Allison then explains the paradigm for the organizational model, which he notes is not in the existing literature, and therefore is Allison’s creation. The unit of analysis is government action as organizational output. Here Allison uses the analogy of a game of chess. Model I states that the players would be national government, who move the pieces (military, diplomats, economy) at will, according to strategies, with winning the game being the goal. For the Model II analyst, the setup is different. The pieces are still the governments organizational units, their movements signifying actual occurrences. The national actor cannot move them as he pleases, however, because the way in which each piece can move is limited by the SOPs of that unit (a military unit will not engage in diplomacy, for instance), and the unit’s SOPs limit the options of the leader. Thus, the outputs each unit are capable of narrow the options of where they can be placed on the chessboard. The dominant inference pattern of the Model II paradigm is time. Allison argues that to understand an organizational actor at a given time (t), the analyst must study the organization at t-1. In other words, you can best guess what an organization is capable of today, by studying what the organization did yesterday. Also, if you want to best guess how an organization will look or act tomorrow (t+1) you must study them at the present . Also important in this paradigm is the treatment of SOPs, which allow large organizations to act in concert, and also provide the leadership with an idea as to their options . Allison notes that this paradigm will improve on Model I by delving deeper into how organizations come to decisions, and he calls for more research in this area. Behavioral studies of government organizations, studies of the tendencies of those groups, as well as looking at the sources of organizational changes will all help to explain the SOPs (and by extension, the outcomes) of governmental organizations .
He finishes the third chapter by outlining how the application of this paradigm will increase the clarity of the missile crisis. One of the major questions unanswered by Model I is why the Soviets would make an offensive strategic move while in the UN, they seemed to be pursuing a relaxation of relations with the US. To the Model I analyst, this incongruence was unexplainable. The Model II analyst, as Allison notes, can marry these two notions, because in a large organization such as a national government, it’s possible that the “left hand and right hand of a government” are acting independently of each other .
The fourth chapter is the second look at the missile crisis, through the prism of the Model II paradigm. He once again divides the chapter along the major three pivots, or events, of the missile crisis. In discussing the Soviet decision to deploy the missiles, he addresses the inconsistencies mentioned above that Model I was not able to deal with. One that illustrates Model II well is the issue of Soviet failure to camouflage their missile sites. American intelligence could not believe that the Soviet, who had an American U-2 plane in their possession, and thus knew a bit about our reconnaissance capabilities, did not hide the missile construction sites in Cuba. Had they done this, it is entirely possible that we would have never discovered the missiles. The other problem, from the US perspective, was that the surface-to-air missile (SAM) batteries, which would have been able to detect and shoot down a U-2 were not operational before the Soviets started building the missile sites. This baffled our intelligence services, mainly because from a rational perspective, it didn’t add up. Allison uses these two observations to display the benefits of Model II. Once you examined the internal structure (organizations) of the Soviet military, these two “inconsistencies” are readily explainable. The missile sites were not camouflaged because the unit that was building them, part of the Strategic Rocket Forces, had never built them outside of the Soviet Union, and domestically, there had never been a need to hide the building of missile sites . The SAM units were under the Air Command, a separate unit in the Soviet Army. The left hand and the right hand were operating in secrecy, on different schedules. Here, Model II was able to square away events that the rigidity of Model I was unable to handle.
The section on the US decision to blockade examined the organizations in the US government, and how they were limited both by their SOPs, and their environment. A good example of this was the October 14 U-2 flight that actually discovered the missiles. Intelligence reached Washington on October 4 that indicated the possibility of missiles being in Cuba. So, a decision was made to fly a U-2 over western Cuba. The ten days it took to actually get the flight in the air is noted by Allison as a partial “failure” because organizational factors were responsible for the delay . It took five days for the CIA and the Air Force to settle which one of them would fly the mission. The CIA’s U-2 was better outfitted to the reconnaissance mission, but the Air Force, mindful of the two that had been shot down, wanted trained officers manning the flight. In the end, it was settled that the Air Force would put their pilots on a CIA, but this resulted in five days of requisite training for the pilots . Those ten days could have made a major difference in the outcomes of the crisis. Allison also reaches two conclusions that are important for organizational theory, and are illustrated by this example. One, when dealing with intricate organizations, tasks stubbornly do not often fit into specific jurisdictions, and two, that bureaucratic units are imperialistic . This would be supported by such theorists as Anthony Downs, who posits that bureaucracies need to be imperialistic with tasks in order to justify their continued existence.
The fifth chapter consists of Allison’s third model for looking at the crisis. Just as Model II required more knowledge of the organizational structures than Model I to render effective analysis, Model III requires more knowledge of the individuals involved. In this model, government actions are not the result of organizational outputs, but rather are the result of a bargaining circle of powerful players, who all possess rational differences that stem from different points of view . At the time, Allison noted some studies of this model, but none in the way he is using it. One of the theorists who has studied this phenomena of “decentralized coordination” is Charles Lindblom. Lindblom, Allison notes, had a different aim in his study, as he was examining how to make public policy, while Allison’s model aims to examine government outcomes . Lindlom, in The Science of ‘Muddling Through’, said that the test of a good policy was that the various analysts agree it is a good policy, even though they might disagree that the particular policy is optimal. For Allison, he notes that the interests of various government heads of department are parochial, but that they are able to reach agreement in terms of national interest . Heads of two different agencies may have their own goals, but they are able to coordinate. The key to the coordination lies in the main executive, the President.
Allison notes the dearth of study in this area, but for one expert, Richard Neustadt, in his book Presidential Power. Allison gives Neustadt a large amount of credit for influencing his Model III, because it was Neustadt’s conclusion that presidential power lies in the power to persuade that is central to Model III . The head of an organization’s main role, in terms of running an effective organization, is his ability to corral the divergent interests of his department heads (“chiefs”), harness their strengths, and coordinate their resources to the unitary goal, in this case “national policy.” Allison also looks at this from the opposite angle. In relaying a story about the Truman-MacArthur feud, he notes that Truman’s chiefs were in agreement on how he should handle MacArthur, but none of them told Truman, each expecting the other to do tell him . So, an executive can “fail” when his subordinates do not support him fully, implying a necessary mutuality, similar to the one espoused by Taylorism.