An Introduction to Allison's Wonderland

Bureaucracy and how American foreign policy is made.

Writers note: this is an essay I wrote for an International Relations course. The articles this essay references are “Are Bureaucracies Important? (Or Allison’s Wonderland)” by Stephen Krasner and “Conceptual Models and the Cuban Missile Crisis" by Graham Allison. I encourage anyone who finds this topic interesting to read both articles.

     In the world of foreign policy, analysis of policy decisions is often broken down into abstract models intended to provide a relatively simple and effective way of looking at the complex decisions governments make regarding these issues. In the 1969, Graham Allison outlined three basic models for this analysis; the rational, organizational, and bureaucratic models. In the aftermath of the Cuban Missile Crisis and the United States’ increasing involvement in Vietnam, his bureaucratic theory of policy-making became popular among analysts for explaining why the policy outcomes that come out of Washington do so. In his essay published in 1972, “Are Bureaucracies Important? (Or Allison Wonderland)” Stephen Krasner cautions against this view, arguing that it obscures the influence the President has on final policy outcomes and that it provides elected officials an excuse for policy that is unpopular among the electorate. This essay will argue that Krasner’s criticisms are well-founded and that the bureaucratic model of policy analysis proposed by Graham fails to adequately explain foreign policy decision-making. The bureaucratic model misunderstands the relationship between the presidency and the positions the presidency fills, assumes a legitimacy to subordinate preferences running contrary to presidential wishes that does not exist, and falsely asserts that bureaucratic institutions dealing with foreign policy have an independent inertia.

     The first criticism of Graham’s model is one of hierarchical concerns. Graham asserts that in the “game” of bureaucratic politics, each position represents a player that come to their positions in equal fashion and as such each have ability to determine policy outcomes from their point of view (Graham 709). This assertion ignores the practical reality of both democratic politics and the structures of government. As Krasner points out, the only elected position that exerts influence on decision-making is that of the president, and all other positions derive their authorities from him by virtue of appointment (Krasner 166). The President appoints the other “players” that Graham asserts have influence on policy-making; their values often align with the president’s, and his appointment of them to their respective positions is an endorsement by the president of their preferences. As Krasner notes, “They [individuals appointed to these positions by the president] also feel some personal fealty to the President who has elevated them…” (Krasner 166). This loyalty is often, by virtue of the hierarchical structure, one-sided and dependent on the president’s appointees making decision conducive with presidential values.

     This mistaken assertion by Graham is important to note because it is connected to the next point this essay examines. Inherent in Graham’s bureaucratic model of policy analysis is an assumption that policy decisions that run contrary to the president’s wishes are legitimate. As Krasner correctly notes, however, disagreements in policy proposals and outcomes between the president and his appointees will, at some point, lead to the dismissal of the appointees (Krasner 166). Graham’s discussion of bureaucratic politics as a game independent of presidential choices ignores the fundamental reality of executive power structures; all power descends from the president, and actions that run contrary to presidential values and choices are illegitimate use of that power. Graham writes, “Action does not presuppose intention…rather separate individuals with different intentions contributed pieces which compose an outcome distinct from what anyone would have chosen” (Graham 711). This ignores or excuses the nature of executive policy-making; a choice is always required, and it is always the president’s choice that matters. Graham himself implicitly notes this in his discussion of the early days of the Cuban Missile Crisis when he notes that McNamara, upon realizing that Kennedy was dead-set on presenting a strong response, pivoted to supporting the blockade as an alternative to an air strike (Graham 715). What he reads into this as an example of bureaucratic pressures on the president—he describes a coalition that pressured Kennedy into supporting the blockade as opposed to an air strike—is not a matter of bureaucratic choices, but one of rational decision based on values. Krasner notes that the supposed “ability” of bureaucracies to establish policies independent of the president are a function of presidential inattention (Krasner 168). This is a value choice, just as attentions paid by the president are value choices. Kennedy leaving Washington in the early days of the crisis and allowing the blockade coalition to form is a representation of these value choices, and the decision that was borne of it is a result of presidential preferences.

     Tied to the point above is this essay’s third critique of Graham’s model. His model presupposes independent inertia by these bureaucratic agencies. The assumption that these bureaucracies (at least those dealing with foreign policy concerns) have an impetus to act outside presidential purview once again ignores the power of the president to set or not set policy. Krasner notes this in his assertion that bureaucratic policy choices are determined by presidential attention, which is itself a byproduct of presidential value choices (Krasner 168). Bureaucracies can put into motion policies without explicit presidential consent if they so choose, but their ability to do so is inherently dependent on implicit consent from the president through ignorance of the policy. On matters where the president does direct his attention, bureaucracies must bend to his will. Kennedy micromanaged the crisis to the best of his abilities; his attentions were focused entirely on the issue, and the bureaucracies he commanded implemented policies entirely with his approval, whether explicit or tacit. Attempts to do otherwise (such as the Navy’s obstinance in the face of what they saw as presidential meddling in their affairs) were ultimately overruled. As Krasner shrewdly observes, “Policies which violate Presidential objectives may bring Presidential wrath” (Krasner 169). Krasner also points out that the bureaucratic structures that Graham believes restrain presidential action can be changed or undone with an exerted presidential focus. This logic holds considerable weight on this particular matter, because Graham’s assertion of independent bureaucratic inertia depends upon these structures as self-propelling. Should the president choose not to exert his energies, and instead allow these procedures to fulfil themselves, he is himself putting these policies in motion with his tacit approval of the procedures (and therefore the policies they may produce) already in place. Graham attempts to argue that the policy choices presented to Kennedy, and the one he ended up choosing, were a result of independent impetuses within the bureaucracy as well as external pressures over which Kennedy had no choice. As Krasner notes, and as this essay has laid out, every action requires choice. Kennedy was not forced into the blockade by forces he had no control of; his concerns with the air strike proposal, as well as his personal worries of impeachment, were both value choices he himself weighed and chose to act in response to. They were not forced upon him, and had he chosen a different outcome a different outcome would have arisen. As is often said colloquially; the buck stops with the president. The buck also starts with him, ultimately.

     All of these points flow to a proposition not directly asserted by Krasner but implied in his logic; rational choice theory is a superior method of analyzing policy outcomes due to its adherence to the notion that, ultimately, value choices are the prime determinant of policy. The president is the final decision-maker on all policy fronts, whether through direct action or indecision, and to argue otherwise is to ignore a fundamental truth; all actions, whether individual or bureaucratic, are ultimately value-based. The power the president holds over these bureaucratic structures is something that must be reckoned with at all times when attempting to explain policy choices, because it is from the president that bureaucratic authority on foreign policy matters stems and is put into action.

     In the course of this essay, Allison Graham’s bureaucratic model of policy analysis is critiqued through examination of Stephen Krasner’s article on the matter. Because it broadly dismisses individual choice as a determining factor, Graham’s model fails to understand how a president and his subordinates interact, confers a legitimate position to bureaucratic decisions that run counter to presidential wishes, and makes the false assumption that bureaucracies have independent impetus on these matters. Understanding these models and recognizing the importance of value judgements, especially those that the president makes concerning foreign policy, is as vitally important today as it was in the 1960s. In the always-complex world of foreign policy, the value judgements elected officials make should not be discounted. To do so risks abdication of responsibility, both by the officials and by the citizens that elect them.

Scott Howard is an alumnus of the University of Florida and executive editor of the Freemen Newsletter.