger Launch Decision Most abstract knowledge (farthest away from the actual work) → Idealism—affiliated with investors and citizens (Congress, the president, shuttle contract investors, scientists with interests outside of manned spaceflight, and the American public) Rationalism—affiliated with chief administrators (NASA administrator Graham, Lucas, and other NASA officials, particularly at the highest levels) Science—affiliated with management scientists and midmanagers (midlevel managers such as Kingsbury, Reinartz, and Mulloy at NASA and Lund at Morton Thiokol) Realism—dominant among those who ultimately and directly produce the goods and services and those who use them (Morton Thiokol engineers such as Boisjoly and Thompson, and the seven Challenger astronauts) Most realistic knowledge (closest to everyday work experience)

Figure 1: Outline of Decision Making in the Challenger Case Study SOURCE: Carnevale and Hummel (1996, pp. 17-18), Garrett (1997, pp. 13, 188-189).

We can see the result of the decision to launch on those who experienced the loss of their lives: the astronauts. One of the results of the Rogers Commission was that those affected by the launch decision directly, the astronauts, would have representation on any launch board or panel (Report to the President: Actions to Implement the Recommendations of the Presidential Commission on the Space Shuttle Chal- lenger Accident, 1986, p. 2). This is surely a correct and appropriate response to the tragedy, and it should have been in place prior to the Challenger incident. The fact that astronauts were not involved in deliberating whether the shuttle should be launched is indicative of the predominance of management ideology. It took a Garrett / ATF RAID AND CHALLENGER DECISION 81

A. Structural Outline of the February 28, 1993, ATF Raid on the Branch Davidian Compound in Waco, Texas Normal, everyday activity Crisis ATF raid, February 28, 1993 Decision options Critical factors Rationalization and the considered affecting decision-making (by managers) decision considerations process (deliberation) (a) Raid/siege/ (a) Element of surprise was (a) Relatively little emphasis was given arrest Koresh necessary; children were to workers; undercover agent off compound potentially at risk; Davidians involved in the work was ignored were heavily armed and had at a critical time in the deliberative previously used weapons process against perceived enemies Raid option (b) Local news media ran a series (b) Belief that odds were in favor of selected of stories on the “Sinful Messiah”; success, based on numerical previous raids had proven to be superiority; tactical science gave successful, siege had proven not the ATF an edge in the decision to to be because of a failure against confront Koresh with an arrest the CSA in Arkansas warrant; Davidians were religious zealots who would not offer resistance once their leader was taken into custody; time limitations and scarce resources emphasized in decision B. Phenomenological Outline Based on Carnevale and Hummel’s Knowledge Analytic Applied to the Branch Davidian Compound Raid Most abstract knowledge (farthest away from the actual work) → Idealism—affiliated with investors and citizens (Congress, the president, secretary of the Treasury, assistant secretaries, local news media interested in a scoop, the law enforcement community, and the American public) Rationalism—affiliated with chief administrators (Director Stephen Higgins, Associate Director for Law Enforcement Daniel Hartnett, Deputy Associate Director for Law Enforcement Edward Conroy, and Intelligence Division Chief David Troy) Science—affiliated with management scientists and midmanagers (raid commanders such as Assis- tant Special Agent in Charge (Houston) Chuck Sarabyn and Phillip Chojnacki, special agent in charge of ATF’s Houston division) Realism—dominant among those who ultimately and directly produce the goods and services and those who use them (the four special agents killed in the raid, agents involved directly in the shoot-out, agent Rodriguez and the Branch Davidians) Most realistic knowledge (closest to everyday work experience)

Figure 2: Outline of Decision Making in the ATF Raid Case Study SOURCE: Carnevale and Hummel (1996, pp. 17-18), Garrett (1997, pp. 13, 188-189).

tragic event of the magnitude of the loss of Challenger and the realism of what happened to the clients (client realism) to force a change in management thinking. 82 ARPA / March 2001

We have seen from examining the Challenger launch decision case study a clear distinction between the relationship of knowledges between the various interests in the NASA organization and the Morton Thiokol organization. Knowledge incompatibility and the question of the ultimate reconciliation of the divided reality of everyday working knowledge, scientific knowledge, and investor knowledge are central to the knowledge analytic:

The question now becomes: What is the relation between these different kinds of knowledges? ...Isacomprehension of these types of knowledge by one of them possible? If not then all management reforms will continuously follow the chimera of a unified knowledge system when the reality—which someone has to absorb—is one of different kinds of knowledges in a state of mutual misunder- standing. (Carnevale & Hummel, 1996, p. 31)

The primary compatibility problematic21 for organizations is the differences in knowledges between science and the everyday work experience of the workers:

Ultimately, management science “proves” it is in certain ways superior to working knowledge. This certainly holds true for scientific management’s ability to control energy input into work (economy) and in reducing the ratio between energy input and output (efficiency)...scientific management and science in general must deny the validity of a working knowledge that is adequate for its own purposes....It must also deny any worker’s ability today to make judgments about that which he directly experiences unless these judgments can be generalized. What needs to be considered here is the possibility that scientific knowledge and everyday knowledge are far removed from one another in the definition of experience and things. The two may, in fact, be so far apart as to be possibly incompatible: i.e., requiring transformations into each other’s terms that leave essential characteristics and knowables behind....Aslong as there are economic and other power interests that value such scientific products as control, economy and efficiency, they can also compel a worker attitude that pretends to be appreciative of science’s findings and, as management science, its working imperatives. (Carnevale & Hummel, 1996, pp. 47-48)

We see clear evidence of the incompatibility of scientific knowledge and worker (engineer knowledge) as depicted in the testimony of Morton Thiokol engineer Roger Boisjoly given during the Rogers Commission hearings. Despite repeated warnings to the NASA managers in both written and oral form, Boisjoly’s knowledge about the possibility of O-ring failure based on experience, a gut feeling, and intuition (or worker realism) was ignored by the NASA managers charged with ren- dering a management decision (or management science) who wanted Boisjoly to quantify his position in a limited amount of time. We see another representation of the knowledge analytic in action with the Waco/ATF raid. The raid managers were warned by the undercover agent Rodri- guez and proceeded with the raid despite Rodriguez’s efforts. The effect of the decision was similar to that of the Challenger incident: Human beings lost their lives as a result of management dismissal of the worker’s practical knowledge and experience. Garrett / ATF RAID AND CHALLENGER DECISION 83

Four ATF agents died and 20 more agents were wounded as a result of the initial raid. Even more catastrophically, the seeds had been sown for the subsequent deaths of the victims at Waco. Further effects of the knowledge analytic are illustrated by Assistant Secretary Noble’s rational detachment from the raid events.

CONCLUSION

The knowledge analytic as presented by Carnevale and Hummel (1996) does not pretend to be a perfect theoretical conception. There is a recognition, instead, that imperfection is an ever-present fact of human existence. What the knowledge analytic provides for managers, practitioners, and academicians is a systematic means to recognize the reality of differences of knowledges in organizations. To reform mistakes and/or errors in organizations, the manager or analyst must be cognizant of the complex reality of the modern organization. Much of the humanness and complexity of the life-world in organizations is lost with the oversimplifying ten- dencies of idealized management theories dominant today. The implications for the knowledge analytic in critically understanding public administration and the development of public policy are tremendous. Further research using the concept of the knowledge analytic based on a case study approach would be desirable. Stories that managers (and other organizational participants) tell provide for a richer context to begin to capture useful knowledge for the betterment of human organizations (Hummel, 1990, 1991). Respect for, appreciation for, and real interest in all members of an organization and their knowledge could result in improvements for participants in public organizations and people receiving responsive and responsible services from public agencies.

NOTES

1. See Carnevale and Hummel (1996). The Carnevale and Hummel paper will be appearing as a chapter in the book Why Management Reforms Fail (Washington, DC: Georgetown University Press, in press). 2. See Schmidt (1993). Schmidt argues that science, engineering, and bureaucratic institu- tions, under a common model of reality, often ignore and suppress insightful kinds of knowledge (that of the worker). 3. See Garrett (1997) for a thorough exploration of the Challenger launch decision incident. 4. See Garrett (1996). In this case, the fire chiefs (managers) in the aftermath of the Oklahoma City bombing recognized that many organizational decisions were personal and should be inextrica- bly linked to worker judgment. The managers discovered that they (the managers) did not have a monopoly on the best judgment for every circumstance or situation facing the rank-and-file members of the Oklahoma City Fire Department on April 19, 1995. 5. See Garrett (1997). There is an impressive array of theoretical explanations of the event. For analytical convenience, I have categorized theories concerning the launch into four Weltanschauungen (or worldviews) in Images of Decision Making and the Launch of the Challenger Space Shuttle. Based on Gibson Burrell and Gareth Morgan’s Sociological Paradigms and 84 ARPA / March 2001

Organizational Analysis (1979), there are four major paradigms: (a) functionalist, (b) structural Marxist, (c) interpretivist, and (d) radical humanist. Most theoretical explanations of the Challenger launch fall into the first category, as do most theoretical explanations of human behavior in the social sciences. 6. See, for example, Howard Schwartz’s Narcissistic Process and Corporate Decay (1990). Schwartz’s theory is based on a composite of the psychoanalytical theories of Sigmund Freud and Carl Jung. NASA administrators were found to be narcissistic. Another recent psychological analysis was put forward in Guy B. Adams and Danny B. Balfour’ Unmasking Administrative Evil (1998) in which Marshall Space Flight Center Director William Lucas was found to be arrogant-vindictive rather than narcissistic (p. 130). In any case, both theoretical interpretations would fall into Burrell and Morgan’s (1979) radical-humanist paradigm. See also Vickers (1965/1995) for one of best and most comprehensive interpretations of how judgment affects public policy. 7. See, in particular, Diane Vaughan’s The Challenger Launch Decision (1996). Vaughan’s theory is deterministic and falls well within the realm of structural-functionalism. 8. The tension between quantitative knowledge and knowledge based on experience is key to knowledge differences between managers and engineers. Instructive here is the philosophy of Edmund Husserl (1931/1969): I am present to myself continually as someone who perceives, represents, thinks, feels, desires, and so forth; and for the most part herein I find myself related in present experience to the fact-world which is constantly about me. But I am not always so related, not every cogito in which I live has for its cogitatum things, men, objects or contents of one kind or another. Perhaps I am busied with pure numbers and the laws they symbolize: noth- ing of this sort is present in the world about me, the world of “real fact.” And yet the world of numbers also is there for me, as the field of objects with which I am arithmetically busied....The arithmetical world is there for me only when and so long as I occupy the arithmetical standpoint. But the natural world, the world in the ordinary sense of the word, is constantly there for me, so long as I live naturally and look in its direction....Ifmycogito is active only in the worlds proper to the new standpoints, the natural world remains uncon- sidered; it is now the background for my consciousness as act, but it is not the encircling sphere within which an arithmetical world finds its true and proper place. The two worlds are present together but disconnected, apart, that is, from their relation to the Ego, in virtue of which I can freely direct my glance or my acts to the one or to the other. (pp. 93-94) 9. McDonald, Boisjoly’s colleague, presented three reasons why the shuttle should not have been launched. First was the concern for cold O-rings; second, the booster recovery ships were in a survival mode, with seas as high as 30 feet and winds of 50 knots and gusts of 70 knots. “Recovery ships were heading for shore. Under those conditions it would be highly unlikely that the Solid Rocket Booster parachutes or the thrustums would be recovered....[Third was] the formation of ice on the launch pad as...Mr.Aldrich did not feel that the launch should be scrubbed because of high seas. The loss of parachutes and thrustums was acceptable, and it was felt that the Solid Rocket Boosters would not be put in undue jeopardy” (Charles, 1996, p. 119). 10. Tompkins was previously an organizational communication consultant to von Braun at Mar- shall Space Flight Center. 11. A version of this story is recounted in Feynman’s tome titled What Do You Care What Other People Think? (1988b). 12. Feynman’s simplified theory has been replicated somewhat by the organization decay models of organization theorists. See and compare, for example, Schwartz’s (1990) psychological (radical humanist) model and McCurdy’s (1989) structural-functionalist model. 13. See also Adams and Balfour (1998) in their interview with Roger Boisjoly, who is quoted as saying, “Lucas was notorious for reprimanding—or, more accurately, verbally tearing apart—sub- ordinates who made mistakes, in public meetings” (p. 131). Adams and Balfour’s concept of “administrative evil” set the stage for organizational dysfunction, that is, poor decision making Garrett / ATF RAID AND CHALLENGER DECISION 85 made possible by the “inner logic” and culture developed and possibly “masked” under the former Nazi Wernher von Braun (pp. 133-134). 14. According to testimony by Deputy Director for ATF Enforcement Daniel Hartnett, The return on the warrant has come back and everything that we thought was in there was in there. There were a total of 237 weapons; 44 were converted to machine guns; 1.5 mil- lion rounds of ammunition; live hand-grenades. They found after the fire three live ones; at least two or three were thrown at our agents, and there were hundreds of these casings that had not been filled. I say hundreds—there was a huge pile of them....There were 16 silencers, 12 shotguns, 57 pistols, 101 rifles, just everything that we thought was in there. We found a bipod for an M-60 machine gun. We didn’t find the receiver. It could have been burned away. But an M-60 is obviously stolen military property and has tremendous fire- power. Two 50-caliber semi-automatic rifles, which of course can penetrate even a Bradley. So everything we thought was in there was there. (Treasury, Postal Service, and General Government Appropriations for Fiscal Year 1994, 1993, pp. 90-91) 15. This option was summarily dismissed out of hand, even though the Report of the Depart- ment of Treasury on the Bureau of Alcohol, Tobacco and Firearms Investigation of Vernon Wayne Howell Also Known as David Koresh (1993; hereafter referred to as Report of the Treasury) pointed out that the logic for the judgment was in error. The reason for rejecting the option was “because of intelligence reports that Koresh rarely ventured off Compound grounds” (Report of the Treasury, 1993, pp. 9, 54). Koresh was seen frequently in Waco well before the day of the raid by many witnesses. 16. The siege option was rejected because the ATF believed that there would be a mass suicide within the compound and that perhaps evidence would be destroyed (Report of the Treasury, 1993, pp. 9, 38). 17. It is not my intent here to assess blame for managers and agents not involved in the subsequent takeover of the operation after the failure of the initial ATF raid. The FBI’s handling of the case is worthy of analysis elsewhere, but due to time and space limitations I have covered only the ATF’s part in the incident. 18. Most of the information for this case study comes from the Treasury document Report of the Treasury (1993). 19. In fact, this was basically the conclusion of investigators in Report of the Treasury (1993). 20. See, for example, Kant’s The Critique of Pure Reason (1781/1984), Husserl’s Experience and Judgment (1948/1973), and Heidegger’s Being and Time (1926/1962). 21. The compatibility problematic shares the same puzzling aspects as the functionalist problematic for theorists of a scientific or functionalist orientation. Burrell and Morgan (1979, p. 27) demonstrate how sociological positivism attempts to reconcile its explanatory shortcomings by incorporating more radical influences from the radical structuralist and interpretive paradigms. The knowledge analytic makes a clear distinction between subjective worker knowledge and the alleg- edly more objective management science, acknowledging their mutual incompatibility.

REFERENCES

Adams, G. B., & Balfour, D. B. (1998). Unmasking administrative evil. Thousand Oaks, CA: Sage. Burrell, G., & Morgan, G. (1979). Sociological paradigms and organizational analysis. London: Heinemann. Carnevale, D., & Hummel, R. P. (1996, November). Why management reforms fail: A knowledge analytic. Paper presented at a conference of the Oklahoma Political Science Association. Charles, M. T. (1996). The last flight of space shuttle Challenger. In R. J. Stillman II (Ed.), Public administration: Concepts and cases (6th ed.). Boston: Houghton Mifflin. 86 ARPA / March 2001

Feynman, R. P. (1988a, February). An outsider’s inside view of the Challenger inquiry. Physics Today, pp. 26-37. Feynman, R. P. (1988b). What do you care what other people think? New York: Norton. Garrett, T. M. (1996). The art of judgment: A case study organizational analysis of the Oklahoma City Fire Department, April 19, 1995. Oklahoma Politics, 5, 31-44. Garrett, T. M. (1997). Images of decision making and the launch of the Challenger space shuttle. UMI Dissertation Abstracts. Heidegger, M. (1962). Being and time [Sein und Zeit] (J. Macquarrie & E. Robinson, Trans.). San Fran- cisco: HarperCollins. (Original work published 1926) Hummel, R. P. (1990). Uncovering validity criteria for stories managers hear and tell. American Review of Public Administration, 20(4), 303-314. Hummel, R. P.(1991, January/February). Stories managers tell: Why they are as valid as science. Public Administration Review, 51(1), 31-41. Husserl, E. (1969). Ideas: General introduction to pure phenomenology (W. R. Boyce Gibson, Trans.). London: Collier. (Original work published as Ideen au einer reinen Phanomenologie und phanomenologischen Philosophie in 1931) Husserl, E. (1973). Experience and judgment (J. S. Churchill & K. Ameriks, Trans.). Evanston, IL: Northwestern University Press. (Original work published 1948) Kant, I. (1984). The critique of pure reason (J.M.D. Meiklejohn, Trans.). Chicago: Encyclopedia Britan- nica. (Original work published 1781) McConnell, M. (1987). Challenger: A major malfunction. Garden City, NY: Doubleday. McCurdy, H. E. (1989, November). The decay of NASA’s technical culture. Space Policy, pp. 301-310. Presidential Commission on the Space Shuttle Challenger Accident. (1986). Challenger accident. In Report to the president by the Presidential Commission on the Space Shuttle Challenger accident,5 vols. Washington, DC: Government Printing Office. Report of the Department of Treasury on the Bureau of Alcohol, Tobacco and Firearms investigation of Vernon Wayne Howell also known as David Koresh. (1993, September). Washington, DC: Govern- ment Printing Office. Report to the president: Actions to implement the recommendations of the Presidential Commission on the Space Shuttle Challenger Accident. (1986, July 14). Washington, DC: Government Printing Office. Schmidt, M. (1993, November/December). Grout: Alternative kinds of knowledge and why they are ignored. Public Administration Review, 53(6), 525-530. Schwartz, H. S. (1990). Narcissistic process and corporate decay: The theory of the organization ideal. New York: New York University Press. Tompkins, P.K. (1993). Organizational communication imperatives: Lessons of the space program. Los Angeles: Roxbury. Treasury, Postal Service, and general government appropriations for Fiscal Year 1994: Hearings before a subcommittee of the Committee on Appropriations, House of Representatives. (1993). 103rd Con- gress, first session, Subcommittee on the Treasury, Postal Service, and General Government Appro- priations: Bureau of Alcohol, Tobacco and Firearms Operation Trojan Horse; raid of the Branch Davidian compound, Waco, Texas. Washington, DC: Government Printing Office. Vaughan, D. (1996). The Challenger launch decision: Risky technology, culture, and deviance at NASA. Chicago: University of Chicago Press. Vickers, G. (1995). The art of judgment: A study of policy making. Thousand Oaks, CA: Sage. (Original work published 1965)

Terence M. Garrett received his Ph.D. from the University of Oklahoma in 1997. He is currently a lecturer at the University of Texas Pan American, Edinburg, Texas, and an associate graduate faculty member of Central Michigan University’s College of Graduate Studies. His areas of interest include public administration, international relations, and comparative politics.