Figure 3.1 Involvement of Groups in Foreign Policy Decisionmaking
No matter how influential or mercenary, a single leader cannot make and implement foreign policy by himself or herself. In fact, in most countries, foreign policy decisions are always made in a group setting. And these policies are virtually always carried out by particular organizations or arrays of organizations (bureaucracies).
We might consider using the following flowchart to help us orient ourselves to the role of groups in foreign policy decisionmaking. It illustrates how, depending on the nature of the foreign policy problem at hand, we can usefully conceptualize group decisionmaking in foreign policy as occurring either among a small and select group of individuals (small group dynamics) within a single organizational entity (organizational process), or between organizations (bureaucratic politics). This approach also continues our journey of progressively considering explanans that operate at increasingly higher levels of abstraction away from the mind of the individual decisionmaker.
Of course, these distinctions cannot be precisely drawn. Small groups may devolve to bureaucratic politics depending upon the group’s membership. Organizations must implement decisions regarding nonroutine problems. Nevertheless, the locus of decision in a particular foreign policy situation is likely to follow tendencies as portrayed in figure 3.1. In the remainder of this chapter, we will investigate Foreign Policy Analysis (FPA) theory regarding each of these types of groups.
Most high-level foreign policy decisions are made in small groups, meaning approximately fifteen persons or less. This is not to say that only fewer than fifteen persons are involved in any particular issue, but serious discussion of, say, a crisis situation almost demands that a leader be able to sit around a table with a set of peers and engage in candid and far-ranging debate of policy options. As a result, the study of small group dynamics has received considerable attention in FPA.
We have mentioned in chapter 2 that a leader’s personality will play a role in his or her choice of close FP advisors. Some personalities prefer groups that defer to the leader’s opinions; others want to hear dissenting views. Some personalities desire a more methodical process of decisionmaking, while others do not want to take the chance that a methodical process might either stifle creativity or lead to second guessing. Experience matters too. For example, experienced leaders are more capable of delegating, monitoring their advisers, controlling the flow of information they receive (Saunders, 2017).
Charles F. Hermann (1978) asserts that elements of the group’s structure, such as the distribution of power within the group as well as the type of role played by the group’s members, will have important consequences for group process, which in turn may have ramifications for FP choice. Groups wherein the leader holds primary power will behave differently from groups wherein the president may have considerable power but must share that power with other members at the table, such as the military chief of staff in a nation heavily dependent on the military’s sanction for rule (see also George, 1980; Greenstein, 2004; Mitchell, 2005). Likewise, members of the group may view themselves as differing somewhat in their role at the table. Some members may view themselves as loyal staff, whose presence must help facilitate promotion of the leaders’ preferences. Others may view themselves primarily as delegates of external entities, whose main purpose in the group is to clarify and argue for the perspective of that entity. So, for example, the director of central intelligence (DCI) may feel less like a staff member and more like the representative of his analysts and agents when part of a National Security Council (NSC) meeting. Still others may view themselves as autonomous actors, who are completely beholden neither to the leader nor to an external entity. These are often some of the most powerful players in the small group, because it is assumed that as they are beholden to none, their analysis is more clear-sighted, less constrained, and thus more valuable. Furthermore, the consent of these powerful players may be necessary to implement any resulting decision. In the United States, the secretaries of defense and state are often relatively autonomous players in FP decisions. For example, the United States-led bombing of Belgrade in 1998 over the Kosovo crisis was often called “Madeleine’s War” because of Madeleine Albright’s strong, almost single-handed insistence on retaliatory action against the Serbs, even in the face of a more cautionary stance taken by the Pentagon and even NATO allies.
Many FP issues are often relegated to interagency committees for initial discussion, and these are then tasked to report to the higher levels of decisionmakers. Though these interagency committees are often technically “small groups,” we will not consider them in this section because they are almost always “all-delegate” groups, whose interactions can only be understood by reference to theories concerning bureaucratic politics (addressed in the third section of this chapter).
Hermann extended his analysis of groups to talk about a more nuanced view of member role than the simple staff/delegate/autonomous actor categorization. In later work, Hermann began to develop indicators of whether, on a particular issue, a member would be an advocate of a specific policy, a “cue-taker” who would see which way the decision was going and bandwagon, or a “broker” who would use his or her influence to create a consensus position through coalitions and bargaining (P. Stewart, Hermann, and Hermann, 1989). Having identified which members of a small group would play each of these roles, Hermann then created a set of rules that helped him determine which members would take what positions, and which views would prevail as a result. Though the data requirements for such an exercise are quite high, this exercise is no different in kind than that performed by top-notch investigative journalists as they try to piece together, say, how the NSC came to a particular FP decision.
Clearly, a range of factors influence small group dynamics. And yet a crucial question remains: Does decisionmaking in small groups lead to better FPDM than if a leader were to decide alone? In light of the discussion in the previous chapter about the limitations of each human mind, it might seem reasonable to expect that surrounding a leader with advisers and experts would help mitigate the negative impacts of the cognitive biases we examined in chapter 2. In other words, “Do groups usually correct individual mistakes?” Sunstein and Hastie respond to this question with a definitive “no.” “Far too often,” they find, “groups actually amplify those mistakes” (Sunstein and Hastie, 2015, 2).
Irving Janis offered a similar observation at the beginning of Victims of Groupthink, a work that has become the most seminal on small groups in foreign policy decisionmaking (FPDM): “Groups, like individuals, have shortcomings” (1972, 3). Victims of Groupthink compares decisionmaking processes in four U.S. foreign policy fiascoes with those that operated during two successful crisis responses. Janis’s analysis finds that the small groups that oversaw fiascos became “victims of Groupthink,” which he defined as a “mode of thinking that people engage in when they are deeply involved in a cohesive in-group, when the members’ strivings for unanimity override their motivation to realistically appraise alternative courses of action” (Janis, 1972, 9). Janis’s term for this dysfunction, a deliberate allusion to the doublespeak vocabulary created by George Orwell for dystopian novel 1984, rapidly entered the lexicon. Even more impressive, the ideas Janis encapsulated in this term continue to provide rich material for scholars within International Relations (Janis himself revised (1982) and extended (1989) his initial analysis) and, increasingly, in related fields concerned with optimizing group behavior, such as psychology and business management. For example, the prominent scholars Cass Sunstein and Reid Hastie (2015) recently published Wiser: Getting Beyond Groupthink to Make Groups Smarter with the Harvard Business Review Press, in part to appeal to business leaders eager to optimize group performance (a subject we will return to below).
From the outset, it is important to qualify that not all small groups are dysfunctional. Sunstein and Hastie (2015, 2) stress that groups can correct for the shortcomings of the individual mind, provided they follow good processes. That said, the particular characteristics of FPDM—high stress, high stakes, ambiguity, uncertainty, secrecy, risk—make such episodes prone to groupthink (Janis, 1982; see also Janis, 1989 and Janis and Mann, 1977). Fear-inducing circumstances prompt us to find the emotional support that will enable us to decrease our fear to manageable levels. That emotional support is first and foremost sought through the small group itself, often because the foreign policy issues involved cannot be revealed outside of the group. Janis feels groupthink is a form of group derangement, a parallel to the derangement we often note in larger groups as “mob psychology.” Groupthink is a form of dysfunctional group cohesiveness.
In Janis’s original theory, groupthink does not arise from conscious manipulation of group members by the leader for his or her own ends, but rather from a subtle social dynamic that evolves over time. However, as ’t Hart and others have noted (1997), it is quite possible to create groupthink-like processes and outcomes in ways other than those posed by Janis. For example, the context may include a high level of threat from the leader himself or herself. For example, Jerrold Post (1991) relates the anecdote of Saddam Hussein calling together his inner circle for advice at a crucial juncture early in his reign. One minister opined that perhaps Saddam should relinquish leadership for a short while until the crisis at hand passed. Saddam thanked him for his opinion and later that day the minister’s body—chopped into pieces and placed in a plastic bag—was delivered to his wife. Needless to say, such an unusually coercive context will almost certainly promote groupthink as well. Other possibilities that may lead to the development of groupthink-like processes and outcomes would include the presence of a highly charismatic leader who elicits in noncoercive fashion an unusual degree of loyalty; a larger cultural context in which unanimity and consensus are highly valued (about which we will say more in a subsequent chapter on culture); or an issue about which the society allows for little deviation in acceptable viewpoint.
Though there are several routes to groupthink, we will examine in greater detail Janis’s original conception of the social dynamics of groupthink where the variables noted above are not in play. In the original conceptualization, group dynamics produces subtle constraints, which the leader may inadvertently reinforce, that prevent members of the group from exercising their critical powers and from openly expressing doubts when the majority of the group appears to have reached a consensus. There may certainly be sincere agreement with the emerging consensus, but Janis points out that in a groupthink group, there is a significant degree of insincere agreement as well. We have all participated in group deliberations where we went along with a decision with which we did not feel comfortable and then watched in dismay and sometimes horror when the decision turned out as badly as we thought (to ourselves) that it would. How do rational, educated persons find themselves in such a situation, assuming they are not members of Saddam Hussein’s group of advisors?
Janis opens his analysis by means of an illuminating field observation made when he was studying the social dynamics of smokers at a clinic set up to help them stop smoking:
At the second meeting of one group of smokers, consisting of twelve middle-class American men and women, two of the most dominant members took the position that heavy smoking was an almost incurable addiction. The majority of the others soon agreed that no one could be expected to cut down drastically. One heavy smoker, a middle-aged business executive, took issue with this consensus, arguing that by using willpower he had stopped smoking since joining the group and that everyone else could do the same. His declaration was followed by a heated discussion, which continued in the halls of the building after the formal meeting adjourned. Most of the others ganged up against the man who was deviating from the group consensus. Then, at the beginning of the next meeting, the deviant announced that he had made an important decision. “When I joined,” he said, “I agreed to follow the two main rules required by the clinic—to make a conscientious effort to stop smoking and to attend every meeting. But I have learned from experience in this group that you can only follow one of the rules, you can’t follow them both. And so, I have decided that I will continue to attend every meeting but I have gone back to smoking two packs a day and I will not make any effort to stop smoking again until after the last meeting.” Whereupon, the other members beamed at him and applauded enthusiastically, welcoming him back to the fold… . At every [subsequent] meeting, the members were amiable, reasserted their warm feeling of solidarity, and sought complete concurrence on every important topic, with no reappearance of the unpleasant bickering that would spoil the cozy atmosphere. (1972, 8)
This case, because of its extremity, reveals some of the dynamics at work. When a group is formed, two separate forces are set in motion. The formation of the group sets in motion a decision process to tackle the issue or problem at hand. But the formation of the group also sets in motion a social institution that is to be maintained over time. Thus the group has, in a sense, two goals: to effectively address the problem that catalyzed its formation and to continue to function as a group. These two goals are neither intuitively nor inevitably at odds. But in groupthink groups, such as the smoking clinic, they become at odds over time.
Group cohesiveness is a powerful source of emotional support for small group members. We see this dynamics at work in families, in gangs, in sports teams, in military platoons, in groups of friends, among online communities, in business departments, and so forth. The rest of the world may not appreciate you or even like you, but as long as the people who interact with you in a salient small group (and thus arguably know you best) appreciate and like you, what the rest of the world thinks may not cause you psychological distress. Conversely, the capacity to produce psychological distress for its members is heightened in small groups that interact over time. The source of that stress is fear—fear of ostracism by the group.
This shift in the source producing the emotion of fear is extremely consequential. The original fear of failure in addressing the problem that catalyzed formation of the group is now compensated for by the emotional support provided by the group itself—but then the prospect of potentially losing that support produces a fear of group ostracism that may dwarf the original fear of task failure. Thus maintenance of group cohesiveness may evolve into the group’s primary purpose, supplanting the original task-oriented purpose for which the group was formed in the first place. When this occurs, groupthink exists. What one begins to fear most is to be labeled as a deviant from the group. As noted in the smoking clinic example, if a group member expresses deviance, the other members of the group will try to influence him or her to revise or tone down their dissident views. If they are not successful in bringing the deviant back into the fold, he or she will be excluded from the group—at first subtly, and then more overtly. Insincere agreement to avoid ostracism may then arise.
In addition to the group’s purpose being supplanted and insincere agreement occurring, Janis notes several other hallmarks of groupthink. First, the group’s standards of judgment are changed and lowered. The group’s standards for judging a matter may stray from more objective reasoning to reasoning based on the desire to prevent deviance or lack of cohesiveness and preserve amiability above all else. Second, groupthink groups begin to think very well of themselves and their members. A groupthink group will feel that it and its members are wiser, more powerful, more knowledgeable, and more virtuous than those who do not belong to the group. This inflated self-image may have several consequences. For one, nonmembers may be dehumanized, especially those who are seen as competing with the group. Nonmembers may be seen as inferior or evil, and action that might not usually be considered moral might be deemed appropriate to deal with nonmembers. For another, inflated self-image may lead to the “risky shift”: the propensity for groupthink groups to collectively decide on more risky behavior than any one member of the group would have chosen individually (this is sometimes called “group polarization”). An easy analogy is to teenage gangs. Often these gangs are capable of risky, violent, criminal behavior on a level that no one teen in the group would dare attempt.
In sum, Janis asserts that groupthink groups are hard-hearted but soft-headed. This softheadedness can also manifest itself in sloppy decision practices due to lowered standards of judgment and inflated self-image. In his case studies of foreign policy fiascoes, Janis finds that the groups in question usually examined only two options to deal with the problem they faced, and that the group would quickly seize on one of the two options that would never again be critically examined for weakness. He also found very little effort on the part of these groups to obtain information from knowledgeable nonmembers but instead found a selection bias in the evaluation of information to favor the preferred option, and an utter failure to establish contingency plans in case the preferred option was unsuccessful. Sloppy decisionmaking did not induce psychological stress because there existed a compensatory inflated self-image: the groupthink groups thought of themselves as not only omniscient, but also as invulnerable. And immoral decisionmaking likewise did not induce stress because loyalty to the group had become the highest form of morality.
Janis is quick to note that not all foreign policy fiascoes are produced by groupthink groups. Furthermore, it is possible for a groupthink group to operate without producing a fiasco. Steve Yetiv asserts that the inner circle of foreign policy decisionmaking under the George H. W. Bush administration arguably suffered from groupthink but was nevertheless capable of sound foreign policy decisions (Yetiv, 2011; see also Garrison, 2012). However, ceteris paribus, it is much more likely for a groupthink group to create fiascoes than otherwise, given its dysfunctional attributes. A case in point, argues Janis, is the 1961 Bay of Pigs episode.
That the Bay of Pigs invasion was a fiasco by any standard is not in doubt. On April 17, 1961, about fourteen hundred Cuban exiles, trained by the United States for this purpose, invaded Cuba at the Bay of Pigs. By the second day, the brigade of exiles was completely surrounded by over twenty thousand Cuban troops. By the third day, about twelve hundred (all who had not been killed) were captured and sent to prison camps. About twenty months later, the United States ransomed most of these with $53 million in food and medicine. The European allies, the United Nations, and friendly Latin American regimes were outraged, and the invasion may have been the catalyst for new military agreements between Cuba and the USSR, which would eventually culminate in the Cuban missile crisis. Even John F. Kennedy, president at the time, asked rhetorically, “How could I have been so stupid to let them go ahead?” (Janis, 1982, 16).
Janis points to the underlying dynamics of Kennedy’s first foreign policy inner circle, which included Dean Rusk (secretary of state), Robert McNamara (secretary of defense), Robert “Bobby” Kennedy (attorney general and the president’s brother), McGeorge Bundy (special assistant for National Security Affairs [ANSA]), Arthur Schlesinger, Jr. (White House historian), Allen Dulles (director of central intelligence [DCI]), and Richard Bissell (deputy director of central intelligence [DDCI]). Kennedy had only been in office a very short time. He was under stress to perform well in foreign policy, since he was the youngest president ever elected, he was a Democrat, and he was a Catholic. Kennedy was not the only “greenhorn” in the group: McNamara, Bobby Kennedy, Bundy, and Schlesinger were all new to government, not to mention high-level government office. In the recent presidential campaign, his opponent Richard Nixon had painted Kennedy as too young and inexperienced to stand up to the Soviet threat. Was Kennedy tough enough?
Dulles and Bissell, both holdovers from the previous Eisenhower administration, briefed Kennedy on the ongoing plan for the exiles’ invasion of Cuba. The plan, therefore, was the plan of his predecessor: Dwight Eisenhower, two-term Republican president, hero of World War II, and a man about whom no one had qualms about “toughness.” Fear of failure in standing up to the Soviet threat was to be extinguished for Kennedy via the emotional support he would get from his small group of advisors. But most were newcomers themselves and had as great or greater fear of failure as he did. Emotional support from Dulles and Bissell, then, would be key. Since they had crafted the invasion plan, this needed emotional support would only be forthcoming from these men if the plan were accepted. This social dynamics sets the stage for groupthink.
Janis points to additional factors auguring in favor of groupthink. Kennedy’s election had ushered in a sense of elation and invulnerability among his inner circle. Schlesinger later put it this way: “Euphoria reigned: we thought for a moment that the world was plastic and the future unlimited” (Janis, 1982, 35). Janis also identifies Bobby Kennedy as a self-appointed “mind-guard” who would attempt to corral deviants who expressed second thoughts privately: in one instance, Bobby, in the midst of his wife Ethel’s birthday party, accosted Schlesinger about the latter’s doubts with, “You may be right or you may be wrong, but the President has made his mind up. Don’t push it any further” (Janis, 1982, 40). Furthermore, Schlesinger himself noted at the time “a curious atmosphere of assumed consensus” (38). No one spoke up against the plan in the group’s meetings, even though numerous members apparently did harbor doubts. Silence was interpreted as consent.
In this context, then, group decisionmaking processes deteriorated in quality. Though the press had leaked the invasion plan, the plan proceeded. The State Department and British intelligence contradicted the CIA position that Castro’s army and air force were weak, but there was no attempt to discover which position was correct: the CIA’s position was accepted uncritically. One assumption of the plan was that the invasion would ignite the Cuban underground, which would then revolt in the cities. Janis points out that not only did no one think to let the underground know that an invasion was imminent, but that since Castro was alerted by U.S. press reports to the plan, he took preemptive measures to round up dissidents. An egregious error was the decision to move the landing site from Trinidad to the Bay of Pigs—without looking at a topographical map that would show that the Bay of Pigs was a swamp far removed from the Escambray Mountains (which is where the invaders were to flee if they ran into trouble).
Though the Bay of Pigs invasion was a fiasco, Janis argues that Kennedy learned invaluable lessons that prepared him for the higher stakes of the Cuban missile crisis, which would play out only eighteen months later. Among other things, ExCom (the small group formed in response) proceeded quite differently in the second crisis, despite featuring most of the same key players. A wide range of options was considered, and Kennedy refused to allow the group to move swiftly to adoption of a preferred option. Experts, particularly from the military, were grilled instead of being shown deference. Dissension was encouraged, and Bobby Kennedy often assumed the role of devil’s advocate. Participants were explicitly asked to be skeptical. There was no formal agenda and no protocol. Subgroups of ExCom met with or without President Kennedy. Often lower-ranking officials were asked to meetings to which their bosses were not invited. Contingency plans were extremely detailed. Kennedy fostered an air of discomfort and reminded all of the grave dangers involved. Issues of morality were openly raised. Reversals of judgment were frequent. Kennedy had members role-play Khrushchev and Castro, pushing for a nonstereotypical view of the enemy alongside themes of nonhumiliation and non-underestimation.
Janis argues that if we are pleased with the result in the Cuban missile crisis, part of the credit must go to Kennedy being scrupulous and diligent in avoiding groupthink at all costs. Janis feels this case illustrates that it is possible to consider measures to head off this pernicious social dynamics. In his research, Janis explores a variety of ways to defuse this all-too-frequent phenomenon. He encourages leaders to avoid homogeneity in the background of group members and to refuse to dissipate stress and discomfort through group amiability. Leaders might do well to appoint a devil’s advocate, though that role may have to be rotated over time so that the person’s views are not automatically dismissed due to role expectations. Janis urges leaders not to make the group too insular, to invite in outsiders and experts to openly challenge group assumptions. Kennedy’s use of subgroup meetings is a good way to make room for dissent, especially if the leader himself is not present. Janis also counsels leaders to hold their opinions to themselves as long as feasible, so as to not inadvertently close off dissent. A checklist of good decision practices might be used to ensure that no important steps have been omitted. Role-playing and study of the other nations involved in order to construct realistic alternative scenarios are very useful. And finally, Janis notes that a variety of cultures have norms of the “last chance” meeting, where after a decision is finalized, participants often get drunk (or otherwise lower their social inhibitions) and then meet again to see if they still agree on the decision made.
In retrospect, the symptoms of groupthink can often appear strikingly obvious. Likewise, the measures required to mitigate the onset of this group pathology—adhering to good process, deliberative seeking of outside opinions, encouraging skepticism, and the like—can seem like common sense. When considering cases where groupthink emerged, we can find ourselves paraphrasing JFK, asking “how can they be so stupid?” Yet in much the same way that an awareness of our own vulnerability to cognitive biases does not make us immune from them, so an understanding of the susceptibility of small groups to engage in groupthink does not necessarily prevent us from succumbing to its detrimental and often-imperceptible (especially at the time) effects.
This tendency to unwittingly repeat the mistakes of the past is illustrated in the lead-up to the U.S. invasion of Iraq in 2003. There is now widespread scholarly agreement that the small group dynamics operating during this FPDM episode exhibited “a classic groupthink dynamic” (Mintz and Wayne, 2016, 92) given many of the symptoms that Janis identified are also clearly detectable in this much more recent episode (see especially Badie, 2010, but also Mitchell and Massoud, 2009; Mintz and DeRouen, 2010; Schafer and Crichlow, 2010; Yetiv, 2013).
George W. Bush’s personality and leadership style made the deliberations with his close advisors especially vulnerable to groupthink. Temperamentally, Bush was an instinctive decisionmaker inclined to “trust his gut” rather than reaching a verdict by presiding over extended formal decisionmaking processes (Mitchell and Massoud, 2009, 273). Furthermore, at the time of the 9/11 attacks, Bush was a first-year president with little previous foreign policy experience to draw on. Indeed, this inexperience had earlier prompted Bush to assemble a highly regarded and experienced foreign policy team around him during his presidential campaign. This group, which came to be known as the Vulcans (Mann, 2004), included senior figures from the previous two Republican administrations. The Vulcans’ key members shared a neoconservative ideology, which proved crucial in the aftermath of 9/11, not only for the nation, but for President Bush himself (Renshon, 2005).
The decisive impact of 9/11 was to change “the administration’s view of Saddam from troubling dictator to an existential threat to US security” (Badie, 2010, 277). This shift in perception happened almost instantly. Recalling the moment he was informed that two planes had flown into the world trade center, Bush later acknowledged to reporters that he “made up [his] mind at that moment we were going to war” (Woodward, 2002, 15). Even allowing for the fact that the more immediate U.S. response to the 9/11 attacks was to invade Afghanistan, it is possible trace how, almost from the outset, Bush and key advisors began to conflate Iraq with the War on Terror. (For an example which traces this process using the public discourse of the Bush administration, see Widmaier, 2007). This conflation, argues Badie (2010, 278), was “pathologically driven by groupthink.”
Badie goes on to explain how once “members of the group internalized the image of Saddam Hussein as an existential threat, the logical policy prescription became immediate military action” (2010, 291). Yetiv agrees. “It was Saddam’s misfortune,” he observes, “that Iraq represented precisely what the Bush administration feared after 9/11: a dictator with connections to terrorist groups developing WMD” (Yetiv, 2013, 55). Adding to the momentum toward invasion, the key neoconservatives within Bush’s inner circle—Vice President Dick Cheney, Secretary of Defense Donald Rumsfeld, and Deputy Secretary of Defense Paul Wolfowitz—were predisposed toward regime change (Renshon, 2005; Smith, 2008, 103; Badie, 2010, 283). Smith (2008, 103) found that “the eventual invasion [of Iraq in 2003] was driven by a group of highly placed officials with a pre-existing ideological commitment to getting rid of Saddam.”
Once Bush and several of his more hawkish advisors had decided on a preferred course of action, they neglected to seriously consider other options. Bush, for his part, rarely asked probing questions of his team (Mitchell and Massoud, 2009, 281). Outside experts were not actively canvassed (Smith, 2008, 102). Most notably, a leading and unlikely critic, Brent Scowcroft, who served as National Security Advisor under George H. W. Bush and Gerald Ford (and was therefore well known to the members of Bush’s foreign policy team), was kept at a distance, left to prosecute his opposition to the invasion in the opinion pages of leading newspapers (Mann, 2004, 340). These dynamics contributed to a culture of “anticipatory compliance,” whereby the principals would align themselves with Bush’s preferred policies (Mitchell and Massoud, 2009, 277–279). Skeptical group members, notably Secretary of State Colin Powell and Director of the CIA George Tenet, were pressured to conform, with Cheney playing the role of mind-guard in much the same way Bobby Kennedy did for JFK in the lead-up to the Bay of Pigs invasion four decades earlier.
Another symptom of groupthink—perhaps the signal failing of this decisionmaking process in the minds of the public—were the “information problems” that came to light post-invasion. Without getting bogged down in the detail of these problems, it suffices to say that crucial intelligence either did not reach the group, was not properly considered, or was ignored (Yetiv, 2013: 62). At least some of these oversights can be explained by an atmosphere of presumption that existed, at times, within Bush’s team; indeed, Yetiv labels the 2003 invasion of Iraq as a “war of overconfidence” (Yetiv, 2013, Chapter 4). Bush’s infamous “mission accomplished” speech on the deck of the USS Abraham Lincoln came to symbolize this dynamic as it became clear over time that his administration had not adequately planned for the likely contingencies in post-invasion Iraq.
The U.S. invasion of Iraq in 2003 illustrates why groupthink is, and will remain, a crucial lens through which to examine FPDM, especially in those instances where decisionmaking processes appear to have been deficient. That said, before moving on from our discussion of groupthink, a final caveat is warranted. Though we have spoken very negatively of groupthink, for good reason, it is possible that the attempt to foment group cohesiveness might have its uses. One such documented use was the 1994–1995 talks between Palestinian and Israeli negotiators to work out the details of the Oslo Accord regarding the West Bank. At the Patio Hotel in remote Eilat, the Israelis took the third floor, the Palestinians took the second, and the talks were held on the first—and no one was allowed to leave for months. As an Israeli negotiator put it, “We created a setting in which there was no physical way out without an agreement” (Schmemann, 1995, A1). The article goes on to note:
You could watch the peace process develop like one of those American soap operas. You saw who went to whose room, who was negotiating with whom.” The delegates ate together, went to the health club together, Israeli generals took saunas with Palestinian guerrillas. “It created a club mentality vis à vis everybody else. We needed a common enemy, and it became the media. We developed a deep understanding of each other’s paranoias, we created a certain trust among representatives of total mistrust.
Even here, we see the power of the emotional support that small groups can provide: power enough to overcome historical hatreds (at least temporarily). The influence of small group dynamics on foreign policymaking should never be underestimated, but rather studied, understood, and used to promote functional ends. However, if you still harbor doubts after considering both the Bay of Pigs and Iraq 2003 cases, perhaps you will be persuaded by the behavior of the corporate sector, which is increasingly taking steps to optimize group dynamics.
Cultivating high-performing teams is not only crucial in preventing foreign policy fiascos, it is valuable in driving corporate growth and innovation. Consider Google. In 2012, the famously data-driven company launched Project Aristotle, an internal research initiative that sought to understand the conditions for collective intelligence. Google wanted to know why when they grouped some of their best and brightest employees together, their collective performance was not as good as might be expected, while some other teams reached a level of performance that seemed to exceed the sum of their parts.
Google’s research team identified two behaviors that good teams shared. First, the good teams displayed “equality in distribution of conversational turn-taking” (Duhigg, 2016). Everybody contributed to the conversation in roughly the same proportion. Second, high-performing groups consisted of members who were tuned in to the feelings of their fellow members (that is, they displayed high “average social sensitivity” in the words of Project Aristotle).
Quite simply, Google found that “what really mattered was less about who is on the team, and more about how the team worked together.” And most important of all was the presence of “psychological safety,” a term first advanced by Amy Edmondson (1999). According to Google’s guide for understanding team effectiveness, for a small group to function well, group members needed to “feel confident that no one on the team will embarrass or punish anyone else for admitting a mistake, asking a question, or offering a new idea” (re:Work with Google, n.d). In essence, as we highlighted at the outset of this discussion, small groups function best when they provide the emotional support needed to alleviate feelings of fear and vulnerability. The question of how to create a group that simultaneously offers psychological safety but also clear-eyed realism and due diligence to standards and ethical principles is the new frontier of small group research.
There are other scholarly insights on small group dynamics that deserve mention. For example, the psychologist Garold Stasser noted that most small groups tend to rely primarily on information about the problem that is already known to all or nearly all group members before group discussion commences (Carey, 2005, 1, 3). Important information that only a few members of the group hold will probably not be used and is likely to be overlooked in the group discussion. Apparently, the easiest psychological route to agreement is not learning new premises for a decision but discovering common premises that already exist within the group.
Ryan Beasley’s work on how small groups come to agree on a problem representation moves the small group dynamics research agenda forward in significant ways (1998). Beasley believes that small groups are not identical: there is a taxonomy of groups according to characteristics such as the centrality of particular individuals, the complexity of group discussion, the degree of alternation between speakers, the continuity of the discussion, and so forth. Thus, each type of group may be predisposed to a certain style of group decisionmaking. Beasley postulates several varied processes for group aggregation of individual understandings: simplicity (“classic” groupthink), single representation embellishment (leader-drive groupthink), factionalism, common decomposition, common alternatives, and expertise. In a study of meetings of the British Cabinet over the Munich crisis, Beasley found that each of these types of decisionmaking was used over time. Groupthink-style processes occurred in only five of the twelve meetings. Thus there may be more nuance and complexity to small group dynamics than the work of Janis might suggest.
Donald Sylvan and Deborah Haddad (1998) suggest also that in cases of group conflict over problem representation, the technique of “storytelling” begins to dominate, in which participants compete with each other to provide the most articulate causal argument concerning a particular problem. The views of those with the most persuasive story will become the basis for decisionmaking by the group.
The fine volume Beyond Groupthink, edited by ’t Hart, Stern, and Sundelius (1997), suggests that the “group-as-decisionmaker” might be too simplistic. The small group in FPDM may play a variety of roles that should be considered, not just “command center,” but also sanctuary, smoke screen, and arena. Furthermore, the effects of leader personality, culture, and institutional context on small group structure and function need further attention. For example, Stern and Sundelius believe the Bay of Pigs fiasco is better explained as “newgroup syndrome” than classic groupthink à la Janis (Stern, 1997), and Marsh explains Obama’s 2009 surge decision in similar fashion (2014). Paul Kowert has also extended groupthink research by examining groupthink’s relationship to learning. He presents the interesting argument that while “some leaders thrive on diversity of opinion, others are immobilized by it” (2002, 4). He suggests that careful attention must be paid to matching a leader’s style of learning and tolerance of disagreement with the organization and membership of the foreign policy advisory group. It is a failure in learning, argues Kowert, that leads either to groupthink or to deadlock. Additional important work on the relationship between leaders and their foreign policy advisory groups can be found in a body of scholarly work produced by Thomas Preston, whose analysis of LBJ’s advisory structure we examine below (Preston, 1997, 2001; Preston and ’t Hart, 1999; see also Glad, 2009). In addition, Mark Schafer and Scott Crichlow have added a more formal modeling component to the interaction of leader characteristics and the production of groupthink (2010). Furthermore, Vertzberger suggests that scholars look more deeply into the cultural context of small group dynamics, pointing to the guru–chela (teacher–disciple) template for political relationships in India as an example (1997; see also 1990).
Going further, Hoyt and Garrison wonder why strategic manipulation of a small group by political “gamesmen” has not been researched more fully (1997; see also Garrison, 1999): tactics such as noninvitation to meetings, nonsharing of information, destroying a member’s credibility, casting a member as an insubordinate when the member refuses to be silenced or excluded, duplicating another member’s assignments to provide alternative information, dropping an item from the agenda, and so forth. One devastating example of the outsized effects on foreign policy of this gamesmanship can be seen in the Tonkin Gulf incident of 1964 that led to the introduction of regular U.S. combat troops to fight in the Vietnam War (Porter, 2014). Robert McNamara, Secretary of Defense, purposefully did not tell President Johnson that the commander who had reported the attacks on the U.S. patrol boats had subsequently informed McNamara that he had serious doubts the boats were actually attacked and he was launching a full investigation. Instead, McNamara urged Johnson to retaliate, and Johnson agreed. Even more egregiously, after the executive order was given, commanders in the field again expressed doubts and asked for a delay of execution to continue the investigation. Without telling Johnson of these further misgivings by those on the ground, McNamara reiterated the executive order to strike.
On a more encouraging note, sometime the chief executive is a better gamesman than his or her staff. For example, after the Israelis discovered the construction of a nuclear facility in Syria, they notified the George W. Bush administration in 2007 that they were going to take military action. Bush’s closest advisors were deeply divided on how to respond to the Israelis, and after much deliberation Bush sided with Condoleezza Rice that the evidence should be brought to the IAEA (International Atomic Energy Agency) in Vienna instead. The Israelis rejected that decision, of course. About two months later, Prime Minister Olmert called George W. Bush just as the Israelis were in the middle of blowing up the site. After the call, Bush said of Olmert, “that guy has guts,” and his advisors realized he had delayed taking the issue to Vienna and maintained absolute silence in order to give the Israelis time to carry out their plans. Rice had not actually won the day, though that was not understood by his advisors at the time (Abrams, 2013).
What makes a group “work”? In newer research, the concept of “group efficacy” has been applied to foreign policy decisionmaking. Charles Hermann (2012) and Thomas Preston (2012) examine the inner circle of foreign policy decisionmaking in the Lyndon B. Johnson administration, with an eye on the degree to which the group as a whole shares a belief that the group has the ability to achieve a certain goal. Such “high-efficacy groups” are much more likely to stay committed to a particular chosen course of action despite adverse feedback. The sources of this group efficacy belief tie in fairly closely with the course of groupthink but are not identical: homogeneity in the background, group cohesion, appraisal of the past performance of the group members, and support and confidence expressed by those external to the group. Hermann shows how it took a series of punishing foreign policy failures, including the famous “loss” of Walter Cronkite, an influential and longstanding television news anchor often referred to during his career as “the most trusted man in America,” to get this group to see the reality of the situation.
Preston notes that the real turning point probably took place in February 1968 when Dean Acheson, one of the Wise Men called upon by LBJ for the needed external support, walked out on Johnson in the middle of a one-on-one meeting with him because Johnson spent forty-five minutes telling him how well the war was going. According to Preston, “When NSC Advisor Rostow phoned to ask why he had walked out, Acheson replied, ‘You tell the President—and you tell him in precisely these words—that he can take Vietnam and stick it up his ass.’ ” Johnson then got on the line and told Acheson to return; Acheson did and in effect told Johnson the Joint Chiefs of Staff had fed him (Johnson) misinformation. When Johnson said he was shocked, Preston reports that Acheson countered, “Then maybe you should be shocked” (71–72). At the final Wise Men meeting of March 25 before President Johnson’s fateful speech of March 28 in which he not only shifted course on Vietnam but also announced his unwillingness to run for a second term, Johnson was mostly silent, which was quite unusual. In this setting, an intense debate broke out. After hearing General Wheeler argue that the purpose of the American military presence in Vietnam was to stave off a Communist victory and not to win the war, Acheson exploded: “Then what in the name of God are five hundred thousand men out there doing—chasing girls? This is not a semantic game, General: if the deployment of all those men is not an effort to gain a military solution, then words have lost all meaning” (76). The fairly abrupt shattering of the group’s sense of efficacy over the course of a few months was necessary for an about-face by the president.
Finally, the notion of “polythink” is probably the most distinctive recent theoretical development in the study of small group dynamics (Mintz and Wayne, 2016). Polythink is groupthink’s mirror image, for though a polythink group is almost the opposite of a groupthink group, the results for foreign policy can be equally disastrous. Mintz and Wayne offer the following definition: “Polythink is a group dynamic whereby different members in a decision-making unit espouse a plurality of opinions and offer divergent policy prescriptions, and even dissent, which can result in intragroup conflict and a fragmented, disjointed decision-making process” (2016, 3). Mintz and Wayne view groupthink and polythink as the two ends of a continuum of how convergent or divergent the beliefs of the group are, and they urge leaders to try and avoid both these extremes, harnessing divergent viewpoints into “a single, cohesive policy direction” (6).
Since FPA scholars must often diagnose group character without benefit of insider information, Mintz and Wayne suggest analysts look for the symptoms of polythink—turf battles, framing battles, leaks, siloing of information, lowest common denominator decisions, and even decision paralysis. Furthermore, circumstances may dictate whether a group veers in the direction of groupthink or polythink—that is, the same group may exhibit differing characteristics based on the situational context. For example, Mintz and Wayne’s case studies suggest that the decision to enter wars may emanate more from a groupthink process, whereas the decision to exit a war may emanate more from a polythink process. Adding another layer of analysis, there can be subgroups within the decision-making group that have a groupthink character, but these subgroups are operating in a larger polythink group setting.
Mintz and Wayne apply their framework to several cases, for example, understanding how the 9/11 hijackers were not prevented from carrying out their terrorist plot despite the fact that both the FBI and the CIA were aware of strange doings by these individuals with terrorist group ties (see also Norton and Stigler, 2015). Unfortunately, neither agency communicated with each other nor the White House about these facts. Pre-9/11 there was also a deep divide within the Bush White House as to whether to treat terrorism as a crime or an act of war. This dispute had important ramification for how to treat governments from whose territories the terrorists operated—were they equally culpable as the group and thus justifiably a target, or equally endangered as the United States and thus justifiably in need of U.S. support? The upshot of the divergent views, the ambiguity of the intel, and the siloing of needed information was “a lowest common denominator decision in which intelligence operations would continue unabated, but military action would not proceed” (Mintz and Wayne, 2016, 53). Of course, after 9/11, the Bush NSC veered swiftly to converge on decisions for military action.
In conclusion, there is much more ground to be plowed in FPA concerning the analysis of small group dynamics. A new generation of scholars is producing an impressive body of work adding to our understanding of this important topic.
Though small group dynamics are extremely important in understanding foreign policy behavior, it must not be overlooked that most high-level foreign policy decisions are implemented through large executive organizations, such as departments and agencies. Furthermore, the government’s “senses” are these same organizations: the gathering of information and the initial processing of information are performed for the most part by organizations. Governments both perceive and act primarily through organizations.
This situation invites us to explore the degree to which the government is not a unitary rational actor. Given the prominence of organizations in the government’s ability to conduct foreign policy, it might be more useful at times to view the government as a matrix of organizations, or, in other words, as a national bureaucracy. There are multiple actors in a national bureaucracy, not one unitary actor. And just as we have found that a collection of individuals within a small group might not act in classically rational fashion, so we can also speculate that the actions of the multiple bureaucratic players might also result in behavior that is less than optimally rational and coordinated. Those who have had the opportunity to work within a large organization, whether that be a government agency, a business corporation, a university, or school system, or even an organized religion, inevitably discover that sometimes the collective is less intelligent than the sum of its members.
So why have organizations at all? Organizations exist to provide capabilities that otherwise would not exist. Consider the case of space exploration, such as sending probes to Mars or Saturn’s moon Titan. When one details all the subtasks involved in accomplishing those larger tasks, it becomes clear that without large collectives of people pooling resources, knowledge, labor, and leadership, no space exploration would ever have taken place. Tasks such as space exploration, or even the fielding of an army, require specialization so that larger tasks may be divided into smaller, more feasible ones. Such endeavors also require a tremendous amount of coordination and communication, with the ability to preserve memory as particular individuals enter and leave the larger organization. Remember that some large organizations relevant to foreign policy, such as the U.S. Department of Defense, have an immense number of employees, including 700,000 civilians in addition to 1.4 million in military service (Walt, 2018, 97). Furthermore, the task of coordination is growing ever more difficult, as government organizations expand rapidly. For example, the President’s foreign policy staff consisted of under twenty people in the early 1960s and has grown steadily to a point where President Obama’s foreign policy staff numbered over four hundred (Walt, 2018, 97). In this section on Organizational Process, we focus heavily on the U.S. case, for while the dynamics we discuss operate in all types of organizations across the globe, the most detailed FPA literature on organizational process remains largely U.S.-focused, though there are exceptions we highlight later in the chapter.
A common reaction to the outsized influence of these organizations on U.S. foreign policy is to anthropomorphize them and speak in such terms as “The Defense Department wanted greater authority to collect intelligence, and it got what it wanted.” This type of language, again connoting a unitary rational actor but at a lower level of government, conceals a more complex reality. Though large organizations contain many human beings, large organizations are arguably a simpler form of life than a human being. First, they have a constrained functionality related to the purpose of their creation. It is useless to ask NASA to plan the invasion of Afghanistan. It is useless to ask the State Department to send a man to the moon. Of course, some organizations may be interested in expanding their functions, but by and large that cannot happen quickly. Organizations will develop specific skill sets, which will constrain what they are able to do. Second, this will give rise to an organizational culture, which is an understanding by the humans in the organization as to the organization’s identity and mission and vision. Morton Halperin calls this an understanding of the organization’s “essence,” which, once entrenched, is almost impossible to change.
One’s essence leads to the staking out of particular “turf,” meaning an understanding of which issues the organization can claim a “stake” in, or organizational interest. Concerning some issues the organization may view itself as the primary “stakeholder” within the bureaucracy, and in other issues it may view itself as a lesser stakeholder. Primary stakeholders may well assert primary authority to make decisions, or at minimum, a veto power. Lesser stakeholders may assert that no decision can take place without their consultation and input.
An organization’s resources include not only its personnel and their capabilities and talents, but also a standard set of resources such as budget and personnel, influence, morale, and autonomy, in addition to turf and essence, all of which we will discuss in due turn.
An organization’s self-understanding of what it is and does is crucial to its ability to function effectively. An organization’s sense of identity and mission provides its members with a vision of why what they are doing is important and necessary, and how what they are doing differs from what other organizations are doing. Without this focus and vision, an organization may not develop the special skill set needed to possess influence within the bureaucracy, and it may also lose its ability to instill morale in its members. An organization’s essence will lead it over time to develop a distinctive organizational culture, with norms of dress, behavior, thinking, and value prioritization. A legendary case in point is the differing corporate cultures of Microsoft and Apple. Not only can one tell the employees apart, one can also tell the customers apart! The situation can be even more fraught when military and civilian organizations must work together, and intercultural communication skills may need to be explicitly taught in order to avoid serious problems (Barno and Bensahel, 2016; see also Keane, 2016 for an excellent case study of the impact of this divide on American counter-narcotics strategy in Afghanistan). Even within the diplomatic arm of the government, cultural differences may abide.
As these examples suggest, while the development of bureaucratic identity is virtually inevitable, and potentially a strength, this process also carries attendant risks. The most salient risks are empire-building and interorganizational xenophobia, both of which phenomena may also be observed taking place at suborganizational levels. Though organizations are designed to be tools of a higher-level elected executive, in many ways they are far more powerful than that executive. They are going to last much longer than he or she will; they directly control large sums of money and personnel; they exercise capabilities on the ground and “own” information no one else may have; they are not under electoral accountability. It is not surprising, then, that many governmental organizations begin to act as autonomous entities—empires, almost—that are not in the business of obeying directives so much as in the business as negotiating directives with an eye to their organization’s advantage. One president (FDR) put it this way:
The Treasury is so large and far-flung and ingrained in its practices that I find it is almost impossible to get the action and results I want… . But the Treasury is not to be compared with the State Department. You should go through the experience of trying to get any changes in the thinking, policy, and action of the career diplomats and then you’d know what a real problem was. But the Treasury and the State Department put together are nothing as compared with the Na-a-vy… . To change anything in the Na-a-vy is like punching a feather bed. You punch it with your right and you punch it with your left until you are finally exhausted, and then you find the damn bed just as it was before you started punching. (Eccles, 1951, 336)
One of the most chilling accounts of this disconnect between the executive and a governmental organization was penned by Franklin C. Miller, who was appointed director of Strategic Forces Policy in the Office of the Secretary of Defense in October 1981 and kept that position until 1992. Miller eventually uncovered that much of the Single Integrated Operational Plan (SIOP), the plan guiding deployment U.S. nuclear weapons, was frankly inconsistent with the Presidential Guidance on the use of nuclear weapons. Part of the reason stemmed from organizational culture; General Curtis LeMay, founder of the Strategic Air Command, wanted to ensure maximum autonomy from Washington, going so far as to move SAC’s headquarters out to Omaha, Nebraska. The resulting culture of zealously guarding information from the rest of the Defense Department undermined rational force and operational planning for the U.S. nuclear arsenal to the point where the president would have been left without the use of options specifically ordered. Furthermore, the bureaucratic truce brokered between the Air Force and the Navy involved neither force taking account of the other’s targeting, meaning the stockpile of nuclear weapons was almost twice as large as necessary to ensure full target coverage. It took Miller years to penetrate the bureaucratic fortress that had been erected and reconnect presidential orders with actual operational plans (Butler and Miller, 2016).
In a sense, President Trump points toward some of the same frustrations when he bristles about the influence of what he calls the “deep state.” As William Ruger (2018) explains, Trump and other commentators use this term to refer to the idea that powerful government organizations pursue their own agendas to the point where “the preferences of elected officials and political appointees are subverted, undermining the will of the people and our republican form of government” (Ruger, 2018). An anonymous New York Times editorial in September from a member of the government “resistance” to Trump stoked considerable debate about the degree of the deep state’s influence, the purity of its motives, the extent to which it is merely a politically contrived foil leveraged by Trump and, in fact, whether the “deep state” even exists (Anonymous, 2018). An overlapping and related question is the influence of what has traditionally been called the foreign policy establishment and its neoliberal consensus, which has recently become referred to in the United States as the “Blob” after President Obama’s Deputy National Security Advisor Ben Rhodes used it as a term of derision in a now-infamous interview published in the New York Times Magazine (Samuels, 2016). However, as the foreign policy establishment also includes members of nongovernmental organizations (NGOs) and think tanks, as well as senior members of the foreign policy bureaucracy, we will deal more with the degree to which the “foreign policy establishment” carries sway in FPDM when we discuss domestic politics in chapter 5.
As these examples suggest, essence can breed distrust and resentment of those who are different, whether they be in other organizations, or even within one’s own organization. As noted previously, the infamous antipathy between the FBI and CIA arguably contributed to some of the intelligence failures that led to 9/11. In the wake of that horrific event, the heads of both agencies publicly accused the other of incompetence and noncooperation. Even the intelligence reform of December 2004, with its creation of a director of national intelligence (DNI) and two new interagency intelligence centers, did not stop the bickering between the two organizations (see Zegart, 2005).
But this xenophobia also extends within the organization. Those who are not “like” those who identify with the essence of the organization may be targeted for harassment and even expulsion. Some scholars have used the term cloning to refer to an organization’s tendency to employ only those who embrace the organization’s essence and culture, which may, as a result, become even narrower over time. Hudson once overheard a conversation between two FBI agents, they discussed the dismissal of another colleague. One said, “Yeah, he’d show up to work in sandals and chinos. It’s true he was very bright—possibly the brightest in the office—but there was no way the Bureau was going to keep him. He just didn’t fit in.”
Similar dynamics can also operate at a whole-of-organization level. When Australia’s now-defunct aid agency, AusAID, was integrated with the Department of Foreign Affairs and Trade (DFAT) in 2013, the ensuing “cultural revolution” was described as pitting the “aidies” against the “pinstripes”—the latter a reference to the preferred suits of Australia’s diplomats (Dobell, 2015). After almost forty years of operating independently, these organizations had developed distinct identities and cultures, despite routinely working closely together. The difficulty of assimilating these two cultures was overlooked when the decision was made to integrate, with serious and ongoing repercussions for organizational performance and priorities (Day, 2016). For example, many senior “aidies” took the opportunity to leave DFAT, leaving it bereft of experienced international development experts.
In some circumstances, groups of individuals remain marginalized within their organization, developing what effectively becomes a “sub-identity.” The position of submariners within the Navy has always been precarious, because the essence of the Navy is sailing ships on the water, not under the water. Likewise, the Army was eager to be rid of the Army Air Corps in the wake of World War II because the pilots were seen as undermining the essence of the Army: boots on the ground. Even when the essence seems similar between two organizations, each may view the other as an interloper. So, for example, in the wake of the 9/11 reorganization that created the National Counterterrorism Center (NCTC), Nolan relates how one interviewee explained how the “CIA is more like the old money, whereas NCTC is the nouveau riche [who] will never be accepted into the old money culture … NCTC is like an annoying younger sibling who wants to play the game but isn’t capable of playing it at the same level. But your parents force you to play with the younger sibling” (2013, 83). What’s at issue in such cases is not essence, but turf.
Essence will help shape “turf,” referring to the substantive and skill domains in which the organization believes it has a primary claim to influence and expertise (and hence authority) within the national bureaucracy. As we have just noted, sometimes an organization’s essence leads it to shun or treat lightly particular turf that it sees as unimportant or subversive to that essence. But much more often than not, organizations are greedy for additional turf, and zealously guard what turf they already possess. The reason is simple: more turf means a larger sphere of influence, more personnel, a larger budget, perhaps even greater autonomy. Losing turf means a concomitant loss in each of these areas. Thus, though the Navy and the Air Force do not view sealift and airlift in support of the Army as expressing their respective essences, the two services resist efforts by the Army to create its own lift capabilities, such as the TSV (theater support vessel). Turf battles over close air support of troops between the Air Force and the Army and over amphibious operations between the Army and the Marines are long-standing and legendary.
New missions will create turf battles, often fierce. After 9/11, President Bush determined that a Department of Homeland Security (DHS) should be created to handle the newly urgent mission of protecting the American homeland from direct attack, particularly from terrorism. Tom Ridge would be the first director, but first the various functions determined to be pertinent to the new mission of DHS would have to be put together in one piece. Andy Card, the president’s chief of staff, recalls how the twenty-six heads of executive entities that would need to cede turf to DHS were assembled in the Roosevelt Room. Each person stated that their agency or department was the answer, and no DHS was necessary. No one would cede any turf at all. In the face of this opposition, Ridge was ready to resign. Card got permission to assemble a small group of individuals in the PEOC (the President’s Emergency Operations Center), which had no windows. This group was from non-DHS-relevant agencies like OMB (Office of Management and Budget), and personnel from the top, middle, and even lower levels of bureaucratic rank were tapped. The group started with a blank sheet of paper and were asked what would the ideal DHS look like? After this exercise, the group then determined which battles would never be won, such as trying to put the FAA (Federal Aviation Administration) under DHS, or eliminating the Department of Transportation. These were taken out of the plan. But they put into the bureaucratic game plan the battles they thought they could win, such as obtaining the Coast Guard, the TSA, obtaining Immigration from Justice, and so forth. The famous (or infamous) spaghetti drawing that pictured what parts of what executive entities would comprise the new DHS was the result, and what now exists as DHS looks very much like the game plan drawn up in the PEOC stipulated. But without circumventing turf battles in this rather secretive and circuitous fashion, it is unclear what would have become of the idea of a DHS.1
Issues of turf can also determine access to information within the bureaucracy. Since access to information is a form of power and control, fights over such access can become especially intense. Such “siloing” of information arguably made it easier for the 9/11 hijackers to carry out their plan, as we have seen. Information access ties in directly to influence: an organization cannot afford to have its policy stances ignored because it is perceived as not knowing what is really going on. However, communication between agencies that need to be coordinating their efforts can be daunting in the extreme. Bridget Nolan recounts her experience at NCTC, which was created to overcome the pre-9/11 lack of communication between intel agencies:
Each agency has its own information that is organized into one or more databases, and being assigned to NCTC is not a guarantee that an analyst will be granted access to the necessary databases. In fact, the mere bureaucratic act of transferring to NCTC, even on a one-year rotation, may result in the analyst’s losing access to information systems he had at the home agency, and several weeks or months may pass before the access is reinstated. An analyst’s cubicle features at least two computer screens and anywhere from two to five “pizza boxes,” which are hard drives corresponding to various agencies’ databases. Each “pizza box”—so called because the rectangular hard drives are stacked horizontally—could feature many individual databases, but the box’s presence in a cubicle does not mean that the analyst actually has access to it. Moreover, access to certain systems may be delayed by six months or more, which certainly impedes the analyst’s ability to do the job … Although Sametime, the instant messaging system, was frequently cited as a form of technology that improved communication, its usefulness exists only as long as an analyst physically works at NCTC or CIA because Sametime is a feature of the CIA computing system only. An FBI analyst, for example, could come to NCTC and build a strong network via Sametime, but as soon as the rotation is over those contacts are much more difficult to maintain because there is no Community-wide instant messaging system. (2013, 27–28, 30)
It is instructive that even when a center such as the NCTC has been created to specifically overcome the stove-piping of information, the stove-piping not only remains, but is actually intensified. This suggests the siloing of information is not a bug, but rather an enduring feature of organizations vying for influence within a bureaucracy.
The size of an organization, operationalized as the amount of funds allocated for its budget and the number of personnel assigned to the organization, is a primary indicator of the strength an organization can bring to bear in bureaucratic battles. The budget of the entire CIA is less than one-tenth that of the Pentagon, for example. Civilian employees of the Department of Defense (DoD) alone number over seven hundred thousand, with military personnel adding about 1.4 million more. Estimates of the number of CIA employees range from twenty-five thousand to thirty thousand. Though popular perception, promulgated through Tom Clancy novels and the like, might lead one to conclude that the CIA is on an organizational par with the DoD, nothing could be further from the truth. Compared to the CIA, the DoD is an eight-hundred-pound gorilla, and the social dynamics of interagency working groups reflect this. For example, when ODNI (Office of the Director of National Intelligence) was first created, its creators envisioned it setting priorities and overseeing the budgets of all the intelligence agencies, including those of DoD. How far do you think that idea went, given that ODNI is even smaller—much smaller—than the CIA?
With regard to budget, it is also worth remembering that relative budget increase is as important to track as total budget figures. The proportion of the armed services budget that goes to each of the three major services is arguably more an issue of contention between the services than is total level of funding. Often a wary peace develops where entities keep bureaucratic conflict under control by a de facto agreement to keep budgets proportionality static, no matter which direction total figures track, up or down. This conflict-avoidance measure can readily undercut the ability of the secretary of defense to make significant alterations in the nation’s fighting force.
One of the most important objectives of any governmental organization is influence: influence with policymakers and comparative influence on matters affecting one’s turf within the bureaucracy. One of the crucial elements upon which influence depends is access. For example, even though the CIA is a considerably smaller organization than the DoD, it was CIA personnel who provided the president with his daily morning security briefing (the PDB) (until the DNI office was established). This unparalleled access provided the CIA with influence far in excess of what its size would forecast. Now that the office of the DNI has taken over this function, the CIA has lost one source of influence as an organization.
Sometimes influence is obtained not through access to policymakers, but through acquiring an interagency reputation. The very small INR office of the State Department (the Bureau of Intelligence and Research) maintains influence completely out of proportion to its size because it has developed a reputation for skewering the intelligence estimates of its larger sister organizations, particularly the DoD and CIA. Because INR is so small, they have nothing to lose and everything to gain by questioning the estimates of these larger organizations. If the INR is proved right, as they sometimes are, this further establishes their reputation as being hard-nosed objectivists who operate unconstrained by organizational pressures to conform their analyses to the accepted or acceptable wisdom. Nevertheless, it is still true as a generality that the larger and more well-funded an organization, and the larger the scope of its expertise and turf, the more likely that organization will have veto power over other organizations in interagency working groups, or even primary authority to make certain decisions.
Morale, though less tangible an asset than funding or personnel, is still vitally important to organizations. Demoralization can lead to an exodus of personnel, or a decrease in productivity among those who remain. A demoralized organization is in a weaker position within the bureaucracy, and may have to fight harder to retain what influence, turf, and budget it once had. For example, this occurred to the CIA after the scandals and hearings about its covert operations in the 1970s. When William Casey became DCI in 1981, his top priority was to reestablish morale in the agency, for he felt it had lost its “nerve” and had become risk averse, threatening not only its ability to carry out its assigned functions but also constraining its influence. But political appointees can also undercut morale. For example, under the short-lived tenure of Secretary of State Rex Tillerson in the Trump administration, Tillerson made plain to the employees of Foggy Bottom in 2017 his intention to slash State’s budget by a third and eliminate up to 2,000 diplomatic positions. The backlash was swift; hundreds of diplomats signed petitions protesting foreign policy positions endorsed by Tillerson, publicly embarrassing him. But “scores” of diplomats also resigned or retired, some being told their positions were being eliminated, others removed from their positions but never reassigned. One anonymous State Department employee commented of this time period, “There’s furniture stacked up in the hallways, a lot of empty offices. The place empties out at 4 p.m. The morale is completely broken” (quoted in Filkins, 2017). This hollowing-out of the State Department coincided with a severe loss of influence until Tillerson was replaced.
However, while good morale is important, sometimes organizational attention to morale can take unusual, sometimes counterproductive, forms. Halperin recounts how it was issues of morale that led the Army to implement shorter tours of duty for officers than enlisted personnel during the Vietnam War (Halperin, 1974). Officers who aspired to a long career in the Army needed combat experience to qualify for field grade rank. The Army felt that providing combat experience for the maximum number of officers possible would thus boost morale. Unfortunately, it led in some cases to resentment by seasoned enlisted personnel of “green” officers looking for glory and willing to engage in risky operations to get it. There were reports that especially gung-ho junior officers were as much at risk from their own platoons as they were from the Viet Cong.
Morale is also linked to what is considered a success by the organization, as well as who is allowed to take credit for that success. Unfortunately, the organization’s take on success may not fit well with the task at hand. Norton and Stigler (2015) point out that the pre-9/11 FBI was particularly unsuited for counterterrorism work. The FBI was extremely decentralized with 56 different field offices that operated fairly autonomously; success was defined in terms of arrests and convictions instead of prevention; and only the first field office involved with a case would be given charge over it, making other offices reluctant to contribute to something they could never take credit for. This modus operandi was ill-suited for unified and coordinated action against terrorist threats to the nation.
It is very difficult for two or more organizations to jointly plan an operation. Each has a different culture, different skills, different procedures, different equipment, and different priorities. Furthermore, each is vying with the others for influence and turf in matters where these overlap between organizations. Thus, one objective of organizations is to operate as autonomously as possible. An excellent example of this was the political jockeying over the creation of the DNI position, alluded to previously. The 9/11 Commission, which spurred the creation of this new position, wanted the DNI to have budgetary authority over all intelligence units scattered throughout the federal bureaucracy, as well as the power to set priorities for intelligence gathering by these units. The major opponent to this conceptualization of the DNI’s power was the DoD, naturally enough: these proposed DNI powers would severely cut into its autonomy. Mustering its influence and resources, the DoD fought and won the concession that military requirements could override DNI requirements when the lives of American military personnel were “at stake.” Given the DoD’s preference for autonomy, we would expect that exceptional condition to become a chronic condition.
Combined with this understanding of what drives organizations, it is now important to understand how large organizations operate. At their most fundamental, organizations exist to reduce complexity. There are several aspects to this complexity: complexity of information processing and decisionmaking, complexity of task execution, and the complexity of coordinating the efforts of the organization’s numerous human employees.
The attack an organization makes on complexity is a simple one: break up a complex whole into pieces that are easily understandable, easily executable, and easily standardized. In a way, the last thing an organization really wants to do is have to think about something from scratch. More efficient is to view something new as an instance of something already known, or something new to do as an extension of something the organization already does. This approach is not irrational in the least: remember that typical government organizations may have hundreds, thousands, or even hundreds of thousands of human employees. And these employees are not static over time; on any given day, some employees are leaving, some are staying, and some are entering employment with the organization for the first time. No one human being within the organization can know all there is to know about it. No one human being possesses complete institutional memory concerning what the organization has done in the past. No one human being has the skills and know-how that the complete organization has. If the organization is to function, such global knowledge must be made as irrelevant as possible.
Though the organizational approach is not irrational, it is decidedly different from what we consider to be normal behavior for a human being, where global knowledge is prized. Let us consider some of the major differences.
Organizations are simply not very responsive to change. Inertia is a strong force within organizations, which may result in a lack of creativity, a lack of flexibility, and a lack of adaptability to new circumstances. The National Security Agency (NSA) admitted it had hundreds of hours of captured pre-9/11 conversations among individuals suspected of having planned or taken part in those attacks that still had not been translated months afterward because it did not have enough Arabic translators. The FBI spent several years and over $170 million to update its computerized file management systems to allow easier dissemination of information across units, only to scrap the entire project and decide to start all over. Nearly two years after the 2003 invasion of Iraq, Special Operations forces were finally given permission to pay field informants cash. Armor for Humvees and body armor for soldiers were not provided in sufficient quantities for the Iraqi invasion because the working assumption was that most troops would not encounter enemy forces. The notion of a hard-core insurgency that would attack American troops anywhere within Iraq, even within “secured areas,” was apparently not a scenario seriously considered during contingency planning. Even the Israeli intelligence establishment refused to consider the possibility that Bashar al-Assad of Syria would attempt to build a nuclear reactor until four years after he assumed power in 2000. The director of the Ministry of Intelligence purportedly considered such a move “illogical,” and therefore “due to inertia, Israel’s intelligence remained mired in the old conception for several years” (Herzog, 2018).
Responsive learning can be painfully slow, imperiling important priorities. Usually incremental learning is the norm for large organizations, where baby steps toward change are undertaken over a significant time period. The most reliable guide to organizational action at time t is organizational action at time t – 1. For example, the journalist Fred Kaplan notes it took twenty-one months after 9/11 for the DoD to come up with a nineteen-page planning document to improve language skills pertinent to the war on terror. This document called for another eleven months to come up with guidance to create new programs, thirteen months to come up with an index to measure readiness in language, sixteen months to establish a database of current language capabilities, nineteen months to enunciate language requirements, twenty-eight months to disseminate a language aptitude test, thirty-seven months to establish crash courses for deploying personnel, and forty-nine months to create a personnel information system containing data on language skill. By forty-nine months after the original planning document, no actual language training programs outside of the crash course for deploying personnel would actually have been established. Kaplan points out that seventy months after 9/11 we still were not yet offering additional language training to meet national priorities—almost six years! He notes it took far less time than that for the Americans to enter World War II and help defeat the Axis powers, and far less time for America to undertake profound reforms after the Soviet launch of Sputnik in 1957. When the time period is this extended, incremental learning almost becomes no learning at all (F. Kaplan, 2005).
Organizations interpret orders according to their existing understandings and capabilities, which results in an implementation gap between what policymakers believe they have ordered and what organizations actually do to execute such orders. March and Simon have called this the “logic of appropriateness,” where actions are chosen on the basis of pattern recognition from knowledge already stored in the system (March and Simon, 1993). For example, when John F. Kennedy ordered the Navy to quarantine Cuba, the Navy heard “blockade,” because that was the closest match in their knowledge base. But there were several key differences between what Kennedy wanted the Navy to do and what the Navy thought a blockade entailed. For example, the Navy wanted to force Soviet subs in the area to surface and determined to sink ships that refused to stop or be boarded. Kennedy did not want either to occur. Fortunately, Kennedy was able to recognize these differences and intervene to clarify in very precise terms what would and would not happen during the quarantine.
Organizations develop standard operating procedures (SOPs) in place of thinking through every new situation from the ground up. However, in addition to simple mismatch of definitions, as noted above, there is also the possibility that the existence of an SOP has short-circuited acknowledgment of obvious extenuating circumstances, resulting in wildly inappropriate behavior on the part of the organization. In his book Essence of Decision, Allison recounts such a case concerning the camouflage of Soviet intermediate-range ballistic missiles (IRBMs) and medium-range ballistic missiles (MRBMs) during the Cuban missile crisis. The missiles were extremely well camouflaged during transportation and unloading at Cuban ports. However, once dispersed to their construction sites, the missiles were not camouflaged at all. They were constructed in the very same configuration as missile sites in the USSR, allowing for easy identification from U-2 imagery. Some U.S. officials even speculated that the USSR wanted the United States to know about the presence of these missiles as they were being emplaced. That was not the case, however. According to Allison, the excellent in-transit camouflage was due to the efforts of Soviet intelligence. But once ashore, the missiles were placed under the Group of Soviet Forces in Cuba, whose commander placed them in control of his staff from the Strategic Rocket Forces (SRF). Now, the SRF had never placed missiles outside of the Soviet Union. Here they were, thousands of miles away from the USSR on a small tropical island. What to do? What they knew how to do: SOP for missile placement in the Soviet Union, which SOP did not include camouflage but did include a standard configuration for the silos. After the Americans announced they had discovered the presence of the missiles, camouflage was hastily improvised. The Director of Central Intelligence (DCI) at the time, John McCone, could not help but wonder how much worse the situation would have been if the missiles had not been discovered before the IRBMs could be made operational. Fortunately, we will never know.
SOPs also create an explicit chain of command. The degree to which hierarchy permeates decisionmaking within an organization has been related by scholars to both the organization’s culture and the culture of the larger society in which it is embedded. In some cultures, “jumping” the chain of command can be grounds for termination. Even serious questioning of a superior’s decisionmaking assumptions or information, let alone the actual decision, may cause career disruption. Though all members of the organization in a sense comprise the brain of the organization, possessing some knowledge that may not be duplicated in the knowledge base of others, some brains may be more valued than others. Unfortunately, it tends to be those most removed from the “ground” whose judgment prevails. This creates the undesirable circumstance in which higher-level decisionmakers within the organization may not even know what they lack in terms of important information about a particular situation. And subordinates may feel discouraged from bringing this lack to their superior’s attention, for fear of personal repercussions. This catch-22 is, of course, the basis for federal and state protection of “whistle-blowers.”
Organizations are motivated primarily by factors discussed above, such as essence, budget, influence, and autonomy. These will not be sacrificed for the sake of executing orders or requests for information issued by policymakers. For example, there is no doubt that organizational reporting on the situation in Vietnam during the Vietnam War was inhibited by the memory of the “China Hands” in the State Department who had been sacked during the McCarthy era for having written the truth about the relative strength and popularity of the Communist and Nationalist forces. Organizations were very leery about reporting the weaknesses of the South Vietnamese regime or the strengths of the North Vietnamese forces. Outright falsification was not necessary: tone and emphasis, coupled with strategic omissions, could hide the truth well enough. That such “altered” reporting did nothing for the quality of U.S. decisionmaking during this era was not uppermost in the calculations of these organizations.
In conclusion, then, organizations are necessary to government. Yet, organizations produce unintended negative consequences on a regular basis and often at the most inopportune moments. How can foreign policymakers use organizations without being undermined by them? First, it is crucial that leaders, usually through their staff, delve into the arcane structure and SOPs of organizations through which they are trying to implement policy. In this way, leaders can work with SOP rather than against it, by finding appropriate units and more closely matched SOPs within the organization and steering executive orders in that direction. Second, leaders can try to force a change on an organization through budgetary “feast or famine.” Offering more money to do something new can be attractive to an organization. Taking money away—especially if it upsets budgetary “truces” between organizational units—can also be a catalyst for change. Leaders can also be alert to scandal and egregious failure within an organization, which can be the justification for extreme change. For example, the Federal Emergency Management Agency (FEMA) will probably not survive as an influential organization, given its abysmal handling of the Hurricane Katrina disaster in 2005. Third, a leader can use turf wars to his advantage, by putting two or more organizations in competition and tying factors like turf, budget, and personnel to the outcome of that competition. Finally, a leader can give up and create a new organization to do what the old organization cannot or will not do. This was a major consideration in the creation of the Directorate of National Intelligence. In the end, leaders cannot do without organizations and must be prepared to deal with them on their own terms in order to effectively use them—and not be used by them.
An excellent way to see how these principles play out when violated is the extensive report on the Columbia shuttle disaster of 2003 (NASA, 2003). Though not a foreign policy case per se, the investigative report is one of the best we have seen in terms of identifying important issues of organizational culture and their intersection with organizational process and tasks. The entire crew of the Columbia space shuttle was lost on February 1, 2003, when their reentry vehicle disintegrated. The accident was caused by a breach in the wing after a large piece of foam dislodged and struck the wing eighty-one seconds into launch. While the foam strike was noticed in the launch footage two days after the launch, NASA ultimately treated the foam strike as an event that would not compromise flight safety.
How did NASA come to this faulty conclusion? The following excerpts from the Columbia Accident Investigation Board (CAIB) Report provide a tragic summary of that organizational decision, making plain the problems inherent in organizational decisionmaking. (Background information and some paraphrased transitions are indicated using italics, rather than by using brackets, to avoid interrupting the flow of the text.) Indeed, the CAIB explicitly states, “In our view, NASA organizational culture has as much to do with this accident as the foam. Organizational culture refers to the basic values, norms, beliefs, and practices that characterize the functioning of an institution. At the most basic level, organizational culture defines the assumptions that employees make as they carry out their work. It is a powerful force that can persist through reorganizations and the change of key personnel. It can be a positive or a negative force” (chapter 5, 97). As you read these report excerpts, pay close attention to how the panel, which included scholars of organizational behavior, points to several of the factors that we have discussed in order to explain the tragedy.
Upon learning of the debris strike on Flight Day Two, the responsible system area management (United Space Alliance and its NASA counterpart) formed a team—the Debris Assessment Team—to analyze the debris strike in accordance with mission rules requiring the careful examination of any “out-of-family” event, meaning an event that could affect the safety of the flight. However, Mission Management did not designate this team as a Tiger Team (meaning a group tackling problems that could cause significant alteration of the mission). Because of its lack of such designation, no subunit “owned” the Debris Assessment Team, and its status was in a rather limbo-like state.
Using film from the Intercenter Photo Working Group, a NASA group that provided photos to the Marshall, Kennedy, and Johnson Centers, Boeing systems integration analysts prepared a preliminary analysis that afternoon. (Initial estimates of debris size and speed, origin of debris, and point of impact would later prove remarkably accurate.) As Flight Day Three and Four unfolded over the Martin Luther King Jr. holiday weekend, engineers began their analysis… . Debris Assessment Team members judged that the actual damage would not be as severe as predicted because of the inherent conservatism of a mathematical model that was used, called the Crater model, and because, in the case of tile, this mathematical model does not take into account the tile’s stronger and more impact-resistant “densified” layer, and in the case of RCC (reinforced carbon-carbon material), the lower density of foam would preclude penetration at impact angles under 21 degrees.
On Flight Day Five, impact assessment results for tile and RCC were presented at an informal meeting of the Debris Assessment Team, which was still operating without direct Shuttle Program or Mission Management leadership. Mission Control’s engineering support, the Mission Evaluation Room, provided no direction for team activities other than to request the team’s results by January 24. As the problem was being worked, Shuttle managers did not formally direct the actions of or consult with Debris Assessment Team leaders about the team’s assumptions, uncertainties, progress, or interim results, an unusual circumstance given that NASA managers are normally engaged in analyzing what they view as problems. At this meeting, the Debris Assessment Team’s participants agreed that an image of the area of the wing in question was essential to refine their analysis and reduce the uncertainties in their damage assessment.
Each member supported the idea to seek imagery from an outside source. However, because of the team’s limbo-like status, without guidance from the Mission Management Team or Mission Evaluation Room managers, the Debris Assessment Team did not know how it should get the imagery, and so chose an unconventional route for its request. Rather than working the request up the normal chain of command—through the Mission Evaluation Room to the Mission Management Team for action to Mission Control—team members nominated Rodney Rocha, the team’s Co-Chair, to pursue the request through the Engineering Directorate at Johnson Space Center to the Department of Defense. As a result of the team moving outside of what Mission Management thought of as proper channels—even though the team arguably had no proper channels because it was not designated a Tiger Team by Mission Management—even after the accident the Debris Assessment Team’s request was viewed by Shuttle Program managers as a non-critical engineering desire rather than a critical operational need… . (above paragraphs from p. 167).
In the meantime, unbeknownst to the Debris Assessment Team, there had been two other requests for DoD imagery. One came from the boss of some members of the Debris Assessment Team, who asked the Space Shuttle Integration Office to contact DoD. That call was made, and though no formal request was made, DoD began to organize for taking such imagery. The Debris Assessment Team then reiterated its request to the Shuttle Engineering Office at Johnson Space Center, constituting a third request.
At 8:30 a.m. the same day, the NASA Department of Defense liaison officer called US STRATCOM and cancelled the request for imagery. The reason given for the cancellation was that NASA had identified its own in-house resources and no longer needed the military’s help. The NASA request to the Department of Defense to prepare to image Columbia on-orbit was both made and rescinded within 90 minutes. Upon further investigation, the Board found what it believes happened during that 90-minute period. The head of the Mission Management Team heard about the second request and asked who was requesting the imagery—because the request had not gone up the line through Mission Management. Several individuals were queried, and these individuals all stated that they had not requested imagery, were not aware of any “official” requests for imagery, and could not identify a “requirement” for imagery. The head of Mission Management later told several individuals that nobody had a requirement for imagery.
However, when the second image request was cancelled by Mission Management, the unforeseen consequence was that the first and third image requests—both made by the Debris Assessment Team—were cancelled as well. Interestingly, while the head of Mission Management publicly stated she did not know of the Debris Assessment Team members’ desire for imagery, she never asked them directly if the request was theirs, even though they were the team analyzing the foam strike. This speaks to the status, or lack thereof, of the very body tasked with determining if there was a serious problem or not.
Also on Flight Day Seven, the head of Mission Management, Linda Ham, raised concerns that the extra time spent maneuvering Columbia to make the left wing visible for imaging would unduly impact the mission schedule; for example, science experiments would have to stop while the imagery was taken. According to personal notes obtained by the Board: “Linda Ham said it was no longer being pursued since even if we saw something, we couldn’t do anything about it. The Program didn’t want to spend the resources.” Shuttle managers, including Ham, also said they were looking for very small areas on the Orbiter and that past imagery resolution was not very good. The Board notes that no individuals in the Columbia mission’s operational chain of command had the security clearance necessary to know about National imaging capabilities. Additionally, no evidence has been uncovered that anyone … sought to determine the expected quality of images and the difficulty and costs of obtaining Department of Defense assistance. Therefore, members of the Mission Management Team were making critical decisions about imagery capabilities based on little or no knowledge… . (above paragraphs from pp. 153–54).
The Debris Assessment Team was demoralized at the rejection of their request for the needed imagery. Debris Assessment Team members speculated as to why their request was rejected and whether their analysis was worth pursuing without new imagery. Discussion then moved on to whether the Debris Assessment Team had a “mandatory need” for Department of Defense imaging. Most team members, when asked by the Board what “mandatory need” meant, replied with a shrug of their shoulders. They believed the need for imagery was obvious: without better pictures, engineers would be unable to make reliable predictions of the depth and area of damage caused by a foam strike that was outside of the experience base.
However, team members concluded that although their need was important, they could not cite a “mandatory” requirement for the request. Analysts on the Debris Assessment Team were in the unenviable position of wanting images to more accurately assess damage while simultaneously needing to prove to Program managers, as a result of their assessment, that there was a need for images in the first place.
After the meeting adjourned, the member of the Debris Assessment Team that had made the first and third request, Rodney Rocha, read the 11:45 a.m. email from his contact at the Johnson Space Center, Paul Shack, which said that the Orbiter Project was not requesting any outside imaging help. Rocha called Shack to ask if Shack’s boss, Johnson Space Center engineering director Frank Benz, knew about the request. Rocha then sent several emails consisting of questions about the ongoing analyses and details on the Shuttle Program’s cancellation of the imaging request. An email that he did not send but instead printed out and shared with a colleague follows.
In my humble technical opinion, this is the wrong (and bordering on irresponsible) answer from the SSP and Orbiter not to request additional imaging help from any outside source. I must emphasize (again) that severe enough damage (3 or 4 multiple tiles knocked out down to the densification layer) combined with the heating and resulting damage to the underlying structure at the most critical location (viz., MLG door/wheels/tires/hydraulics or the X1191 spar cap) could present potentially grave hazards. The engineering team will admit it might not achieve definitive high confidence answers without additional images, but, without action to request help to clarify the damage visually, we will guarantee it will not. Can we talk to Frank Benz before Friday’s MMT? Remember the NASA safety posters everywhere around stating, “If it’s not safe, say so”? Yes, it’s that serious. [SSP—Space Shuttle Program, MLG-Main Landing Gear, MMT-Mission Management Team]
When asked why he did not send this email, Rocha replied that he did not want to jump the chain of command. Having already raised the need to have the Orbiter imaged with Shack [i.e., in the third request for imagery], he would defer to management’s judgment on obtaining imagery… . (above paragraphs from p. 157)
Mission Control personnel thought they should tell Commander Rick Husband and Pilot William McCool about the debris strike, not because they thought that it was worthy of the crew’s attention but because the crew might be asked about it in an upcoming media interview. Director Steve Stitch sent the following email to Husband and McCool and copied other Flight Directors… . (158)
“The impact appears to be totally on the lower surface and no particles are seen to traverse over the upper surface of the wing. Experts have reviewed the high speed photography and there is no concern for RCC or tile damage. We have seen this same phenomenon on several other flights and there is absolutely no concern for entry”… . (159)
When the Debris Assessment Team made their report the morning of January 24, there was standing room only, as all team members apparently felt some emotion regarding what was happening. However, they failed in creating any sense of urgency or danger in the Mission Management attendees. At the Mission Management Team’s 8:00 a.m. meeting [on January 24, when a final decision about the return flight was to be made subsequent to the final report of the Debris Assessment Team], Mission Evaluation Room manager Don McCormack verbally summarized the Debris Assessment Team’s brief. It was the third topic discussed. Unlike the earlier briefing, McCormack’s presentation did not include the Debris Assessment Team’s presentation charts. In addition, as analyzed by Edward Tufte, the legendary scholar of visual representation, the PowerPoint slide actually used that morning did not visually communicate any sense of urgency and the text was unintentionally misleading [p. 191]. The Board notes that no supporting analysis or examination of minority engineering views was asked for or offered, that neither Mission Evaluation Room nor Mission Management Team members requested a technical paper of the Debris Assessment Team analysis, and that no technical questions were asked… . (161)
According to a Memorandum for the Record written by William Readdy, Associate Administrator for Space Flight, Readdy and Michael Card, from NASA’s Safety and Mission Assurance Office, discussed an offer of Department of Defense imagery support for Columbia. This January 29 conversation ended with Readdy telling Card that NASA would accept the offer but because the Mission Management Team had concluded that this was not a safety-of-flight issue, the imagery should be gathered only on a low priority “not-to-interfere” basis. Ultimately, no imagery was taken… . (166)
After the Challenger disaster, NASA had contracted out to SAIC the role of providing third-party safety personnel to all meetings to raise concerns that were being overlooked for one reason or the other. [S]afety personnel were present during these key meetings about Columbia but passive and did not serve as a channel for the voicing of concerns or dissenting views. Safety representatives attended meetings of the Debris Assessment Team, Mission Evaluation Room, and Mission Management Team, but were merely party to the analysis process and conclusions instead of an independent source of questions and challenges. Safety contractors in the Mission Evaluation Room were only marginally aware of the debris strike analysis. One contractor did question the Debris Assessment Team safety representative about the analysis and was told that it was adequate. No additional inquiries were made. The highest-ranking safety representative at NASA headquarters deferred to Program managers when asked for an opinion on imaging of Columbia. The safety manager he spoke to also failed to follow up. (170)
Engineering experts at Kennedy and Marshall, the other space centers, were anxious enough about the foam strike to consider how the Columbia crew could fix the impact damage, or barring that, whether the crew could be rescued by sending up Atlantis, or whether Columbia could crash-land in the water. But all of the misgivings of the engineers, stated and unstated, were of no consequence: Mission Management did not see that the foam strike caused a serious safety issue. Columbia was instructed to reenter according to the original flight plan, resulting in the catastrophic loss of the entire crew and the shuttle.
Several comments by the Investigation Board are worth contemplating by the student of organizational process. Below, we share a few of these:
Another factor that enabled Mission management’s detachment from the concerns of their own engineers is rooted in the culture of NASA itself. The Board observed an unofficial hierarchy among NASA programs and directorates that hindered the flow of communications. The effects of this unofficial hierarchy are seen in the attitude that members of the Debris Assessment Team held. Part of the reason they chose the institutional route for their imagery request was that without direction from the Mission Evaluation Room and Mission Management Team, they felt more comfortable with their own chain of command, which was outside the Shuttle Program. Further, when asked by the investigators why they were not more vocal about their concerns, Debris Assessment Team members opined that by raising contrary points of view about Shuttle mission safety, they would be singled out for possible ridicule by their peers and managers. (169)
Communication did not flow effectively up to or down from Program managers. As it became clear during the mission that managers were not as concerned as others about the danger of the foam strike, the ability of engineers to challenge those beliefs greatly diminished. Managers’ tendency to accept opinions that agree with their own dams the flow of effective communications. (169)
After the accident, Program managers stated privately and publicly that if engineers had a safety concern, they were obligated to communicate their concerns to management. Managers did not seem to understand that as leaders they had a corresponding and perhaps greater obligation to create viable routes for the engineering community to express their views and receive information. This barrier to communications not only blocked the flow of information to managers, but it also prevented the downstream flow of information from managers to engineers, leaving Debris Assessment Team members no basis for understanding the reasoning behind Mission Management Team decisions. (169) … Managers’ claims that they didn’t hear the engineers’ concerns were due in part to their not asking or listening. (170)
After Program managers learned about the foam strike, their belief that it would not be a problem was confirmed (early and without analysis) by a trusted expert [Calvin Schomburg] who was readily accessible and spoke from “experience.” No one in management questioned this conclusion. (172)
Managers asked “Who’s requesting the photos?” instead of assessing the merits of the request. Management seemed more concerned about the staff following proper channels (even while they were themselves taking informal advice) than they were about the analysis. (172)
In both the Mission Evaluation Room and Mission Management Team meetings over the Debris Assessment Team’s results, the focus was on the bottom line—was there a safety-of-flight issue, or not? There was little discussion of analysis, assumptions, issues, or ramifications. (172)
Organizations with strong safety cultures generally acknowledge that a leader’s best response to unanimous consent is to play devil’s advocate and encourage an exhaustive debate. Mission Management Team leaders failed to seek out such minority opinions. Imagine the difference if any Shuttle manager had simply asked, “Prove to me that Columbia has not been harmed.” (192)
Notice in the account and in the comments of the Investigation Board several of the factors we have discussed previously about organizational process: the inflexibility of SOPs, the chilling effect of hierarchy, the compartmentalization of knowledge, the indifference by more senior personnel to the resynthesis of that compartmentalized knowledge, the issue of organizational “face” vis-à-vis the Pentagon, the facade of attention to safety belied by the actual organizational culture of “can do.” The full report on the Columbia shuttle disaster is over six hundred pages long and is a testament to the inherent problem of creeping dysfunctionality in large organizations. It is well worth the effort for the foreign policy analyst to peruse this report.
Thus, despite elaborate organizational charts to ensure that all aspects of a problem would be considered, despite overt rhetoric about the importance of safety and speaking up, despite the personnel of NASA being highly accomplished in their respective fields, the same old issues of turf, lack of communication, SOP, and organizational culture directly contributed to the deaths of the Columbia crew. Without the benefits provided by large organizations, there would have been no shuttle program. Without the disadvantages of large organizations, the lives of these astronauts might not have been lost.
Bureaucratic politics is a complex intersection of small group dynamics, organizational process, domestic political forces, and the personal characteristics of relevant individuals. Most bureaucratic politics takes place in interagency groups, which are one of the foremost means for important, but noncrisis, situations to be addressed within government. Though positions taken by the participants in such interagency groups may be roughly predictable, predicting which position(s) will prevail is sometimes possible, sometimes impossible, but always an extremely complex calculation. Though important matters are generally tasked to an interagency group to develop a series of options or recommendations for higher-level small groups, such as the NSC, to address, it is still likely that the interagency group is not only subject to influence attempts by the participating organizations, but also vulnerable to domestic political pressure and even electoral imperatives. Further complicating matters is the impact of diverse personalities assigned to the interagency group, as well as underlying networks of friendship and conflict that enmesh these personalities. In short, bureaucratic politics produce the most intriguing soap operas to be found in government. While the game of international relations may be played according to national interest, there is also a second game being played within each government, a game of personal and/or organizational interests and ambitions, which may in fact be more determinative of a nation’s foreign policy than the game of national interest. I’m tempted to call it the “game of small thrones.” Allison and Zelikow put it this way:
Choices by one player (e.g., to authorize action by his department, to make a speech, or to refrain from acquiring certain information), resultants of minor games (e.g., the wording of a cable or the decision on departmental action worked out among lower-level players), resultants of central games (e.g., decisions, actions, and speeches bargained out among central players), and foul-ups (e.g., choices that are not made because they are not recognized or are raised too late, misunderstandings, etc.)—these pieces, when stuck to the same canvas, constitute government behavior relevant to an issue. To explain why a particular formal governmental decision was made, or why one pattern of governmental behavior emerged, it is necessary to identify the games and players, to display the coalitions, bargains, and compromises, and to convey some feel for the confusion. (1999, 257)
Some key concepts help us frame the dramas, large and small, produced by bureaucratic politics:
Stakeholders, sometimes called “players,” are those whose roles, expertise, turf, or sheer political power coupled with strong interest allow them to affect a bureaucratic outcome. Powerful stakeholders may be able to claim authority within a decision context, or, in other settings, claim the right to be consulted during decisionmaking, or even claim the right to veto certain types of decisions. Stakeholdership itself may be the subject of politicking; powerful stakeholders may attempt to block the claims of other stakeholders with less power. For example, well-credentialed government nuclear scientists propounding that current nuclear warheads are not reliable and must be replaced have been disinvited from key interagency meetings where the future of the U.S. nuclear arsenal is discussed. Thus the very composition of interagency groups, and other issues such as the chairmanship of such a group, are subject to political jockeying. In general, sheer political power trumps claims to stakeholdership based on role, and role trumps claims to stakeholdership based on expertise. For example, Congressman Dan Burton, the grandfather of a child with autism, was able to force the Food and Drug Administration (FDA) to reinvestigate links between thimerosal in childhood vaccines and autism, but the FDA was simultaneously able to effectively marginalize the views of physician-researchers who felt they could show such a link empirically. But those with political power are not always victorious; there are cases where the relatively powerless have prevailed, and we will deal with these in the section on “equalizers.”
Another generalization about stakeholders is the adage “where you stand depends upon where you sit,” implying that at least in the case of role of stakeholders, organizational affiliation will largely determine the stance taken in bureaucratic negotiations. This may inevitably produce long-standing bureaucratic rivalries and competitions that outlast the tenure of any careerist in the organizations. For example, in interagency discussion between the FBI and CIA, we are not surprised when the one argues for greater powers vis-à-vis the other. Furthermore, we are not surprised when outsiders demand greater cooperation between the two organizations and try to institutionalize that through standing interagency “centers,” such as the National Counterterrorism Center. But then we are also not surprised when assignment to such centers is regarded as the kiss of death for one’s career within one’s home organization.
Those of us who work in large bureaucracies know that the only way to be an effective player is to know the action channels—whom to see and where to go and what to do to make something happen. For example, just to make something trivial happen at Hudson’s university—getting a new key to a new office—requires that she finds the proper form, obtain the signature of her chair and her dean, and walk the form over to a particular obscure building on the margins of campus to pay a fee and get the key. Changing from PC to Mac at the university office? Hudson must give a statement to her chair saying why the change is needed, her chair must write a statement justifying Hudson’s justification, the result must be forwarded to the college computing committee by a particular date, and the committee must in turn relay its decision to the comptroller who buys the equipment.
At least Hudson navigated the process far enough to be able to document the required steps. Day was recently informed that in order to install a requested piece of software on his laptop, he would first be required to install the software on his office desktop computer. However, when he organized for this to happen, IT staff informed him that his office desktop was too old and slow for the software to work. When Day was told the fix—ask his Department to buy him a new desktop computer—he gave up on obtaining the software he had originally wanted (for his laptop, no less). We are all familiar with the plethora of procedures and committees facing us when attempting to do most anything within the bureaucracies of our universities.
So it is within government and the foreign policy establishment. Though it is always instructive to look at organizational charts, “boxology” does not tell you how to actually get something done. For example, how do you declare that the official U.S. government position is that Saddam Hussein has weapons of mass destruction (WMD)? This is actually quite complicated. The president just can’t say, “Saddam Hussein has WMD.” No, the president asks the DCI if Saddam Hussein has WMD. The DCI asks the Intelligence Community Executive Committee, which asks the National Foreign Intelligence Board, which asks each of its member intelligence organizations to independently answer the question. After each intelligence organization hashes out its own answer, interagency committees are set up to debate the answer among agencies. The resulting opinions and minority opinions and dissenting opinions will then be sent to the board, which will discuss them and send them up to the Executive Committee. The Intelligence Community Executive Committee will further discuss the issue and then make a report to the National Intelligence Council. The NIC will make their own investigation of all the facts and analysis put forward by the intelligence community. At some point, the particular member of that office charged with oversight of the broad issue area of proliferation will issue a National Intelligence Estimate (NIE). That official NIE is then presented to the president, who can now say, “Saddam Hussein has WMD.” If you don’t know the action channels, you cannot act.
Those who study bureaucracy are often reluctant to call the outcomes of bureaucratic politics “decisions.” After all the stakeholders have pulled and hauled to the best of their power in a particular direction, what is left over is better seen as something less than a decision, which term connotes some processual rationality. Resultant connotes that the outcome would probably not coincide with one chosen by any unitary rational actor. It is usually the lowest common denominator outcome: the outcome upon which a majority of the participants in the process can agree. In general, of course, unless there is a threatening emergency, most resultants can be characterized as incremental change based upon a papering over of key differences. The vaguer the proposal, the greater the convergence of agreement around it.
Effective political players within large bureaucracies not only know all the action channels—they are also masters at bureaucratic manipulation. The most important tools of manipulation, especially if one can occupy a position of authority within the group such as a chairmanship, are the use of framing, rules, deadlines, information control, and agendas to obtain one’s desired ends.
Framing is a process by which a group comes to understand a situation and define its decisionmaking task. Framing is not only a psychological process for an individual; when it involves persuasion of group members to adopt one’s frame, framing also becomes a very political act. As Entman explains in his classic article on the subject, “To frame is to select some aspects of a perceived reality and make them more salient … in such a way as to promote a particular problem definition, casual interpretation, moral evaluation, and/or treatment recommendation for the item described” (1993, 52). So, for example, some groups refer to a fetus as “uterine material” while others consider it a “preborn person.” Likewise, some framed the contras in Nicaragua during the Reagan administration as “freedom fighters” and others, “terrorist guerrillas.” Was Iran exercising its rights under the Nuclear Nonproliferation Treaty (NPT) with its uranium enrichment program or undermining the NPT? (For a good overview of the role of framing in FPDM, see Mintz and DeRouen, 2010, Chapter 8.)
Ryan Beasley (1998) notes that framing may actually be more important to study in bureaucratic politics than the final decisionmaking process, for choice is constrained by the frame adopted by the group. Beasley finds that a particular frame is more likely to be adopted if it is simple, if it is backed by a strong leader or a member of the group that can claim special expertise in the area, and if it lends itself to a fairly clear-cut course of action. Another aspect is whether the frame of action can be characterized as an incremental outgrowth of what has been done in the past. Frames, once adopted, tend to “set” fairly quickly, and it may take the addition of new personnel to the bureaucratic mix to rethink a long-standing frame.
A famous example of “a frame not taken” occurred near the beginning of ExCom’s deliberations during the Cuban missile crisis. When ExCom was presented with the photographic evidence that missile silos were being placed in Cuba, Robert McNamara, the secretary of defense, opined that any such missiles would have little military significance. As such, they would not be worth taking forceful action that would risk a nuclear war. McNamara had the expertise to make such a claim, and yet his frame was swiftly rejected by Kennedy. Kennedy felt that the Soviets’ move had great political consequences, ranging from the fate of Berlin to his own electoral prospects. Kennedy’s strong opinion that the missiles were a grave threat would frame the rest of ExCom’s meetings.
The rules under which the group operates are also an extremely significant factor in understanding group behavior. Consider the differences between a bureaucratic group that operates by majority rule and one that operates on the principle of unanimity. In the former, coalitions will be important; in the latter, every single individual can be a deal breaker. A group under rules of unanimity will probably make fewer and less specific decisions than a group with rules of majority voting. But voting itself can become quite complicated. For example, in the U.S. Congress, parliamentary rules are coupled with rules on filibuster, cloture, committee passage before floor vote, attachment of bills to other bills for vote, necessity of two-thirds majority for particular votes and for overturning vetoes, reconciliation of House and Senate versions of the same bill, and so forth. A legislator who has mastered the rules by which Congress works is at a significant advantage over one who has not. Other types of group deliberation rules include weighted voting, such as in the International Monetary Fund (IMF); and permanent versus nonpermanent status, such as in the UN Security Council.
In addition to deliberation rules, there are many other types of rules to consider. Certain groups may be given the power to initiate hearings or investigations. Rules may govern which entities have the right to appeal to higher deliberative bodies. Laws and the precedent of court cases may specify the line between legal and illegal behavior in certain cases. Regulations may determine who must and who cannot participate in a certain decision context. Some rules that may seem innocuous are actually quite important, such as the power to keep the calendar for a group. One way to see that power of the calendar is to consider the influence of deadlines on decisionmaking.
Deadlines can literally shape group decisionmaking. Less powerful members of the group can use the deadline as leverage to extract concessions from more powerful members. On the other hand, more powerful members can use the deadline to paint others as obstructionists who are likely to cause the group to miss its deadline. Deadlines can force premature closure of discussion on an issue, but on the other hand, deadlines can also create incentive to compile as much information as quickly as possible in an attempt to carry the discussion and sway undecideds before the deadline occurs.
Timing is also a crucial element of bureaucratic play. There are windows of opportunities for ideas, but they open and close at unpredictable moments (Kingdon, 1995). One must be prepared ahead of time to take advantage of them, because they do not last long. Even if your idea is only tangentially related to some newly perceived need, it may have a far better chance of being considered by powerful players if linked to that need. Heymann notes, rightly, “A proposal will be most successful when it is responsive both to a perceived problem and to the valued political opportunities of elected officials” (2008, 106). Waiting to pounce as soon as this unpredictable “alignment of the stars” takes place requires an almost superhuman level of persistence, but when the alignment occurs, the new wind in your sails may take the idea very far in a relatively short period of time. Kingdon (1995) labeled these precipitous moments “policy windows” and referred to the individual actors who were skilled at identifying and making use of them as “policy entrepreneurs.” Studying the behavior of policy entrepreneurs and their role in shaping foreign policy decisionmaking through their influence is one of the places where FPA most readily crosses over into the field of public policy. Clearly the policy entrepreneur concept is a way to link two levels of analysis: the individual and domestic levels are connected by the actions of a skilled bureaucratic operator who knows not only what bureaucratic levers to pull, but where those levers are and the time it is best to pull them. While the study of policy entrepreneurs has occurred mainly in relation to domestic politics, important steps are gradually being taken to apply this concept to FPDM, including contributions by Heymann (2008), Carter and Scott (2009), Blavoukos and Bourantonis (2012), Mintrom and Luetjens (2017), and Davies and True (2017).
The control of information is a recognized art in bureaucratic politics. It may range from a subordinate telling his or her superior what the superior would like to hear all the way up to outright denial of information to those who have the right to know, usually through exclusion of particular players from meetings in which the information is discussed. Probably the most legendary use of denial of information was ANSA Zbigniew Brzezinski’s feud with Secretary of State Cyrus Vance during the Carter administration. The primary means of undercutting Vance was to exclude him from meetings or hold meetings when he was out of the country. Brzezinski also explicitly told his staff not to tell the State Department about important developments, such as the proposed normalization of relations with the People’s Republic of China. Brzezinski even managed to meet with a Chinese envoy at the White House while keeping it a secret from the secretary of state (Rothkopf, 2006).
The manipulation of group agendas is a related skill that is highly prized in the political arena. In most groups, the chairman decides the agenda, but in some groups the group may actually vote on the agenda. The reason the agenda may become political is that it determines the course of group discussion. Items may be purposefully not placed on the agenda so that they will not be discussed, for example. But other types of manipulation are possible. The chair may set a time limit on the discussion of each item, which may allow him or her to cut off discussion of a contentious issue before all have had the opportunity to speak. This is a common tactic in public hearings where citizens or other interested parties may wish to speak on contentious issues. Another tactic is to allow lengthy discussion of items placed first on the agenda, and thus limit or even prevent any discussion of issues coming later in the agenda.
Unless there is near unanimity on a particular issue, most group interactions become examples of coalition politics at work. Within the constraints of rules and deadlines, the group is usually tasked with making some type of determination or decision. This requires getting agreement among enough group members so that a particular determination or decision carries the day.
The bureaucratic player first attempts to build coalitions through the use of personal relationship networks, which may be the most expeditious way to begin building a front. Philip Heymann recounts the pivotal role of the personal relationship between Senator Orrin Hatch and Michael Pillsbury, a deputy of the Office of Planning and Policy at DoD during the Reagan administration. According to Heymann’s account of Pillsbury’s quest to get Stinger missiles into the hands of the Afghan mujahideen, that one connection, forged when Pillsbury was a staff member of the Republican Steering Committee of which Hatch was a member, provided Pillsbury with several crucial successes, including Pakistani and Chinese cooperation in the effort, which in turn begin to alter the political terrain in Washington, D.C., on the Stinger issue (Heymann, 2008). It is also possible to use chains of personal relationships, where one gains access to friends of friends and acquaintances of acquaintances; at one point in the Stinger saga, Pillsbury persuaded Hatch to phone William Casey, then DCI, with important information that enabled Casey to openly favor the Stinger shipment. This type of access is priceless.
Outside of personal networks, there are generally three ways to assemble a coalition. The first is through compromise, where a minimum winning coalition is built around a position with which coalition members feel comfortable, if not completely satisfied. The second is through quid pro quo arrangements where support on Z’s pet issue A by member Y is linked to support on Y’s pet issue B by member Z, ensuring a win-win scenario for all. The third is through implicit or explicit coercion, where a particular faction uses intimidation, threats, media attention, manipulation of rules, or other means to wilt any opposition to or possible compromise of their preferred position. Needless to say, the first two types of coalition-building efforts are comparatively more stable than the last because those who voted for the particular position have no vested interest in seeing it fail.
A large part of the complexity of coalition building is that each coalition member has multiple interests, and therefore the membership of a particular voting coalition has the potential to change as new or different interests are perceived to be at stake. Likewise, particular individuals in the coalition may play multiple roles within the government. For example, does the secretary of state represent the president or the State Department? The answer may depend on the issue at hand and may also be subject to change as circumstances change, as Marsh finds with the 2009 Obama surge decision (2014). In that particular case, the most proximate advisors to the president (who opposed the surge) lost out to a unified but somewhat unusual coalition of the military and the secretary of state.
Though the individual cog in the bureaucratic machine may have very little power, there are time-honored tactics that can help level the playing field somewhat. Let us suppose you are a middle-level bureaucrat who strongly disagrees with the direction adopted by those at a higher level. What could you do?
Actually, quite a lot. First, you could simply not implement the directives you have been given, without raising a fuss. Oftentimes, officials in high positions may not have the time to check that each and every one of their directives has been carried out. If queried, one could blame overriding circumstances for an unforeseen delay. You could also do something different from what you have been ordered to do, and if questioned suggest that a misunderstanding occurred. You could implement cosmetic, not substantive change, or obey the letter but not the spirit of the orders. Or you could implement your orders in an overzealous fashion so as to showcase the faults you see in the directive.
There are other approaches that can be taken. You could insist upon a personal hearing before implementing your orders and suggest reasons for reconsidering. You could make it known that you are keeping a detailed paper trail and journal of what is happening. You could attempt to make your directives public, either by going to the media, to Congress, to another government, or by writing your own book about the situation. You could resign, or at least threaten to resign. Jack Goldsmith is an instructive case of someone who resigned very well. Jack Goldsmith came on board in the fall of 2003 at the Department of Justice as the head of the Office of Legal Counsel (OLC) after the departures of both Jay Bybee, his predecessor in that position, and John Yoo, who had worked in the OLC and had drafted the infamous 2002 “torture memo” justifying the use of torture in the interrogation of enemy combatants in the war on terror. In April 2004, the Abu Ghraib scandal broke, and then two months later the “torture memo” was leaked to the press. Goldsmith decided he must use his position to “withdraw” (effectively nullify) the memo. But he feared the White House would overrule his decision. What to do?
So [Goldsmith] made a strategic decision: on the same day that he withdrew the opinion, he submitted his resignation, effectively forcing the administration to choose between accepting his decision and letting him leave quietly, or rejecting it and turning his resignation into a big news story. “If the story had come out that the US government decided to stick by the controversial opinions that led the head of the Office of Legal Counsel to resign, that would have looked bad,” Goldsmith [said]. “The timing was designed to ensure that the decision stuck.” (Rosen, 2007, 44)
This is not to say that subversion is always the right thing to contemplate. There are certainly times when subordinates taking matters into their own hands is exactly the wrong thing to do: think, for example, of the human rights violations at Abu Ghraib. But there are sometimes when the actions individuals may take on their own initiative may improve the performance of their government. Consider three examples of individuals playing decisive roles in avoiding nuclear war. In October 1962, during the Cuban missile crisis, Second Captain Vasili Arkhipov was able to convince his commander to await instructions from Moscow rather than launch nuclear torpedoes from the Soviet submarine B-59, casting the sole negative vote out of the three officers needed to launch (Lewis et al., 2014, 9). Almost exactly a year later, in October 1963, Robert H. Johnson almost single-handedly prevented the U.S. government from launching a preemptive attack against nascent Chinese nuclear installations by producing the first of two authoritative reports (the second was published in April 1964) suggesting the Chinese program would not significantly alter the military balance (Gady, 2017). And in September 1983, Lieutenant Colonel Stanislov Petrov decided that an early warning indicator that went off (twice) on September 26, 1983 indicating an U.S. ICBM launch was a false alarm and did not immediately notify his superiors (Lewis et al., 2014, 13).
Halperin offers this equally heartening non-nuclear example from the memoirs of Henry S. Villard, a foreign service officer (FSO) who was ambassador to Libya back when that nation had a king:
The Libyan Prime Minister had resigned and flown off to Rome, his nerves frayed by the thankless task of guiding a newborn state. The King was ill, in seclusion; there was a rumor in the bazaars that he might abdicate. The whole government structure seemed about to collapse. I had just reached a vital point in negotiations for an air-base agreement. So when the Libyan cabinet asked me to fly to Italy and persuade the Prime Minister to return, I cabled the Department urgently for permission to make the try.
Time was of the essence, yet the hours ticked by without response. In Washington, the wheels ground methodically. Committee met with committee, weighing the pros and cons of my recommendation. The Pentagon had to be consulted. Policy factors had to be considered; so did tactics, in light of the progress to date on the air-base negotiations. Suggestions at a lower level had to be referred to a higher level for further discussion. I sent a second cable. No reply.
Finally, I decided to act on my own. I boarded the plane of my Air Attache, flew to Rome, and called on the Prime Minister at his hotel. With all the eloquence I could muster, I urged him to come back and steer the ship of state through the storm, pointing out that the fate of his country—and our delicate negotiations—rested in his hands alone. He heard me in silence, still smarting from the political wounds which had caused him to resign. He would think it over; he would give me his answer that evening.
At eight o’clock I was again at the Prime Minister’s door. His face was wreathed in smiles. He would do as I asked, and to mark the occasion he invited me to dine with him downstairs. With a load like lead off my mind, I was enjoying the repast when I spied an officer of our Rome Embassy discreetly waving a piece of paper from behind the potted palms. I made my excuses, rose, and went over to receive the message—a priority cable to Tripoli, repeated to Rome for information. At long last, Washington had moved. There were my orders. Under no circumstances was I to follow the Prime Minister to Rome for that, the Department feared, might be interpreted as interference in the domestic affairs of a sovereign country. (Halperin, 1974, 277–78)
In seeking to understand bureaucratic politics, it must also be recognized that many games are being played simultaneously, and the set of players in any one game only partially overlaps the set of players in another. At the most microlevel, there may be clashes of personality or will between two or more individuals. There may be conflicts between different offices within one organization. There may be a struggle between two or more organizations within a bureaucracy over turf or budget. There may be a contest for influence among the president’s closest advisors. The larger electoral context between political parties is always a backdrop, and in election years may move to the foreground. Interparty factionalism can also be consequential in the positioning of political parties. And then there are the games in the international arena played out between allies, rivals, NGOs, intergovernmental organizations (IGOs), and so on. In other words, just identifying stakeholders in a particular issue is not enough. One must know how many boards a stakeholder is playing on and who the other stakeholders on each board are.
Furthermore, the pivotal meetings are often not those that include the top decisionmakers, but rather meetings that take place at lower levels, or even informally. In the bureaucratic politics perspective, one has meetings with top decisionmakers only after the sausage has already been made, the ducks are all in a row, and the top leader only needs to nod his head in passing after realizing that all his closest advisors agree on the direction in which the nation should move. However, those highest-level formal meetings are routinely the anticlimax to the more interesting story that preceded it, such as we saw when DHS was created by the Bush administration in the windowless PEOC without any of the primary bureaucratic stakeholders. Especially given the explosive growth of the U.S. national security apparatus, Cooper, Gvosdev, and Blankshain argue that existing FPA research “may not give sufficient attention to the role and influence of officials working at lower bureaucratic echelons” (2018, 521). They suggest we need to look deeper at what they call the “sub-bureaucratic level of analysis” if we seek to understand where key FPDM influence germinates in some cases, given “there is a foment of political continuously happening beneath the surface” (539). What that saying, in essence, is that bureaucratic games take place not only at the top level of bureaucracies, but at all levels at all times.
While all of these battles are worthy of study, some are more consequential than others. That is because there is a “second game,” which is not about the topic at hand, but about something more visceral. Heymann (2008) asserts that in order to see the “second game,” we need to probe how various bureaucratic players understand the “significance of the occasion.” That is, what role do actors believe they have been called upon to play? And how do the actors assess what that will mean for them, personally, in the context of the positions they occupy? In a sense, says Heymann, the “second game” is really about respect, for “players who feel they are being denied respect correctly recognize that they are threatened in their entitlement to play the role they have been assigned and that they value” (136). These may be the most fateful games of all, for they are played with a special intensity and even pettiness. For example, in 2014 when President Obama tapped retired Marine General John Allen to serve as special envoy to the anti-ISIS coalition, the head of CENTCOM (U.S. Central Command), General Lloyd Austin, was not amused, especially since Allen would report directly to Obama, bypassing CENTCOM and even the JCS. Allen requested air transport from CENTCOM to the region for a meeting, and Austin turned him down flat, instructing his staff to tell Allen to ask the State Department. Allen was left “steaming” (Perry, 2014).
Because the most important games may not center around foreign policy at all even though foreign policy is the battlefield on which they are played, it is hard not to come to the conclusions that the “resultants” of bureaucratic politics tend to be penny-wise and pound-foolish more often than not. One of the most exemplary cases in this regard played out under the administration of George W. Bush, and is explicated below.
In order to see some elements of bureaucratic politics in action, we will examine a particular case from the 9/11 time period. The New York Times published a series of articles in 2004 that detailed how a new system of military justice was created in the wake of the 9/11 attacks (Golden, 2004a, 2004b). This system was used to detain suspected terrorists at Guantánamo Bay, Cuba, in a military prison setting. One of the chief lightning rods of the system was the assertion that the men detained did not possess rights as prisoners of war under the Geneva Convention. Over time, this new military system came under attack from many quarters, including the military’s own lawyers.
The assertion of the Times is that bureaucratic manipulation to achieve long-standing ideological aims on the part of key players was the engine driving the creation of this new system. In this recounting, we will refrain from assessing ideological motives and concentrate on the analysis of elements of groupthink, organizational process, and bureaucratic politics. Pay close attention to who “sat” where, who knew whom, who knew what, who was included, who was excluded, and how perceived domestic political imperatives affected the process.
The cast of players included Timothy Flanigan, deputy White House counsel; John Yoo, in the Justice Department’s Office of Legal Counsel; William Barr, the former attorney general when Flanigan served as head of that same office; David Addington, counsel to the vice president; Alberto Gonzales, White House counsel; Pierre-Richard Prosper, the State Department’s ambassador-at-large for war crimes issues; Patrick Philbin, a deputy in Justice’s Office of Legal Counsel; William J. Haynes II, general counsel to the secretary of defense; and John Bellinger, legal advisor to the National Security Council, along with a bevy of higher-ranking officials and lower-ranking attorneys.
The events of September 11, 2001, set the stage for the United States-led war on global terrorism. A key question was how the United States could adopt an aggressive stance toward terrorism and yet negotiate the U.S. legal system, which provides many rights to accused persons, and the international legal system, which also provides significant rights to prisoners of war under the Geneva Convention. The best legal minds in government would be tasked with reconciling what on first glance appeared to be irreconcilable.
The White House counsel’s office became the locus of initiative concerning the development of a new legal paradigm for the war on terror. Flanigan was apparently assigned the lead on this assignment. Flanigan contacted Yoo, a friend, who wrote a twenty-page reply opining that in the context of terrorist attacks, Fourth Amendment rights might not apply.
Flanigan then placed a call to his old boss, William Barr, to ask advice. Barr apparently reminded him that the Justice Department had researched the idea of “special military tribunals” to oversee trials of suspected terrorists almost ten years previously when Pan Am 103 had been blown up over Scotland. Flanigan felt that such military tribunals, later reworded as military “commissions,” would strike precisely the right posture in the new global war on terror. As commander in chief, it would ultimately be the president who would control what these commissions did.
At some point, Flanigan apparently shared his ideas with Addington and Gonzales, and both concurred. Gonzales decided to establish an interagency working group to hammer out options concerning the prosecution of terrorists—already knowing which option he would try to ensure prevailed. Pierre-Richard Prosper from State was assigned to chair the group, and according to the Times account, it was made clear to him by Gonzales that military commissions would be one of the options.
The Prosper interagency group saw three alternatives for prosecuting terrorists: federal courts, military tribunals, and Nuremberg-style tribunals with both military and civilian members. The Justice Department’s representatives to the group insisted that federal courts were adequate. The various counsels from the White House were united in their disagreement. After the options had been researched and debated for approximately a month, the White House pulled the plug on Prosper’s group, and Flanigan was again in charge of developing the new legal framework.
This time the framework would be worked out among the various White House counsels before it was revealed to any other agencies. This is a very risky bureaucratic maneuver. Leaving out whole hosts of lawyers situated across a dozen relevant agencies and departments would virtually invite attack. As we will see, the most damning attack would come from those lawyers who were asked to actually implement the framework’s particulars.
On November 6, 2001, Patrick Philbin in the Justice Department’s Office of Legal Counsel sent, by request, a thirty-five-page confidential memorandum to Gonzales. In it, citing a 1942 case where Franklin D. Roosevelt ordered on his own authority a military tribunal to try eight Nazi saboteurs, Philbin argued that the president had the inherent authority to set up the desired military commissions. He further argued that rights of due process would not necessarily apply in the context of war (including the war on terror).
Based on this memorandum, the various counsels at the White House drafted an executive order, which was apparently approved by John Ashcroft, the head of the Justice Department, and also Donald Rumsfeld, secretary of defense (through his counsel William J. Haynes II). Interestingly, it had been the criminal division of the Justice Department that had argued against military commissions in the Prosper interagency group. How did Ashcroft overcome their opposition? He did not. Ashcroft simply did not tell Michael Chertoff, the head of Justice’s criminal division, about the new order. Chertoff, who later became secretary of homeland security, only saw the orders when they were published. Ditto for the State Department and even the National Security Council.
In the meantime, a group of Army lawyers had tried to meet with Haynes to prevent a fait accompli. Probably sensing that not meeting with them at all would be contrary to public relations interests, Haynes called their leader into his office on Friday, November 9, and allowed him to review the proposed order for exactly thirty minutes. He was not allowed to take notes, according to the Times report.
The next day, Saturday, the Army’s judge advocate general called together a group of senior military lawyers for an emergency meeting. Their purpose was to draft a response that would result in modifications to the order before it was published. But that very same day, the vice president, the attorney general, Haynes, Gonzales, Flanigan, Addington, and others were finalizing the order. The Times reports that Dick Cheney felt the order should not be shown in advance to Colin Powell, secretary of state, or Condoleezza Rice, the ANSA. The vice president and the president discussed the order over a lunch, and the president signed the order on Tuesday, November 13. No press conference was held.
In bureaucracies, however, as we have discussed in this chapter, “faits” are only “accomplis” when play has ceased—or at least become dormant—on the multiple boards of play. The maneuvering of Flanigan and others to make only one board, the White House board, count was doomed to failure.
The Senate Judiciary Committee immediately called for hearings. (Ironically, according to the Times account, the administration tasked Prosper and Chertoff to represent the administration’s view, even though both men had argued against the policy and eventually were excluded from deliberations.) The Department of Defense parried this new attack in preemptive fashion by leaking the draft concerning implementation of the new system, indicating that critics’ concerns had been taken into account. Rumsfeld also assembled a group of external legal experts to offer advice, and some of these held credibility for having worked on the Nuremberg and Tokyo tribunals.
For a moment, it appeared that play had stalled, and the administration’s gambit had worked. However, it would turn out that the Pentagon had overlooked a very important game board. It was not the Senate or the American Civil Liberties Union (ACLU) that the Pentagon should have worried about. It was their own lawyers, military lawyers, over whom they should have lost sleep. Unfortunately, the approach that Haynes took toward the military lawyers was exclusionary. In one exchange reported by the Times, the Navy judge advocate general, Admiral Guter, confronted Haynes directly: “ ‘We need more information.’ Mr. Haynes looked at him coldly. ‘No, you don’t.’ ” Guter would retire soon thereafter, and then sign a “friend of the court” brief on behalf of Guantánamo detainees appealing their detention.
In the meantime, a new issue had been put into play. Could detainees appeal their detention in federal court? Numerous critics had argued detainees must have this right, and then of course the federal courts would judge whether the new legal framework of military commissions was constitutional. The White House team of lawyers saw this chain of reasoning for what it was: a bureaucratic Trojan horse designed to derail the entire military commission idea. Philbin and Yoo from Justice were again tasked with providing relevant legal arguments, this time that detainees could not make such an appeal. Their memorandum dated December 28, 2001, suggested an overseas detention site in order to argue that the detentions were not taking place on American territory. Guantánamo was chosen in accordance with this logic. The first detainees would arrive on January 11.
Furthermore, the White House legal team, again turning to Justice’s Office of Legal Counsel for support, had argued that the Geneva Conventions did not apply to terrorists. Yoo had argued, and Gonzales and Addington concurred, that even the Taliban could be considered terrorists. In fact, even if interrogators could not identify any link to terrorism per se, detainees would be held as “enemy combatants,” with the identity of the enemy force left undefined.
At this point, however, excluded players began to emerge and make their presence felt. Condoleezza Rice wondered why the National Security Council and its legal team had not been involved. Colin Powell complained that given the number of allied nations involved in the situation, State had to be in the loop, too. The FBI and the criminal division of Justice had their own complaints.
In order to reconstruct unity among his bureaucratic players on these important issues, President Bush asked two of the NSC’s staff, including legal counsel John Bellinger, to bring the players together and have them work out the kinks in an interagency committee. Apparently, however, the various players began asking some rather difficult questions, such as how Defense knew these people were enemy combatants. Defense’s first position against such probing was to stonewall. One former Defense official told the Times that “he and others went into interagency meetings on Guantánamo with a standard script, dictated by their superiors: ‘Back off—we’ve got this under control.’ ” Since Defense was following the November order drafted by the White House legal team and approved not only by the president, but also the powerful vice president, this tactic worked—for a while.
According to the Times, in August 2002, the ANSA, through the NSC, made her move. Rice’s NSC staff sent its own Arab-speaking representative, reportedly a “senior intelligence analyst,” to Guantánamo to assess conditions and speak to detainees. His or her report was given to Rice, and the report was purportedly very damning of what appeared to be a completely ad hoc operation. Rice took it to Powell. She also took it to Tom Ridge, advisor to the president on Homeland Security. And, in the coup de grâce, she took it to the criminal division of Justice. She began to build a counterforce to Rumsfeld and Cheney on the issues of detainment and military commissions.
On October 18, members of the cabinet involved with national security affairs met in a high-level showdown. Rice and Powell argued that what was going on in Guantánamo was not what the president had had in mind. They called for most of the detainees to be released. Rumsfeld apparently backed down. He was not interested in being a jailer; he was a warrior. Rumsfeld agreed to brief other agencies about the situation at Guantánamo and agreed that the other cabinet members had the right to approve or disapprove plans for prosecution or release of the remaining detainees.
This last promise was to become the Trojan horse that the White House team had effectively warded off earlier. Now Justice, State, the NSC, the FBI, and other agencies all had to agree to a particular detainee’s prosecution before Defense could proceed. As the Times puts it, “The internal struggle over the prisoners’ fate began to play out in dysfunctional weekly meetings at which officials from across the government assembled by secure video link to consider individual detainees put forward by the Pentagon for outright release or transfer to the custody of their home governments.” Readers of this chapter will not be surprised to learn that these dysfunctional weekly meetings produced almost no transfers, releases, or prosecutions.
Months later, in the spring of 2003, the military commissions had still not tried even one case. But after the Supreme Court agreed to hear a case challenging the legality of the detentions, the Pentagon decided to move forward with a few prosecutions. But they had underestimated their own lawyers.
Military lawyers assigned to defend the detainees took an aggressive stance. They filed a “friend of the court” brief with reference to the aforementioned Supreme Court case. They publicly challenged Pentagon rules that they were not to speak with the media. One military defense lawyer filed suit in a federal district court to block the military commissions.
On June 28, 2003, the Supreme Court ruled that detainees had the right to petition federal courts for their freedom. After that, a significant number of detainees were transferred to the custody of their home governments, where many were simply released from custody. The military commission framework never became fully operational. In July 2006, the Supreme Court ruled that military tribunals had to be explicitly authorized by legislation adopted by Congress before they could be formed, and the White House conceded that the detainees would retain their rights under the Geneva Convention. And when William Haynes was nominated to the federal bench, a whole host of military lawyers signed a letter to Congress urging his nomination be rejected.
This case study is a fascinating tale of groupthink, polythink, organizational behavior, and bureaucratic politics all rolled together into what ended up a colossal policy failure. Consider the personal ties that permitted members of the White House counsel team to work effectively with certain members of the Justice Department, perhaps initiating groupthink. But consider further how intraorganizational cleavages within Justice and Defense undermined the resultant policy. Examine also how tactics to exclude potential naysayers from process, from information, and from access were effectively used in the short term, but then backfired over time. Keep in mind the roles played by the various branches of government, with moves by the executive branch affected by the opening of Senate hearings and rulings by the Supreme Court. Note also the role of organizational essence, with the Pentagon eventually deciding that it was not in the penitentiary business. Do not overlook the role of public embarrassment as military lawyers and judges voiced their open opposition to the plan. Consider finally the deeper context of the second game played among Rice, Powell, Rumsfeld, and Cheney for personal influence and access. Finally, reflect upon the fact that the end stage of interagency meetings, where all naysayers were included, predictably resulted in a de facto gutting of the policy through sheer inability to reach consensus, a polythink situation if ever there was one. This episode offers the foreign policy analyst an insightful glimpse into the complex levels of group forces at work in foreign policy decisionmaking.
★ ★ ★
Virtually all FPDM occurs in groups. Thus it is no coincidence that several of the most influential, celebrated, and cited works in FPA—Allison’s examinations of the Cuban missile crisis (1969, 1971), Janis’s Groupthink (1972), Allison and Halperin’s work on bureaucratic politics (1972), and the first edition of Halperin’s Foreign Policy and Bureaucratic Politics (1974)—deal with group decisionmaking. The fact that the classic works from the classic period of FPA all conceptualize how groups influence FPDM continues to exert considerable influence over the shape of the subdiscipline, but it must be recognized they examine only U.S. cases. This is entirely understandable and justifiable, of course. The decisions made by the United States at this time of history were the most globally consequential, given the preponderance of U.S. power. However, the operation of group dynamics is dependent on domestic political structures. Parliamentary systems operate differently from presidential systems, with resultant differences in how group decisionmaking proceeds. Fortunately, since those early classic works on the U.S. case, we have seen a small, but heartening increased in the number of non–North American works on groups and FPDM. Zhang Qingmin’s (2016) contribution exploring how bureaucratic politics influences Chinese foreign policymaking is but one example (see also Jones, 2017). Klaus Brummer has also made important inroads on the study of bureaucratic politics in the European context (2009, 2013, 2017). In their foundational work on Australian FPDM, Gyngell and Wesley (2007) likewise delve deeply into the foreign policy bureaucracy of that state. We will explore the extension of classic U.S.-centric FPA theory to non–North American nation-states further in chapter 8.
The government, with its small groups, organizations, and bureaucracies, may be the seat of government action, but that seat is embedded in a larger context of culture, society, and politics to which the foreign policy analyst must also pay attention. In chapter 4, we turn to the question of how culture affects foreign policy.