Prying Open the “Black Box” of Decision-Making

Things became a bit more conceptually murky with the defeat of the Axis powers and the coming of the Cold War. Could Soviet national interest fully explain Josef Stalin’s policy of ideological confrontation with the Western powers? At best this explanation seemed incomplete. For some trying to assess the situation through the realist paradigm, the first stages of the Cold War between the Soviet Union and the West were a puzzle that made no sense. For a power that had experienced such material and human losses in the fight against Nazi Germany, traditional conceptions of national interest should have dictated that Moscow pursue a period of external peace so that internal reconstruction could take priority.¹⁰¹⁰ W. L. White, Report on the Russians (New York: Harcourt, Brace, 1945), 36. Others, such as American diplomat George Kennan, maintained that the Stalin regime needed external conflict for internal survival, and highlighted how ideological factors could take precedence over economic needs in determining policy.¹¹¹¹ 861.00/2-2246: Telegram, “The Charge in the Soviet Union (Kennan) to the Secretary of State,” February 22, 1946, archived at: http://nsarchive.gwu.edu/coldwar/documents/episode-1/kennan.htm. In sum, assuming that the Soviet Union was making decisions based only on national interest was insufficient, since that did not explain how Soviet national interest might be conceptualized, prioritized, defined, implemented, and executed – and which individuals, institutions, and organizations could ensure that their version of the national interest prevailed in the councils of government. If this was true of the Soviet Union, then the lesson could easily be extrapolated for other countries. Moreover, an appeal to national interest could not and did not explain why governments made the choices that they did – or why two states that might appear to be equal in international position (in terms of economic and military power, for instance) might choose very different policy options. Thus, there was growing interest, as FPA scholar Valerie Hudson observes, in focusing on how “human decision-makers acting singly and in groups” might indeed be “the ground of all that happens in international relations.”¹²¹² Valerie M. Hudson, “Foreign Policy Analysis: Actor-Specific Theory and the Ground of International Relations,” Foreign Policy Analysis 1 (2005): 2.

At the same time, as the advent of the Cold War necessitated the creation of a large and permanent national security apparatus in Washington, the US government began to recruit growing numbers of social scientists into its ranks, especially in the Central Intelligence Agency (CIA) and the staffs of the White House and its nascent National Security Council (NSC).¹³¹³ David L. Featherman and Maris A. Vinovskis, “Growth and Use of Social and Behavioral Science in the Federal Government since World War II,” in Social Science and Policy-Making: A Search for Relevance in the Twentieth Century, eds. David L. Featherman and Maris A. Vinovskis (Ann Arbor, MI: University of Michigan Press, 2001), 68. Having observed how governments made national security decisions from the inside, some members of this growing cohort of scholar-practitioners would eventually return to academia, where they were now prepared to use these real-world insights to challenge old orthodoxies and to formulate new questions and answers. In 1951, General Dwight D. Eisenhower, then serving as president of Columbia University, created the Institute of War and Peace Studies as a research center that would focus on the causes of conflict between states. This institute at Columbia University, as well as new centers created at other universities during this period to grapple with new issues in national and international security, would serve as an incubator for the emerging new field of foreign policy analysis.

A major conceptual breakthrough came in 1954, when a Princeton professor and former staffer at the Council on Foreign Relations, Richard Snyder, published a monograph, Decision-Making as an Approach to the Study of International Politics. (This work was later expanded and revised into a book by Snyder and two of his colleagues, H. W. Bruck and Burton Sapin, and published in 1962 as Foreign Policy Decision-Making.) Snyder and his associates were interested in probing the “whys” of governmental behavior – to open up and probe the “black box” of the decision-making process so that “one could … recognize the actual complexity underlying decisions (which includes individual biases and bureaucratic processes)”¹⁴¹⁴ Chris Alden and Amnon Aran, Foreign Policy Analysis: New Approaches (New York: Routledge, 2012), 22. (see Box 2.1 The “Black Box” of Decision-Making).

Snyder began to probe “the stories behind foreign policy decisions” and to encourage the creation of new sets “of lenses that bring some focus to the complex picture that emerges.”¹⁵¹⁵ Jean Garrison (ed.), “Foreign Policy Analysis in 20/20: A Symposium,” International Studies Review 5(2) (2003): 155. Snyder also began to ask questions that have driven further research and study, among them the interrelationship of foreign and domestic policy, the influence of cultural factors on policy decisions, and the balance between agency (the ability of individuals to enact their will) and structure (pre-existing arrangements that influence and limit choices).¹⁶¹⁶ Valerie M. Hudson, “Foreign Policy Decision-Making: A Touchstone for International Relations Theory in the 21st Century,” in Foreign Policy Decision-Making rev. edn, eds. Richard C. Snyder et al. (New York: Palgrave Macmillan, 2002), 1–2. This research challenged the pat assumption that a state always developed its national security policy based on a clear understanding of its national interest – and indeed questioned whether the national interest really even exists. Snyder, his colleagues, and other researchers who took up his approach together laid the basis for FPA as a specific field of intellectual inquiry, tasked to probe the rationale behind the decisions taken by governments in the international arena.¹⁷¹⁷ See, for instance, the discussion in Juliet Kaarbo, “A Foreign Policy Analysis Perspective on the Domestic Politics Turn in IR Theory,” International Studies Review 17(2) (2015): esp. 190–192.

Conceiving Three “Levels of Analysis”

The seeds of another major conceptual innovation were planted in 1952, when an aspiring scholar who had served in two wars, Kenneth Waltz, joined Columbia University’s new Institute for War and Peace Studies as a research assistant while completing his doctorate. Upon being awarded his PhD in 1954, he became one of the Institute’s research associates. While at Columbia, he began to expand and revise his dissertation examining the causes of war, which was ultimately published in 1959 as the hugely influential book Man, the State and War. Waltz introduced a tripartite taxonomy to classify explanations for why countries pursue armed conflict, grouped around what he termed images of the individual, the state, and the state-system (otherwise known as the international system), with “each image defined according to where one locates the nexus of important causes.”¹⁸¹⁸ Kenneth N. Waltz, Man, the State and War: A Theoretical Analysis (New York: Columbia University Press, [1954, 1959] 2001), 12. Waltz sought to examine whether the decision to go to war emanated from the emotions and perceptions of individual leaders (human nature), or whether forces in the domestic political system or the international environment were the deciding factors (see Figure 2.2 Levels of Analysis).¹⁹¹⁹ Robert Jervis, “America and the World – 2017 and Beyond: Introductory Essay,” H-Diplo, January 2, 2017, at: http://issforum.org/roundtables/policy/1-5B-Jervis. Waltz concluded that any national security policy decision “is then related to one of our three images of international relations, or to some combination of them.”²⁰²⁰ Waltz, Man, the State and War, 13.

The year after the publication of Waltz’s book, J. David Singer rechristened the “images” as “levels” in an effort to suggest how the three images would interrelate, from the personal to the domestic system to the international environment. In particular, Singer hoped that a terminological shift to levels of analysis might shift the emphasis away from trying to find which level was the right answer for explaining policy and toward using all three levels to look for correlation and causation.²¹²¹ See the overall discussion in Singer, “The Level-of-Analysis Problem in International Relations,” 77–92. Moreover, Singer sought to extend the explanatory reach of this simple framework beyond just the causes of war. Rather, he conceived of levels of analysis as a useful approach to understanding any complex foreign policy decision. This in turn opened the door for a wave of new theorizing about whether and how each of these levels was conceptually important.

Waltz himself leaned toward systemic/international explanations as the driving factors for policy choices, with domestic institutional and personal leadership qualities as secondary ones. A state’s national security policy could best be understood, according to Waltz, by focusing on the unequal distribution of capabilities between states and the overall anarchic international environment.²²²² See, for instance, the discussion in Mark Kramer, “Ideology and the Cold War,” Review of International Studies 25(4) (1999): 539. For instance, following Waltz’s preferred view, a focus on the international level could explain why Poland pursued a consistent foreign policy from 1918 to 1939 despite oscillating between liberal and conservative governments, and parliamentary democracy and authoritarian dictatorship. Surrounded by two powerful enemies in Germany and the Soviet Union, any government in Warsaw would need to secure allies and work to contain both Moscow and Berlin. After all, states exist and operate in different geopolitical structures, and governments must respond to the realities of their immediate international environment. Differences in policies pursued by different states can often be linked to differences in the systemic environment in which they are located.²³²³ Michael Foulon, “Neoclassical Realism: Challengers and Bridging Identities,” International Studies Review 17(4) (2015): 636.

Others, however, were prepared to admit greater influence at the other levels of analysis. For instance, another research associate at the Institute of War and Peace Studies, the sociologist Amitai Etzioni, called attention to the influence of organizational culture on decision-making as well as to the importance of the psychological-cognitive dimension of decision-making.²⁴²⁴ See the discussion in Nikolas K. Gvosdev, Communitarian Foreign Policy: Amitai Etzioni’s Vision (New Brunswick, NJ: Transaction, 2016), esp. 9–12. There were concerns that an overemphasis on the systemic/international level of analysis ran the risk of treating policymakers as “interchangeable generic rational utility maximizers,”²⁵²⁵ Hudson, “Foreign Policy Analysis: Actor-Specific Theory,” 3. and thus ignored the reality that “the particularities of the human beings making national foreign policy were vitally important to understanding foreign policy choice.”²⁶²⁶ Ibid., 7.

This focus on the human element, and particularly on the preferences and political beliefs of leaders, led a number of scholars to put increasingly greater emphasis on the cognitive dimension of policymaking.²⁷²⁷ Robert Jervis, “Do Leaders Matter and How Do We Know?” Security Studies 22(2) (2013): 153–179. These included Alexander George, whose 1964 study of President Woodrow Wilson and his chief advisor Edward House focused on how their personalities – and the nature of their personal interaction – shaped US foreign policy during and after the First World War.²⁸²⁸ Alexander L. George and Juliette L. George, Woodrow Wilson and Colonel House: A Personality Study (New York: Dover, 1964). George would spend much of his academic and government-advisory career studying crisis management and the role of personality in shaping decisions. Other insights were provided by Robert Jervis, who pointed out in a landmark 1968 article that, in contrast to the assertion of leaders being guided by a clear understanding of eternal national interests, policymakers often made decisions based on what they perceived rather than on the actual facts of the operational environment – and could very easily disregard inconvenient pieces of information that contradicted their preferred worldview.²⁹²⁹ Robert Jervis, “Hypotheses on Misperception,” World Politics 20(3) (1968): 454–479. Irving Janis and Leon Mann, in their 1977 book on Decision Making, presented what they termed a motivational model highlighting the role of emotions as a factor in how and why decisions are made.

It is true that a powerful chief executive can shape a nation’s national security policy. At the same time, however, countering views emerged to balance an overemphasis on the individual level of analysis. President Barack Obama admitted as much by noting the limits on his freedom of action to impose and implement his personal foreign policy preferences.³⁰³⁰ Avery Miller and Ali Rogin, “Obama Tells the World Not to Prejudge Trump’s Policies Before He Takes Office,” ABC News, November 20, 2016, at: http://abcnews.go.com/Politics/obama-tells-world-prejudge-trumps-policies-takes-office/story?id=43677921. These constraints have included both the structure and dynamics of the international system and those of the domestic state and society.³¹³¹ Jervis, “America and the World – 2017 and Beyond.” Thus, other scholars and observers have been interested in examining the interplay of organizations and bureaucracies within government, and the extent to which bureaucratic “structure” trumps a policymaker’s “agency.”³²³² See, for instance, the discussion in Walter Carlsnaes, “The Agency–Structure Problem in Foreign Policy Analysis,” International Studies Quarterly 36(3) (1992): 245–270.

Organizational Models

In 1969, a new Harvard professor (who had previously served as a consultant in the Department of Defense) published an article attempting to probe US and Soviet decision-making during what was arguably the most dire crisis in the history of humanity, the Cuban Missile Crisis of 1962 (see Box 2.2 The Cuban Missile Crisis).³³³³ Graham T. Allison, “Conceptual Models and the Cuban Missile Crisis,” American Political Science Review 63(3) (1969): 689–718. For two weeks, the world balanced on the precipice of nuclear war between global superpowers. In choosing this seminal event, Graham Allison wanted to test theories and assumptions about what drove policy choices in Washington and Moscow – and to take advantage of the insights he had gained from his participation in the “May Group” at Harvard (see Box 2.3 The May Group). Allison expanded the work and developed it into a full-length book (Essence of Decision) in 1971. In 1999, with Philip Zelikow, Allison significantly revised and updated the analysis in response to the declassification of formerly secret documents and recordings, and to advances in scholarship on foreign policy analysis. Like Waltz’s Man, the State and War a decade earlier, Essence of Decision offered the reader the opportunity to simultaneously deploy different analytical lenses to probe the choices made by both sides.³⁴³⁴ Allison and Zelikow, Essence of Decision.

Having established a narrative of events, Allison wanted to first test the usefulness of the prevailing explanation for foreign policy decisions, which he termed the rational actor model. As discussed above, this paradigm posited that a government would act as a single unified entity and select the optimal choice for achieving objectives based on its national interests. The rational actor model assumed that, in deciding how best to achieve the national interest, leaders (and the organizations they had created within government) had a process in place that considered all possible courses of action, systematically evaluated their pros and cons, and selected the best option.³⁵³⁵ Hudson, “Foreign Policy Decision-Making: A Touchstone,” 2.

Allison found that this model, which he designated Model I, reasonably explained many aspects of policymaking during the crisis. Meetings of the Soviet Presidium and of President John F. Kennedy’s Executive Committee of the National Security Council (EXCOMM) outwardly embodied the assumptions of this perspective. Representatives of relevant government agencies were assembled in one body, various options were proposed, the costs and benefits of different choices were evaluated, and courses of action were selected on the basis of advancing national interests. Ostensibly, policy choices were made in reaction to both countries’ assessments of their position in the international system – the Soviet decision to deploy missiles in Cuba to counter the perceived advantage the United States enjoyed from having bases close to Soviet territory; the US assessment that Soviet missiles in Cuba would give Moscow a clear strategic advantage in being able to strike American territory with limited warning; and the final resolution of the crisis driven by a clear understanding on the part of both American and Soviet leaders of the risks of nuclear war.

Still, the rational actor model could not explain all aspects of decision-making during the crisis, and indeed displayed noticeable gaps. Why had the Soviets telegraphed their actions in Cuba when greater secrecy would have better served Soviet national interests? Why were leaders on both sides presented with such a limited number of options, and why were some courses of action almost immediately settled upon as the preferred choice, while others were almost immediately dismissed? What was the process by which the agenda was set for discussion in the Presidium and in EXCOMM; were these bodies always the focal points of governmental action; and what might influence what was put on or taken off the list of acceptable options? Indeed, the very notion of “agenda-setting” was not even anticipated by the rational actor model, which assumed that all possible courses of action would be equally considered and addressed, and all relevant pieces of information would make their way to the proper authorities.³⁶³⁶ Jerel A. Rosati and James M. Scott, The Politics of United States Foreign Policy, 6th edn (Boston, MA: Wadsworth-Cengage Learning, 2014), 272. Yet, as Allison observed, in a situation in which all participants cannot agree on what constitutes an optimal course of action, policy might instead be “the result of the triumph of one group [within government] over others.”³⁷³⁷ Allison and Zelikow, Essence of Decision, 707.

Therefore, Allison advanced two other models for explaining decision-making during the crisis: organizational behavior (Model II) and governmental politics (Model III). If Model I assumed that changes in the systemic environment that affected a state’s national interests were the principal variables in policymaking, Model II looked at the organizational level, while Model III dug even deeper to include both organizational factors and individual personalities within the leadership cadre.

Allison’s organizational behavior model represented a significant challenge to the verities of the rational actor model. The starting point for this approach was the assessment that government is not a unitary actor (which the rational actor model presupposes), but a collection of organizations with distinct missions and capabilities. The rational actor model assumed that all agencies and departments of government can be perfectly coordinated to arrive at and execute a single policy choice. Model II, in contrast, took as its starting point the observation that any complex national security problem would be broken up, with pieces assigned to different organizations. These organizations would then rely on pre-existing plans and standard operating procedures – immediate, off-the-shelf responses that, while not optimal in any particular case, perform well on average, reduce uncertainty, and allow the organization to rapidly respond to many problems.

Thus, as Allison noted, American spy planes were able to discover Soviet missiles in Cuba not because the Soviet Union had made a deliberate strategic choice to make the missiles conspicuous, but because the construction battalions sent from the heartland of the USSR to construct the missile emplacements followed a standard template that stood out like a sore thumb on the island of Cuba. In turn, once Soviet missiles were discovered in Cuba, the options available to the president were based largely on the core competencies and existing plans of the organizations that made up the national security establishment (diplomatic action from the State Department, airstrikes from the US Air Force, a blockade from the US Navy).

Allison’s governmental politics model, by contrast, viewed policy as emerging from the bargaining that would take place among senior officials, notably the heads of the relevant bureaucracies. Allison pointed out the extent to which a national security decision might in fact be a negotiated compromise among different senior leaders jockeying for position and influence around the decision-maker, especially to advance their own personal and organizational interests. While notions of policymaking as bargaining had already been explored in the previous decade by two Harvard scholars, Richard Neustadt and Samuel Huntington, Allison explicitly “solidified the bargaining nature of governmental policy into a decision-making model.”³⁸³⁸ Jerel A. Rosati, “Developing a Systematic Decision-Making Framework: Bureaucratic Politics in Perspective,” World Politics 33(2) (1981): 234.

Taken together, Allison’s Models II and III suggested that policy decisions were not necessarily optimal outcomes arrived at rationally after vigorous assessment of various alternative solutions, and, in some cases, could not even properly be called decisions. Instead, foreign policy resulted “from compromise, conflict and confusion of officials with diverse interests and unequal influence.”³⁹³⁹ Allison and Zelikow, Essence of Decision, 162. Drawing upon the observations of political scientist and economist Herbert Simon, Allison concluded that, just like individuals, national governments often gravitated toward a so-called “satisficing” solution⁴⁰⁴⁰ Herbert Simon, “Rational Choice and the Structure of the Environment,” Psychological Review 63 (1956): 129.—adopting the first workable option that could reasonably be expected to bring about success and that could garner sufficient support from the people and interests represented around the decision-making table. Instead of looking at an amorphous “national interest” as the basis for decisions, therefore, Models II and III showed the extent to which US and Soviet policy was affected by the pre-existing plans and missions of governmental organizations (services, agencies, and departments), particularly in how they fell back on set “standard operating procedures” in responding to the crisis. They also highlighted the importance of the bargaining and politicking both among the members of EXCOMM and the Soviet Presidium to control the presentation of information and options, to gain the ears of President John F. Kennedy and Premier Nikita Khrushchev, and to build coalitions among senior officials to support a particular course of action.

Allison’s work proved to be profoundly influential, creating a surge of scholarly interest in examining how “bureaucratic organizations (decision unit structures) and their processes (bureaucratic politics)” might impact and influence the outcomes of national security decisions.⁴¹⁴¹ Charles F. Hermann, “What Decision Units Shape Foreign Policy: Individual, Group, Bureaucracy?” in Foreign Policy Analysis, ed. Richard L. Merritt (Lexington, MA: Lexington Books, 1975), 119. Allison’s insights were further developed by Morton Halperin (who, along with Priscilla Clapp and Arnold Kantor, released his findings in the 1974 book Bureaucratic Politics and Foreign Policy). Halperin was an academic who, along with Allison, Neustadt, and others, took part in the discussions of the “May Group” at Harvard, which focused on how the bureaucracy might set the agenda for national security policy. Like Allison, Halperin also entered the Department of Defense during the 1960s, which, as he noted, allowed him to “learn much about how the bureaucracy functions and about the obligations and responsibilities” of the mid-level political appointees.⁴²⁴² Morton H. Halperin and Priscilla A. Clapp, with Arnold Kanter, Bureaucratic Politics and Foreign Policy, 2nd edn (Washington, DC: Brookings Institution Press, 2006), ix. Halperin extended Allison’s work by examining how policy was shaped by the interplay of leaders, bureaucratic actors, organizational culture, and influences from outside the government. In particular, Halperin focused on how national security organizations within the government could easily substitute their more tangible organizational interests in place of a vague definition of the national interest in order to justify their policy positions and stands. Policy would emerge as a result of the struggle among organizations to enshrine their priorities or to find workable compromises to preserve at least some minimum amount of their equities.

To those who argued, using Waltz’s level of the individual, that presidential appointees could impose the chief executive’s vision on recalcitrant organizations, Halperin warned that outsiders could be “captured” by the organizations to which they were posted and would often come to calculate the national interest “in terms of the organizational interests” of their home agencies.⁴³⁴³ Ibid., 61. Halperin also pushed back against Waltz’s preference to assign greater weight to the systemic level of analysis by pointing out that a careful assessment of bureaucratic positions showed that “it was rare that changes in the international environment alone would lead to changes in participants’ interests or in their judgments about the desirability of particular decisions.”⁴⁴⁴⁴ Ibid., 100. Moreover, when there was a distinct lack of clear and decisive leadership from the president or the White House, departments and agencies would continue to pursue their own vision of the correct course of action, leading to “policies that are devised and pursued in separation.”⁴⁵⁴⁵ Conor Keane, “The Impact of Bureaucratic Conflict on US Counternarcotics Efforts in Afghanistan,” Foreign Policy Analysis 12(3) (2016): 309.

The initial work of Allison on governmental politics, along with the further insights from Halperin and others, laid the foundations for what became widely accepted as a bureaucratic politics model for understanding US foreign policy decision-making. For many, this is the lens that frequently best explains the reality of how the US government functions, particularly in the context of the formal interagency process of the National Security Council (NSC). The principal driver of policy is neither the international system nor the individual preferences of the leader, but the compromises that have emerged from the bureaucracy.⁴⁶⁴⁶ Jervis, “America and the World – 2017 and Beyond.” Thus, observers of US government processes often note:

Much of Halperin’s work focused on the intra-governmental competition among different organizations for defining roles and missions as a key shaper of US policy. This approach drew criticism from some, like Stephen Krasner, who argued that the concept of bureaucratic politics downplayed the role and influence that presidential wishes (and those of the senior appointees in government) played in setting the bureaucracy in motion.⁴⁸⁴⁸ See, for instance, Stephen Krasner, “Are Bureaucracies Important? (Or Allison Wonderland),” Foreign Policy 7 (1972): 159–179.

Attempting to balance between those who stressed the importance of the individual level of analysis in driving policy decisions versus those situating the motivating force within the bureaucracy (the second level), Jerel Rosati put forward the notion of a spectrum of dominance – that the president (as an individual) would dominate in the policy process when the issue was of particular importance to him, and that the bureaucracy would tend to set the agenda when an issue was more routine or did not rise to the level of presidential attention.⁴⁹⁴⁹ Rosati, “Developing a Systematic Decision-Making Framework,” 234–252.

Even for those issues that received a great deal of presidential focus, however, the president’s personal preferences might not always set policy. Halperin, Clapp, and Kantor also noted that the US president, as the senior national security decision-maker, was exposed to a variety of influences. They pointed out that “he often responds at any one time to whichever pressures are momentarily strongest, whether they come from particular elements in the bureaucracy, from foreign governments or from his own domestic political concerns.”⁵⁰⁵⁰ Halperin, Clapp, and Kantor, Bureaucratic Politics and Foreign Policy, 83. The US president is also impacted by the particular constitutional set-up which, although it gives the preponderance of authority in national security matters to the Executive Branch, still vests the other branches, especially Congress, with authority as well.

Two-Level Games

Halperin’s observation that actors from outside the government’s decision-making process might also exert influence gave rise to interest in how a state’s foreign and defense policies might be affected by outside people and institutions within the wider domestic political process.

Researchers took two particular tracks. The first was to become more concerned “not only with the actors involved in the state’s formal decision-making apparatus, but also with the variety of sub-national sources of influence upon state … policy.”⁵¹⁵¹ Alden and Aran, Foreign Policy Analysis: New Approaches, 1. James N. Rosenau, a political scientist and international relations expert, started his academic career investigating the impact of public opinion (both elite and mass) on the conduct of foreign and defense policies. His work challenged earlier assumptions made by Gabriel Almond and Walter Lippmann – the dominant view for much of the mid-twentieth century – that national security policy was of interest only to a small group of specialists within the government; that the public generally could not form coherent positions on foreign policy issues; and that public opinion would have a limited impact on policy formation.⁵²⁵² The Almond–Lippmann consensus is summed up in Ole R. Holsti, “Public Opinion and Foreign Policy: Challenges to the Almond–Lippmann Consensus,” International Studies Quarterly 36(4) (1992): 442. Rosenau’s research suggested that interest in foreign affairs extended beyond government bureaucracies, and his analysis broke down “public opinion” into three broad groups – first, the opinion of “elites” (those in government and the media); second, the opinion of the “informed citizenry” (public intellectuals, the private sector business community); and, finally, the “general public” – to see what impact outside lobbying and advocacy might have in how policy decisions were made.⁵³⁵³ James Rosenau’s classic work is Public Opinion and Foreign Policy (New York: Random House, 1961). Sometimes, the driving force behind a policy decision would not be located within a governmental organization, but would come about as a result of powerful and mobilized domestic constituencies outside of government, yet capable of influencing the national security process. European scholars such as Thomas Risse-Kappen and Harald Muller focused their attention on “the nature of the political institutions, with the basic features of the society, and with the institutional and organizational arrangements linking state and society and channeling societal demands into the political system.”⁵⁴⁵⁴ Quoted in Alden and Aran, Foreign Policy Analysis: New Approaches, 7. For more on the state–society linkages and their influence on national security policy, see, for instance, Harald Muller and Thomas Risse-Kappen, “From the Outside In and From the Inside Out,” in The Limits of State Autonomy, eds. David Skidmore and Valerie M. Hudson (Boulder, CO: Westview Press, 1993), 25–48.

The second area of study focused on how foreign governments and interests might influence the domestic national security process. Rosenau’s work, as well as the research done by other scholars and practitioners, led to a greater appreciation of the “expanding overlap of domestic and foreign affairs,” which meant that, in order to understand national security decisions, it would be necessary to “probe both the internal and external dynamics” which feed into the policy process.⁵⁵⁵⁵ James N. Rosenau, Along the Domestic–Foreign Frontier: Exploring Governance in a Turbulent World (Cambridge: Cambridge University Press, 1997), xiii. Rosenau was one of the first to focus on what he described as linkage politics that connected national and international affairs.⁵⁶⁵⁶ Robert D. Putnam, in his article, “Diplomacy and Domestic Politics: The Logic of Two-Level Games,” International Organization 42(3) (1988): 430, provides a summary of Rosenau’s initial work in this area. In other words, global influences could have an impact on the domestic policy process. In turn, policymakers would have to consider not only the strategic merits of any particular course of action, but also the likely impact on a whole host of domestic and international actors.⁵⁷⁵⁷ Nikolas K. Gvosdev, “Should Military Officers Study Policy Analysis?” Joint Forces Quarterly 76(1) (2015): 33. This led to the realization that the “decision-making process involves both a domestic arena where one set of rules and interests govern, and an international arena where a different set of rules and interests prevails.”⁵⁸⁵⁸ Alden and Aran, Foreign Policy Analysis: New Approaches, 8. In other words, a country’s national security policy was worked out in both domestic as well as in international arenas.

A Harvard social scientist who undertook an in-depth analysis of the Group of Seven summits during the 1970s, Robert Putnam, began to put these pieces together into an analytical framework. Putnam noted, “It is fruitless to debate whether domestic politics really determine international relations, or the reverse. The answer to that question is clearly ‘Both, sometimes.’ The more interesting questions are ‘When?’ and ‘How?’”⁵⁹⁵⁹ Putnam, “Diplomacy and Domestic Politics,” 427.

Putnam came across plenty of anecdotal evidence that suggested that policy was being formulated amid simultaneous, competing international and domestic negotiations. Robert Strauss, the US Trade Representative during the Carter Administration, observed as much, noting: “I spent as much time negotiating with domestic constituents (both industry and labor) and members of the US Congress as I did negotiating with our foreign trading partners.”⁶⁰⁶⁰ Robert S. Strauss, “Foreword,” in Joan E. Twiggs, The Tokyo Round of Multilateral Trade Negotiations: A Case Study in Building Domestic Support for Diplomacy (Washington, DC: Georgetown University Institute for the Study of Diplomacy, 1987), vii, John Dunlap, who served as secretary of labor in the Ford Administration, summed up his experience that any successful bilateral negotiation required three agreements: one across the table, and one on each side of the table.⁶¹⁶¹ Quoted in Howard Raiffa, The Art and Science of Negotiation (Cambridge, MA: Harvard University Press, 1982), 166.

Drawing on the interaction between the systemic/international and state/society levels, adding the personal capabilities of the negotiators, and applying the insights of bureaucratic politics and linkage politics, Putnam’s answer was to put forward the model of a two-level game where policymakers were simultaneously negotiating with their counterparts at the systemic/international level and with actors in their own domestic political system. As he described it:

This “two-level” approach means that any effort to analyze policy would have to move beyond charting and explaining the relationships between organizations within government (per Allison and Halperin). It would have to see additional underpinnings for national security policy in the actions of domestic non-governmental actors, including non-governmental organizations, businesses, interest groups, think tanks, and the media. It would also need to consider the role of international actors, not only other states and groups of states, but also non-state actors, including social, religious, and cultural movements.⁶³⁶³ Stephen Smith, Amelia Hadfield, and Tim Dunne, “Introduction,” in Foreign Policy: Theories, Actors, Cases, eds. Stephen Smith, Amelia Hadfield, and Tim Dunne (Oxford: Oxford University Press, 2008), 5. No two games are identical – each will have a different mix of factors in play. For instance, there would be different outcomes based on the interests and capabilities of the policymaker, whether a statesman is “faced with domestic constraints of mobilization,” or is taking action “to maximize domestic political support,” or “to realize personal goals.”⁶⁴⁶⁴ From Andrew Moravcsik’s introduction in Peter B. Evans, Harold Karan Jacobson, and Robert D. Putnam, Double-Edged Diplomacy: International Bargaining and Domestic Politics (Berkeley, CA: University of California Press, 1993), 16. Capabilities may differ between states – based both on their domestic sources of authority as well as the unequal distribution of power in the international system. All of these factors can add to the complexity of the process.

The two-level games approach, therefore, made it clear that policymakers had to address both domestic and international politics in formulating options. In addition, it provided a way for viewing foreign policy decision-making across the levels of analysis. As Michael Clarke has observed:

In the American context, all these influences come together in a policy process that is attempting, in the words of former National Security Advisor Robert McFarlane, to “work towards a national security policy that is both consistent and responsive to the needs of the country and the views of the President.”⁶⁶⁶⁶ Robert C. McFarlane, with Richard Saunders and Thomas C. Shull, “The National Security Council: Organization for Policy-Making,” in The Presidency and National Security Policy, ed. R. Gordon Hoxie (New York: Center for the Study of the Presidency, 1984), 261.

Evolution of the Classic Organizational Models

First, there is the reality that paradigms have to evolve as governments themselves change. Much of the formative work in the field of foreign policy analysis took place during the Cold War period and reflects a snapshot of the US government as it was configured during that period. This snapshot reflects, in particular, the early effects of the 1947 National Security Act and the steps taken by Franklin D. Roosevelt, Harry S. Truman, and Dwight Eisenhower to shape the institutions of the modern US presidency. The ongoing shift in the United States, where the legislative branch has steadily been ceding ground to an imperial presidency, and the formal cabinet has been losing power and influence to the White House staff, means that models drawn up to reflect how the US government undertook national security decisions in the 1950s and 1960s must be modified to keep up with how things actually function in the second and third decades of the twenty-first century.⁷⁰⁷⁰ Observing the decline in cabinet governance in the United States during the 1970s and 1980s, Samuel Kernell was already warning at the end of the Reagan Administration that it was unrealistic to expect that the model of how things had been done in the past could be easily resurrected. See his “The Evolution of the White House Staff,” in Can the Government Govern? eds. John E. Chubb and Paul Peterson (Washington, DC: The Brookings Institution, 1989), esp. 235. Another reality that must be taken into account is the massive expansion in the size and reach of the US national security establishment, not only in the sheer number of employees but in the proliferation of new agencies across the Executive Branch.⁷¹⁷¹ Derek S. Reveron, Nikolas K. Gvosdev, and Mackubin T. Owens, US Foreign Policy and Defense Strategy: The Rise of an Incidental Superpower (Washington, DC: Georgetown University Press, 2015), 5–6. It is now much harder to negotiate a compromise, say on US Middle East policy, if it is no longer primarily a negotiation between the heads of the departments of State and Defense, but rather between dozens of different organizations clamoring to have their equities represented at the table.⁷²⁷² See, for instance, Nikolas K. Gvosdev, “The Noisy Cabinet,” National Interest, January 14, 2009, at: http://nationalinterest.org/article/the-noisy-cabinet-2975.

Similarly, Putnam’s original framework of a two-level game has also undergone modification over the last several decades. For one thing, more attention has been paid to the role of individuals in how multilevel negotiations are carried out. For instance, the personal rapport between Secretary of State John Kerry and Iranian Foreign Minister Javid Zarif proved to be essential for keeping the negotiating process to reach the Iran nuclear deal alive and on track, when significant pressures from both the Iranian and American domestic political systems threatened to derail the agreement. Because these two specific individuals were in place and in charge of their respective nations’ delegations, however, a degree of trust could be facilitated that enabled negotiations to continue.⁷³⁷³ Robin Wright, “Tehran’s Promise,” New Yorker, July 27, 2015, at: www.newyorker .com/magazine/2015/07/27/tehrans-promise. Modifying the two-level game to incorporate the presence or absence of rapport between the actual interlocutors carrying out the negotiations represents the addition of Waltz’s individual level of analysis to the more impersonal game depicted by Putnam.⁷⁴⁷⁴ See, for instance, how Kevin Marsh adds the “worldview of the president and his advisors” to his list of two-level game factors to be considered, in “‘Leading from Behind’: Neoclassical Realism and Operation Odyssey Dawn,” Defense and Security Analysis 30(2) (2014): 127.

Putnam’s framework has also been revised as different factors (such as greater regional and international integration and developments spurred onward by the current wave of globalization) have both strengthened and weakened the sovereign nation-state as the principal actor. The rise of more powerful regional and transnational associations such as the European Union, to which states have ceded some of their sovereign authority, has created conditions for a three-level game where a state must, in its interactions with another government, take into account both its domestic politics as well as the considerations of the international associations to which it belongs.⁷⁵⁷⁵ See, for instance, Playing Three Level Games in the Global Economy: Case Studies from the EU, ed. Davide Bonvinici, EU Diplomacy Papers 4/2008 (Bruges: College of Europe, 2008). Three-level games are increasingly the norm in Europe, where domestic constituencies can mobilize the supranational institutions of the European Union to pressure or constrain the positions of national governments.⁷⁶⁷⁶ Beach, Analyzing Foreign Policy, 85. Technological changes and the impact of globalization have also changed the nature of the game.⁷⁷⁷⁷ For an overview of these developments, see, for instance, Walter B. Wriston, “Bits, Bytes, and Diplomacy,” Foreign Affairs 76(5) (1997): 172–182. The negotiating teams designated to represent each state are no longer the automatic gatekeepers of the process. Instead, domestic actors in both countries who support or oppose particular outcomes can communicate directly. In some cases, they may be able to coordinate their actions with relevant actors on the other side, bypassing the two negotiating teams. This trend was very much on display in the nuclear talks between the United States and Iran in 2014 and 2015.

Taking into account such developments, Karen Mingst has expanded the number of game players to encompass a total of seven groups of what she terms “linkage actors” – paying more attention to not only international organizations but also the role of international law and regulatory agencies in casting votes on policy, as well as the expert community and the non-governmental sector. In addition, she has called attention to how coalitions can form within a government in support of a specific course of action and reach out across the international divide to find allies in the government of the opposite side.⁷⁸⁷⁸ Karen A. Mingst, “Uncovering Missing Links: Linkage Actors and Their Strategies in Foreign Policy Analysis,” in Foreign Policy Analysis: Continuity and Change in its Second Generation, eds. Laura Neack, Jeanne A. K. Hey, and Patrick J. Haney (Englewood Cliffs, NJ: Prentice Hall, 1995), 229–242. It is now common to refer to foreign policy as a “nested game with many players.”⁷⁹⁷⁹ Laura Neack, The New Foreign Policy: Complex Interactions, Competing Interests, 3rd edn (Lanham, MD: Rowman & Littlefield, 2014), 206.

New Organizational Paradigms for a Changing World

It is thus possible to chart how the original paradigms have evolved over time, and to see how they may continue to evolve looking into the future. For example, the three Waltzian levels are discernible in the four broad “determinants” that Alex Mintz and Karl DeRouen argue create the basis for policy decisions: the individual level in their emphasis on “psychological factors”; the state/societal level corresponding to “domestic factors”; and the systemic represented in “international factors.” The synthesis of these three levels then occurs in the actual “decision environment” where these factors are assessed, weighed, and resolved.⁸⁰⁸⁰ Alex Mintz and Karl DeRouen, Jr., Understanding Foreign Policy Decision-Making (Cambridge: Cambridge University Press, 2010), 4. Any discussion of the decision environment, in turn, brings us to Allison’s models, but the original three models as described by Allison (rational actor, organizational process, and governmental politics) have also evolved as circumstances have changed and other factors have become more important.

Allison, for instance, had argued that actors in his Model III would be motivated to advocate for policy based primarily by their position within government and their particular bureaucratic role.⁸¹⁸¹ Kevin P. Marsh, “Obama’s Surge: A Bureaucratic Politics Analysis of the Decision to Order a Troop Surge in the Afghanistan War,” Foreign Policy Analysis 10 (2014): 267. Yet he also acknowledged that a bureaucratic player within the interagency process would advocate for outcomes “that will advance his conception of national, organizational, group and personal interests.”⁸²⁸² Allison and Zelikow, Essence of Decision, 171. As the scope and complexity of national security issues has grown in the years since Allison developed his models, so too has the size and number of the bureaucratic organizations and associated intra- and interagency processes to deal with them. Because such a sprawling national security apparatus is inherently more difficult to manage from the top, this development raises the possibility that lower-level bureaucratic dynamics may be playing a more important role than in Allison’s day. Given the notable lack of scholarly attention to such subordinate bureaucratic dynamics that persists even today, this represents a gap in FPA knowledge. This strongly suggests there may be a need to consider the usefulness of parsing of the classic bureaucratic politics model by breaking out a discrete sub-bureaucratic politics perspective that sharpens and amplifies the analytic focus in this increasingly important regard.

Likewise, in the decades since Allison started his initial research, the rise of a much more powerful White House staff (especially those connected to the National Security Council) has produced a much larger group of presidential advisors who do not “belong” to any particular bureaucratic unit but instead seek to advocate policies on behalf of their personal interpretation and understanding of the president’s wishes. Moreover, holding an important bureaucratic position is no longer an automatic guarantee of being included within the president’s inner decision circle as was assumed in the 1960s when the May Group was meeting. The Allison and Halperin model of bureaucratic politics is, based on an application of Rufus Miles’ observation that “where you stand” (on any given issue) “depends on where you sit” (within the US government).⁸³⁸³ Rufus E. Miles, Jr., “The Origin and Meaning of Miles’ Law,” Public Administration Review 38(5) (1978): 399–403. Having people influence decision-making not on the basis of representing key parts of the national security establishment, but instead by virtue of their position and access as part of the president’s inner circle, requires adding yet another conceptual model in the form of a palace politics perspective. In reality this is not an entirely new paradigm, but instead represents a parsing of Allison’s broad original conception of governmental politics, breaking out this discrete aspect to give a sharper analytical focus to these developments.

Looking Beyond Organizational Models

There are other areas where the individual choices and preferences of leaders can have tremendous impact on what occurs in the realm of national security policy. A president’s choice of advisors, how he or she chooses to delegate, and the national security structures he or she creates can have a critical impact on the policies that are chosen and executed.⁸⁴⁸⁴ Margaret G. Hermann and Thomas Preston, “Presidents, Advisers, and Foreign Policy: The Effect of Leadership Style on Executive Arrangements,” Political Psychology 15(1) (1994), 75–96. How an individual leader interprets both the systemic level in which his or her country must operate and determines how to navigate the state/societal level in pursuing policy choices also matters for how policy is framed.⁸⁵⁸⁵ See, for instance, the discussions about the foreign policies pursued by Kazakhstan’s president Nursultan Nazarbayev and Azerbaijan’s president Ilham Aliyev, in Nikolas K. Gvosdev, “Keystone States: A New Category of Power,” Horizons 5 (2015), at: www.cirsd.org/en/horizons/horizons-autumn-2015-issue-no5/keystone-states-a-new-category-of-power. Over the last thirty years, there has been a steady cross-pollination of foreign policy analysis with theories originally developed to explain the economic choices of individuals in the market place.⁸⁶⁸⁶ Steven B. Redd and Alex Mintz, “Policy Perspectives on National Security and Foreign Policy Decision-Making,” Policy Studies Journal 41(S1) (2013): S11–S37. If rational choice theory or prospect theory, for instance, could provide insights into why people would buy certain types of insurance, a number of scholars were then interested in their applicability to why leaders might embrace or reject arms control, join or leave alliances, or support or eschew military intervention in Rwanda, the Balkans, or the Middle East.

Building on those conceptual developments, scholars like David Houghton took the three Allisonian models and extrapolated the proposition that there are four different ways to view decision-makers: the decision-maker as the detached cost–benefit maximizer (homo economicus); the decision-maker shaped by his or her bureaucratic and organizational environment (homo bureaucraticus); the decision-maker impelled by his or her own vision, worldview, and sets of experiences (homo psychologicus); and the decision-maker influenced by the dynamics of the group assembled to consider the options and the rules by which that group functions (homo sociologicus).⁸⁷⁸⁷ David P. Houghton, The Decision Point: Six Cases in US Foreign Policy Decision Making (New York: Oxford University Press, 2013), 7–15. Houghton’s work to build on Allison’s models – captured in his homo economicus and homo bureaucraticus constructs – was an effort to capture elements that the original models did not quite encompass and to draw in more recent work about the applicability of economic theory to understanding national security decisions.

Houghton’s next two (homo psychologicus and sociologicus) were an effort to incorporate the new work that has been done in recent years to combine psychological insights with foreign policy decision-making which opened up new areas for exploration. Alexander George’s conception of a leader’s “operational codes” (his or her core beliefs both about the nature of the world and the preference for different types of action) laid the foundation for further study and research about how different people process data, assign value, and ascertain risk, and how this can result in vastly different decision outcomes.⁸⁸⁸⁸ See, for instance, Beach, Analyzing Foreign Policy, esp. 105–111.

In ground-breaking work, Margaret Hermann, Thomas Preston, Baghat Korany, and Timothy Shaw assessed how the strength of a leader’s belief and desire to achieve a particular goal will affect the extent to which he or she can be persuaded to take into account systemic and domestic sources of divergence – say, the opposition of key allies or unfavorable statements made by legislators. Their work reinforced assessments that differences in the cognitive makeup of different people will lead to different policy choices.⁸⁹⁸⁹ Margaret G. Hermann, Thomas Preston, Baghat Korany, and Timothy Shaw, “Who Leads Matters: The Effects of Powerful Individuals,” International Studies Review 3(2) (2001), 93–131. In turn, the types of people a leader surrounds him- or herself with – who has access, how disputes are refereed, the tolerance for debate, and discussion of options – will have a critical impact on the decision environment. Irving Janis pioneered the study of “groupthink”: how the emergence of a real or perceived consensus about the desirability of one course of action will cause the group to ignore alternatives and dissuade individuals from presenting dissenting views. Under conditions of extreme stress (especially time pressure), a group may want to quickly coalesce around a course of action and so will deliberately shut down further discussion that may prolong the decision.⁹⁰⁹⁰ Irving Janis’ main treatment of this subject is found in Groupthink: Psychological Studies of Policy Decisions and Fiascoes (Boston, MA: Houghton Mifflin, 1982). In contrast, the group surrounding the leader may be defined not by groupthink but by “polythink”: a plurality of views and perspectives that preclude the emergence of any clear options and lead the group either to decision paralysis or to adopting a lowest-common-denominator policy choice that does not reflect an optimal option, but that all members of the group can sign on to.⁹¹⁹¹ Alex Mintz and Carly Wayne, The Polythink Syndrome: US Foreign Policy Decisions on 9/11, Afghanistan, Iraq, Iran, Syria, and ISIS (Stanford, CA: Stanford University Press, 2016). In both cases of groupthink and polythink, how the leader structures the decision environment – or how he or she has that environment structured for them by others – can be the critical factor in how and why the decision was made. These observations form the basis of Houghton’s development of the latter two decision types.

Thus, homo psychologicus and homo sociologicus validate the relevance of a palace politics perspective, which calls attention to the influence of staff and advisors operating outside the formal bureaucratic structure and yet who are seen to have growing influence on decision-making. They also point to the emergence of yet another useful new lens in the form of a cognitive perspective to focus analysis on the thought processes of individuals alongside models that focus on institutions and groups.⁹²⁹² See, for instance, Nikolas K. Gvosdev, “Who Will Really Set the Next President’s Security Agenda?” The National Interest, July 12, 2016, at: http://nationalinterest.org/feature/who-will-really-set-the-next-presidents-security-agenda-16929?page=show.

Finally, building on the insights and observations of the earlier pioneers in the field, there is now an effort to combine different approaches.⁹³⁹³ Juliet Kaarbo, Jeffrey S. Lantis, and Ryan K. Beasley, “The Analysis of Foreign Policy in Comparative Perspective,” in Foreign Policy in Comparative Perspective, eds. Ryan K. Beasley, Juliet Kaarbo, Jeffrey S. Lantis, and Michael T. Snarr, 2nd edn (Los Angeles, CA: Sage, 2013), 7–19. This so-called “third generation” approach to foreign policy analysis looks to synthesize insights and observations from the previous two lines of effort as well as take into account major changes in both the domestic and international environments, particularly since the end of the Cold War, to produce analytical tools that can both explain what happens in national security affairs but also help to predict and anticipate the influences on the process.⁹⁴⁹⁴ Christopher M. Jones, “Toward a Third Generation Model: Rethinking Governmental Politics and Foreign Policy Analysis,” paper presented at the annual meeting of the International Studies Association 48th Annual Convention, Chicago, Illinois, February 28, 2007. Thus, in building on the original set of levels, models, and games, the current cadres of scholars and practitioners are looking to develop them “over multiple levels of analysis simultaneously.”⁹⁵⁹⁵ Hudson and Vore, “Foreign Policy Analysis Yesterday, Today and Tomorrow,” 224.

Combining and recombining different levels, models, and perspectives – and adding more and more players into the mix – leads to a proliferation of ever-more complex frameworks. While this can delight academics looking to break new scholarly ground, it can become distinctly unhelpful to practitioners who are looking for a series of basic signposts to delineate the key factors they must consider in the decision environment. Repeatedly, there have been warnings that the growth in theoretical perspectives is “particularly difficult to grasp and deal with effectively in the policymaking process.”⁹⁶⁹⁶ Peter M. Haas, “Epistemic Communities and International Policy Coordination,” International Organization 46(1) (1992): 13. The end result, therefore, is that practitioners simply disregard the insights that theory might provide. Borrowing a page from philosopher Nicholas Rescher, the “web of theory” that is woven around a “given manifold of data” can never be entirely adequate to the complexity of the situation – but such simplifications are necessary in order to “economize our cognitive effort by using the most direct workable means to our ends.”⁹⁷⁹⁷ Nicholas Rescher, Cognitive Complications: Epistemology in Pragmatic Perspective (Lanham, MD: Lexington Books, 2015), 63–64.

Providing an Analytic Toolkit for Practitioners

Intellectual Roots of Foreign Policy Analysis

Different Models of Decision-Making

Foreign Policy Analysis Today and Tomorrow

Taking a Textbook Approach