Causes of Change in National Security Processes: Carter, Reagan, and Bush Decision Making on Arms Control.

The structure of national security decision making at the presidential level is one of the most crucial, yet often neglected, aspects of the study of the presidency. Since the 1980s, neorealist and neoliberal theories that focus on system-level causes of state behavior have become the dominant paradigms for studying international relations. As a result, the flood of scholarly work on state-level theories such as bureaucratic politics, organizational process, and presidential management of national security decision making that dominated research agendas in the 1960s and 1970s has been reduced to a trickle.(1) Scholars of domestic political decision making and the Executive Office of the Presidency, however, have maintained an interest in how presidents organize their staffs for decision making. This article will merge some of the earlier literature on national security decision making with some more recent literature on presidential decision making. Its focus will be the ways in which national security decision-making structures are changed by international and domestic political forces as well as presidential management of the bureaucratic and organizational politics within the executive branch. The importance of this work lies in its portrait of the structure of national security decision making as a dynamic and evolving process. It argues that presidents make modifications in their standard interagency processes for the deliberate purpose of achieving specific political goals. Adjustments in policy are the impetus for adjustments in process.

The work of Walcott and Hult (1995) provides the analytical framework for this article. Walcott and Hult identify the three sources of decision-making structure as (1) the political environment, (2) the organizational dynamics/role of the advisory system, and (3) the role of the president. However, this article adds to their ideas by narrowing the focus to changes in decision-making structures. The dependent variable is defined as temporary or permanent adaptations or modifications in an administration's standard national security interagency decision-making process. The independent variables examined include changes in the international environment, changes in the domestic political environment, and attempts at policy innovation by the president. Since the political environment, the advisory system, and choices made by the president generate an initial interagency structure and process, shifts in these variables may render that original design less useful to the president. He will then make modifications. Case studies of arms control decision making from the Carter, Reagan, and Bush administrations will be used to test this hypothesis. The case studies analyze the Carter administration's March 1977 Strategic Arms Limitation Talks II (SALT II) proposal, the Reagan administration's initial Strategic Arms Reduction Talks (START) proposal of May 1982, and the Bush administration's arms control speech of September 1991, which contained the genesis for the START II agreement.

First, this article examines the influences on decision-making structure and suggests a model of how variations in those influences affect national security decision-making structures. Following this, three briefcase studies explore the initial interagency processes of each administration; the causes of deliberate, presidentially inspired alterations in that structure; and the resulting temporary or permanent adaptations. The following section will revisit the original framework and suggest the addition of three contextual factors--time, ideology, and the international system--that carry special weight in their influence on decision making in each of the three categories of independent variables identified by Walcott and Hult (1995). A final section discusses the implications of the case study findings. In particular, these cases highlight the broad range of influences on the policy process and the inability of the standard interagency model to provide presidents with policy processes that will satisfy their political needs in a timely fashion. It also illustrates three paths to changes in decision-making structure. Carter, as an innovator, was the source of permanent adaptations in his decision-making process when he felt that his interagency process was unable to produce a bold and new approach to strategic arms limitations. Reagan responded to domestic political pressures for progress in U.S.-Soviet arms control with a temporary modification of his standard interagency process. Bush perceived an opportunity to move toward deep cuts in strategic weaponry as the Soviet Union began to crumble following the attempted coup against President Mikhail Gorbachev in August 1991. His reaction to this international development was a temporary adjustment to the standard decision-making structure. In each case, the president's adaptations to the decision-making process are caused by his sense that changes in the political environment have rendered his standard interagency process less effective. This is party a function of learning. Ultimately, this article suggests a reversal of the question of whether different decision processes produce different types of policies. Instead, different policy needs seem to necessitate different decision processes.

Influences on Decision-Making Structure

This section takes a deeper look at the ways in which scholars have analyzed national security policy making. Ultimately, to explain changes in decision-making structures, a model of the various influences on the decision process must be fashioned. The model detailed below is distinct from other foreign policy process models in that it includes variables from several levels of analysis.

Models of Presidential Decision Making

National security and foreign policy decision making have been explained by scholars since the 1960s through the use of three competing models: organizational process, bureaucratic politics, and presidential management. The organizational process model contends that the government is best described as a vast conglomeration of semi-independent departments with interests and perspectives of their own; governmental policy is the output of attempts to merge those competing interests into coherent policy or the uncoordinated aggregate of decisions made by each department (Huntington 1961; Hammond 1961; Schilling, Hammond, and Snyder 1962; Wohlstetter 1962; Art 1968; Smith 1970; Halperin and Halperin 1983-1984; Senate Committee on Armed Services 1985).

The bureaucratic politics model focuses on the perception, interests, and ambitions of individual governmental officials. The president is often portrayed as just another player in the bureaucratic game. Governmental policy, ultimately, is the result of bargaining and compromise between individuals and coalitions of individuals (Hilsman 1959; Neustadt 1990; Schilling, Hammond, and Snyder 1962; Allison 1969, 1971; Destler 1972a, 1972b; Allison and Halperin 1972; Halperin and Kanter 1973; Halperin 1974).

The presidential management model suggests that presidents employ a number of management strategies or decision-making styles to gain control over self-interested departments and ambitious officials. These "management styles" are partly a function of a president's personal leadership style. From his own predispositions and from the advice of scholars and practitioners, a president creates committee structures, decision-making procedures, and roles and responsibilities for specific officials and agencies. Ultimately, the president hopes to make sure that decisions are made the way he wants them made and that the policy outcomes reflect his preferences (Art 1973; Krasner 1972; Rourke 1972; Ball 1974; Pearlmutter 1974; Johnson 1974; George 1980).(2) Within this paradigm, scholars have suggested a range of ways to describe the various methods by which presidents structure their decisions. Johnson (1974), George (1980), Thompson (1992), and Snow and Brown (1997, 102-29) identify three management styles. In a competitive management style, as used by FDR, the president pits advisers against one another and creates overlapping areas of responsibility to force senior officials to compete for influence. Formalistic management styles, as used by Truman, Eisenhower, Johnson, Nixon, and Reagan, create a committee structure and well-delineated roles for all advisers to channel the flow of information and advice to the president. These formal patterns become a regularized method of decision making. Collegial management develops teamwork among senior officials and a decision-making body in which advisers debate as equals in a problem-solving environment.

Study of these management models has continued. Porter (1980), in a study of decision making on economic policy, suggested three alternative models: (1) centralized management in which a hierarchical and formal staff structure exists; (2) adhocracy in which the decision-making structure is more flexible and informal, with irregular decision-making patterns; and (3) multiple advocacy in which a key staff member plays the role of managing the system and making sure that decisions are made with the input of the full range of departmental participation. Multiple advocacy is similar in some ways to collegial management; however, it is more formal in the sense that all relevant departments are guaranteed a place at the table and all views are thoroughly vetted through the decision-making process by a "custodian-manager" of the administration (George 1971, 1972a, 1972b; Destler 1972a, 1972b; Hall 1975; Moens 1990). Campbell (1986), in an examination of the Carter and Reagan presidencies, deepened the focus on management to study presidential preference for centralized or decentralized management, the overall relationship between the presidents and their advisers (teamwork vs. laissez-faire, regular vs. irregular meetings, standing vs. ad hoc committees), and the relationships among the advisers. Hermann and Preston (1994) examined several studies of executive leadership styles. Their research suggests that presidential leadership can best be studied in terms of two key variables: (1) authority patterns that are either formal or informal and (2) policy-coordinating methods in which building consensus (process focus) or policy achievement (policy focus) is the priority.

Importantly, the foundation of these management models is an assumption that presidents seek to create a decision-making system that will produce the types of decisions that they will find most useful. The initial decision-making structures created by incoming administrations represent a president's administrative theory--his preferred method for making decisions given his political goals and beliefs about his relationships to political appointees and the career bureaucracies. This article accepts that assumption as its starting point.

Analytical Framework

An understanding of presidential decision making requires a specific focus on the sources of decision-making structure. Competing models provide a useful framework for comparing the administrative procedures of various administrations, but they usually view the origin and evolution of these procedures in terms of presidential character and style. Walcott and Hult (1995) provide a broader framework for understanding the influences on decision-making structure and the cause of changes in decision making over time in office. Though Walcott and Hult studied the White House Office, their model is useful for examining the evolution and modification of an administration's national security decision-making process. Three different sources of structure are identifiable: (1) the political environment, (2) presidential choice, and (3) organizational dynamics (pp. 16-19).

Political environment. First, the political environment plays a large role. This includes public opinion, domestic and international political actors, and the technology of governing in the modern age. In studying the way in which organizations interact with their environments, the study of organizational behavior has a more well-developed body of literature than political science. Borrowing from this literature allows the executive branch and the national security decision-making process to be seen as "open systems" (Thompson 1967). Their decision-making characteristics are deeply influenced by environmental stimuli. One of the overriding concerns of these open systems is adapting to uncertainty in their environment. The literature on organizational behavior often focuses on how corporations adapt to the market within which they operate. Similarly, public institutions operate within a political market. Presidents certainly understand this. Just as corporations are judged by the quality and price of their products, presidents will be judged by the quality of their policies. Poor products left on the shelf can lead to bankruptcy; poor policies without public support will lead a president to electoral failure or a stained legacy. Structural contingency theory hypothesized that management systems are dependent on the environment within which they operate and that the first task of any organization is to design itself to deal with its environment and the challenges it may pose (Woodward 1958; Burns and Stalker 1961; Lawrence and Lorsch 1967). Hult and Walcott's (1990) own "governance model" examines the settings within which organizations operate by making distinctions between the level of controversy, certainty, and uncertainty over means and ends for policy. In the context of organizational change, national security decision-making processes must adapt to changes in the domestic and international political environment.

Presidential choice. The second variable is presidential choice. It refers to the president's political objectives and strategy for achieving those objectives. Again, the organizational behavior literature provides depth to this concept. In strategic choice theory, a linkage of structure, environment, and choice, Chandler (1962, 11-14) argued that "structure follows strategy." An organization's response to its environment is not like a leaf turning toward the light. It is a measured and calculated response to environmental stimuli based on a reassessment of strategic goals and tactical actions (Child 1972). Importantly, presidential strategy is seen here as a reaction to the political environment. Lake and Powell (1999), in an application of strategic choice to international relations, suggest that the crucial variable is the interaction between actors' preferences (nation-states, political leaders, or political organizations) and the international system. Their work does not deal with decision-making structure, but it does emphasize the interrelationship between strategy and the environment. Structural changes in national security processes are dependent on the changes in a president's strategy for dealing with a given political challenge.

Organizational dynamics. The third factor is organizational dynamics. Once an administration takes office and an initial decision-making structure is put in place, the procedures, committee structures, and delineated roles and responsibilities of senior officials become an influence on the way decisions will be made in the future and the way in which the structure of the decision-making unit will evolve. At first glance, this aspect seems to lean toward tautology. Walcott and Hult (1995) focused on three dependent variables in their work: emergence of White House structures, stability of those structures, and differentiation within those structures. Suggesting that organizational dynamics affects structure is almost like saying that structure is an influence on structure. However, the cause-effect dilemma is less problematic if the nature of decision making is seen in the context of competing forces and as a function of time. Organizational dynamics, in this sense, refers to the interaction of organizational and bureaucratic actors with each other and with the president over an extended period of time. These interactions are a constant factor in the evolution of a decision-making process. Since all presidents have developed some type of management strategy, a refocus on the presidential management models helps clarify the issue.(3)

If structure is defined as the relationships between the president and his senior advisers and the relationships among the senior advisers (Fenno 1959, 4), then the presidential management models can be seen as a way in which presidents order these relationships. A look at several aspects and weaknesses of these management models sets an analytical context for the study of change in decision-making structure. First, a number of scholars accept the notion that more than one management style can exist within a single administration at any given time (Kohl 1975; Porter 1980, 247; Thompson 1992, 113; Walcott and Hult 1995, 14-16). Others are even skeptical about the use of a classification scheme for decision-making styles (Burke and Greenstein 1989, 275). Both of these points highlight a key weakness in these structural models of decision making. They are too static and do not account for evolution or change in the structure of decision making. Even with caveats from the authors of articles and books dealing with management styles about the limited utility of these models, they have become a standard shorthand for judging presidential decision making.

A more accurate way of viewing decision-making structure or management styles is to see them as inherently unstable. These models are best seen as starting points for any administration. They represent the initial design of the administration's decision method. As more and more decisions are made within this process, the process begins to evolve. The three decision-making paradigms of organizational process, bureaucratic politics, and presidential management are usually seen as mutually exclusive. However, it is more accurate to see them as competing forces that always exist within any administration. The organizational process model describes the role organizations play in decision making as they compete for responsibility, authority, and resources. The bureaucratic politics model focuses on the roles of ambitious individuals fighting for their share of power. Presidential management models illustrate the president's attempts to control these forces and gain control over administration decision making. The clash between organizations, senior officials, and the president never ends. A decision-making structure is evident; assigned roles, responsibilities, and committee procedures are defined. However, a closer look at the process will reveal that this structure is like the topography of a battlefield. It is where combatants must fight, adapt to the terrain, and seek advantage within the given geographical context. Reedy's (1970, 88-99) examination of the Johnson administration is a most revealing description of presidential decision making. It compares life in the White House to life in a palace court, with intrigue, schemes, and vicious rivalries. Much of this arises out of a desire on the behalf of subordinates to satisfy the personal, political, and emotional needs of the president. In a sense, any administration's decision making can best be described through an analysis of two issues: (1) the decision-making structure and process and (2) the administrative order of battle--an assessment of alliances, coalitions, disagreements, and rivalries between organizations, officials, and the president in general and on any given issue.

A model of structural change. Accepting that organizations, officials, and the president fight repeated battles leads directly to the notion that any decision-making structure or management style will be unstable and subject to modification and significant change. This is the key assumption that allows the study of the causes of change (why change occurs) and the nature of change in decision-making processes (temporary or permanent). A generic model of change in decision-making structure can be developed. Figure 1 illustrates the influence of the international and domestic political environment on the relationships between the president and his senior advisers within the executive branch. The president's actions are defined by his personal leadership style, his political goals, and management strategies. The decision-making structure is what is produced by the interaction of these intra-administration dynamics. Change is represented by the second part of Figure 1 in which for time [T.sub.1], changes in the political environment and changes in the domestic political environment lead to changes in the organizational dynamics and therefore changes in decision-making structure.(4) The case studies below are designed to seek a better understanding of the causes of change. While changes in the domestic and international political environments and the shifting of the relationships among the senior decision-making officials were crucial, each of the decision-making changes was a deliberate action. These actions represent learning in the broadest sense of the word--officials within the administration feel that they have learned something about their political environment, their fellow decision makers, and/or the administration's decision processes. The specific actions they take reflect that learning.

[Figure 1 ILLUSTRATION OMITTED]

Case Studies

The following case studies are designed to identify the causes of changes in the decision-making structures of the Carter, Reagan, and Bush administrations. These administrations have been chosen specifically for their contrasts. In terms of overall judgments of presidential leadership style and administration decision-making processes, the differences are stark. Carter was often criticized for being an "amateur" in foreign policy, with no set belief system on international politics or coherent views of the national interest. The Carter administration is often seen as amateurish, plagued by bureaucratic infighting, and ultimately a failure in terms of both process and policy in the field of foreign policy (Davis 1979; Garthoff 1985; Smith 1986; Rosati 1987; Spencer 1988; Dumbrell 1993). Within an administration noted for foreign policy success, the Reagan administration was hampered by the president's legendary inattentiveness, an often crippling lack of consensus, and excessive decentralization that led to severe decision-making difficulties; Iran-Contra was only an exaggerated version of a routine problem (Tower Commission 1987; Rockman 1988; Hill and Williams 1990; Anderson 1990; Schieffer and Gates 1991; Cannon 1991). While President Bush was often described as too cautious, too lacking in foreign policy vision, or too wedded to traditional realist ideas, his administration, in contrast to those of Carter and Reagan, is often viewed as the epitome of professional and collegial decision making that led to foreign policy success and stable decision making (Kegley 1989; Mandelbaum 1990-1991; Rockman 1991; Peele 1992; David 1996; Hutchings 1997).

In terms of the domestic political environment and relations with Congress, the differences are a useful contrast. The Democrat Carter shared power with a Democratic Congress. Reagan's fellow Republicans controlled the Senate for the first six years of his terms in office, but Democrats controlled the House of Representatives. Republican Bush faced Democratic control of Congress. The international environment also provides differentiation. Carter was in office as detente began to fade. Reagan's terms in office began with a renewed cold war and ended with the beginning of a permanent thaw in U.S.-Soviet relations. Bush entered office as the USSR collapsed and the cold war ended.

The analysis that follows illustrates that all three presidents decided to modify their decision-making processes (presidential choice) and engineered end runs around their own interagency processes as a reaction to political pressures (political environment) and dissatisfaction with their own standard interagency decision-making process (organizational dynamics). This dissatisfaction can be seen as part of a process through which a president and his advisers learn about each other, their political environment, and their administrative processes. The importance of this aspect of learning will be stressed in the concluding sections of this article. An initial section that illustrates the similarities between the standard interagency design of the National Security Council (NSC) systems for each administration highlights a key point in this study: all administrations begin with a similar decision-making structure of hierarchically connected interagency committees working under the NSC. Carter (the amateur), Reagan (the delegator), and Bush (the professional) initiated similar systems at the beginning of their terms in office, and each eventually felt the need to temporarily abandon or permanently adjust those systems at specific points in their policy processes.

Each case study will use the following structure to examine the key elements that explain how and why the administration's decision process for arms control changed: (1) the standard interagency process used for arms control--the starting point from which to analyze change, (2) the domestic and international political environment within which arms control decisions were made, (3) the organizational dynamics within the administration within the issue area of arms control, and (4) the president's choice to make changes in the administration's decision-making processes. In the following narrative, organizational dynamics and the political environment may be discussed in tandem. Often, it is difficult to separate the different positions of departments and officials from their views of the political situation. Analysis of each case study is intended to identify the nature (temporary or permanent) and causes of these changes.

The Standard Interagency National Security Process

An examination of the way in which national security decision-making structures change requires some background on the nature of national security decision making in general. The interagency process is the backbone of national security decision making. This is true for a number of reasons. First, the basic reality of the executive branch structure means that the processes of the government are interagency in nature. The size of the executive branch, its ever-growing responsibilities (even in an age when the trend is to decrease the size of government), and the diverse range of those responsibilities necessitate that it be composed of numerous agencies with specialized competencies. However, the division of labor is ill defined in most instances. Very few governmental activities can be performed by one department or agency alone. Therefore, decisions will be interagency in nature, and coordination and control become one of the president's key concerns, concerns usually satisfied in part through the use of presidential staff within the Executive Office of the President (EOP) and White House Organization (WHO).

Second, the National Security Act of 1947 still provides the blueprint for designing foreign policy. It was fundamentally a method of creating inclusivity, diversity, and coordinated foreign policy. Inclusivity of agencies and departments allows participation in decision making for those agencies and departments that have important roles in implementation. This is related to the need for diversity. One of the essential premises of cabinet government, the ultimate interagency committee, is that a wide range of views is necessary to make the best decision possible. All agencies and departments with implementation and analytical responsibilities on a given issue can, in theory, bring a unique perspective on problems and solutions. Ultimately, the interagency process is a method of coordination. To make sure that U.S. government policy is truly national policy and not simply the outcome of organizations' undertaking disconnected tasks that may be operating at cross purposes, interagency decision making is necessary. The left hand must not only know what the right hand is doing, but the hands should be working together. The NSC is the ultimate expression of this search for coordination. It is the NSC process that provides the basic formal interagency structure for national security decision making, and it is within the context of the NSC process that changes in decision-making structure will be analyzed.

There are similarities and differences between the basic interagency NSC processes of all administrations. The permanent institutional makeup of the executive branch in general and the EOP and White House staff accounts for the similarities. In national security decisions, the NSC and its accompanying staff represent one such permanent structure or "deep structure" (Heclo 1991) that an incoming president inherits from his predecessor. These institutional structures can become constraints on both the process and substance of institutional decisions. The differences are often explained as the result of redesigning presidential decision making to fit the personal style of each incoming president. The use of the NSC staff and senior officials is an idiosyncratic factor that may change from president to president. Burke (1992) refers to these two perspectives, respectively, as viewing the presidency through "an organizational lens" or "a managerial lens." In essence, the NSC process of any administration shows modifications of the basic NSC interagency process within certain limits; they are variations on a theme.

One additional aspect of the NSC and its staff must be mentioned. In general, since the presidency of FDR, presidents have surrounded themselves with their own personal policy machinery to deal with the expansion of executive branch responsibilities. This has essentially added a fourth or "presidential branch" to the American governmental structure that operates within the EOP and WHO (Hess 1988; Hart 1995). This presidential branch works along with the statutory departments and agencies as complement or rival. It becomes either the coordinator of the interagency process or a replacement for it. In national security affairs, this has resulted in a natural rivalry between the secretary of state/State Department and the assistant to the president for national security (APNSA)/NSC staff. This tension is a fact of life within all three administrations and the most significant development in the national security decision-making process in the postwar era (Destler 1977; Senate Committee on Foreign Relations 1980; Destler 1980; Szanton 1980; Mulcahy 1986; Brzezinski 1987-1988; Shoemaker 1989).

Carter, Reagan, and Bush had somewhat similar views of the way foreign policy decisions should be made. Each believed in a collegial process within which advisers could speak freely and debate issues while advising the president and talked of a strong secretary of state, firmly in charge of U.S. policy making. Carter, however, also stated frequently that he wanted to be his own secretary of state. Ultimately, final decisions were to be made by an active and involved president (Brzezinski 1983, 8, 66, 513; Carter 1982, 52, 59; Vance 1983, 35; "Foreign Policy by Committee" 1977; Bonafede 1977, 1596; Thompson 1990, 6-7; Bush 1987, 173-74; Bush and Scowcroft 1998, 19). Despite these general similarities, the details of how this collegial decision making was to be structured made the three administrations' decision-making processes very different. Carter and Bush were both deeply involved in the policy process and made efforts to maintain the collegial nature of the policy-making group. Bush had great success in this regard, while brutal bureaucratic warfare erupted in the Carter administration between Secretary of State Cyrus Vance and APNSA Zbigniew Brzezinski. Reagan's role in the policy process was minimal. As a result of this lack of leadership, the system was so decentralized that collegiality, if it ever existed, quickly vanished into vicious rivalries and confusion.

In a system codesigned by Carter and APNSA Zbigniew Brzezinski, the NSC system consisted of two committees that would provide support for the statutory NSC itself. It was within these committees that the senior-level interagency process resided. The two committees were called the Policy Review Committee (PRC) and the Standing Coordination Committee (SCC). Attendant to the PRC were lower-level interdepartmental groups (IGs). The IGs were the working level of the PRC process and operated at the undersecretary or assistant secretary level.(5) The chair of the IG depended on the issue area. Importantly, the SCC was always chaired by the APNSA. This was Brzezinski's committee. In his absence, his deputy David Aaron would chair the committee. The SCC was designed to deal with shorter term problems that were interdepartmental by nature such as intelligence, arms control, and crisis management. The SCC contained within it subcabinet-level working groups chaired by NSC staffers. The membership of both the PRC and SCC was cabinet secretary level and included statutory NSC members and whoever else was needed (Bonafede 1977, 1598-1601; Carter Administration 1977; Korb 1979, 115-17; Vance 1983, 36; Brzezinski 1983, 59-60).

Within the Reagan administration, a decentralized structure was created for national security. Under the NSC, there were three main senior interagency groups (SIGs) for foreign policy (SIG-FP), defense policy (SIG-DP), and intelligence (SIG-I). Additional SIGs for other issues were added as needed. The SIGs were staffed by cabinet-level officials and chaired by the cabinet officer whose department had the most direct jurisdiction. Supporting each SIG were a number of IGs chaired and staffed at the assistant secretary level. The number of SIGs and IGs proliferated, often leaving it unclear who was in charge of policy and by what method the issue would be decided.(6) This system was put in place in an ad hoc manner in 1981 but not codified until 1982 during the tenure of Reagan's second APNSA William Clark. Reagan's initial APNSA Richard Allen was hamstrung by a structure that rendered him subordinate to Edwin Meese, counselor to the president, whose priorities lay in domestic affairs. The management of national security during Reagan's first year in office was often an afterthought or the haphazard result of a tug-of-war between Secretary of State Alexander Haig, Secretary of Defense Caspar Weinberger, and Meese. Clark took charge in early 1982, formalized the system, and elevated the NSC staff to its more typical role in national security--coordination of the SIG-IG interagency process and a steady drift toward controlling the policy-making process (Smith 1981; Szulc 1981, 184-85; Reagan Administration 1995b; Barnett 1983,220; Melbourne 1983; Endicott 1984, 193; Campbell 1986, 43).

The Bush system was much more centralized than the Reagan administration and slightly different from, though conceptually the same as, the Carter system. The NSC was supported by the Principals Committee (NSC/PC), a cabinet-level committee composed of NSC members minus the president and vice president and chaired by the APNSA (Brent Scowcroft). Supporting the NSC/PC was the Deputies Committee (NSC/DC), chaired by the deputy APNSA (Robert Gates). Its members were the principal deputies to the cabinet officers on the NSC/PC or key undersecretaries. Both committees added additional officials from the appropriate levels as needed. Below the NSC/DC were regionally or functionally differentiated interagency committees called policy coordinating committees (NSC/ PCCs), each chaired by an assistant secretary from the department whose expertise was most crucial to the issue.(7) The NSC/PC was the key advisory body to the president, setting agendas and preparing issues for presentation to the full NSC. The NSC/DC actually ran the day-to-day operations of foreign policy, tasking the NSC/PCCs and the departmental bureaucracies (Bush Administration 1995a, 1995b; Weinraub 1989; Gelb 1989; Hoffman 1989; Prados 1991, 548; Snow and Brown 1997, 198-201; Hutchings 1997, 22-23; Bush and Scowcroft 1998, 35).

The formal paperwork of the NSC process in each administration was similar, consisting of a review and directive process. For Carter, it was the Presidential Review Memorandum (PRIM), Presidential Directive (PD) system (Bonafede 1977, 1601; Vance 1983, 36-37; Moens 1990, 38). Under Reagan, National Security Study Directives (NSSDs) and National Security Decision Directives (NSDDs) comprised the formal paperwork of the system (Reagan Administration 1995a; Kirschten 1982; Weisman 1982; Kittle 1982). For Bush, it was dubbed the National Security Review (NSR), National Security Directive (NSD) system (Hutchings 1997, 24). PRMs, NSSDs, and NSRs were study papers. Once the president or cabinet-level NSC committees had decided that there was a specific issue or existing policy that needed further study, the APNSA or deputy would draw up a PRM, NSSD, or NSR that would provide instructions for the policy review. Those instructions contained process and policy guidelines. The document mapped out a bureaucratic route for the review along with a timetable for the review process.(8) These reviews were the basic method of staffing out policy analyses that were important enough to be elevated out of the normal routines of bureaucratic standard operating procedures. Once the interagency study had been completed, it would be addressed by the upper-level NSC committee for debate and discussion. While the Carter SCC and PRC did the heavy lifting at this stage, the Bush NSC/DC attempted to hammer out a consensus, ordered more reviews, or documented policy gridlock within a review before an issue was sent up to the NSC/PC. Within the Reagan administration, the policy might go to the SIGs or directly to the NSC or the more informal cabinet-level National Security Planning Group (NSPG). At this level, the study might be revised by the cabinet secretaries or recommended for transmittal to the president, whether consensus or disagreement was the result. This allowed the president to understand where agreements or disagreements among his senior advisers lay. If the issue was serious enough to merit a specific presidential guideline, a directive (PD, NSDD, or NSD) was drawn up by the APNSA for the president to sign, providing official presidential guidance on the policy.

The Carter PRC and SCC, the Reagan SIGs, and the Bush NSC/PC were analogous structures. All were cabinet-level committees that addressed foreign policy issues on behalf of the president. The Bush design eliminated the natural rivalry between committees at the same hierarchical level and added the NSC/DC to oversee the work of the lower-level interagency committees and departmental bureaucracies. All of these functions within the Carter administration fell to the PRC and SCC. However, the Carter SCC did have what was called a "mini-SCC" chaired by the deputy APNSA that assisted the SCC in managing foreign policy. The Reagan administration's proliferation of SIGs and IGs rendered attempts at coordinating policy making nearly futile. Clark fought this battle for just under two years before leaving the APNSA post for the safety of the Interior Department.

The processes of decision making at this formal level were also similar. The basic interagency design of hierarchically related committees channeling information, advice, and analysis up and down the chain of command formed the backbone of all three administrations' decision processes. Despite the differences in their decision processes that resulted, in part, from leadership style and ideological agreement/disagreement among the principal decision makers, the basic structure of all three administrations was the standard interagency model.(9)

The Carter Administration's March 1977 SALT II Decision

The Carter administration's decision-making process for its initial Strategic Arms Limitation Talks (SALT) proposal illustrates why presidents often choose to change aspects of their national security decision-making systems. In the case of the Carter administration, the standard interagency process was perceived to be inadequate because of its failure to produce policy options that met presidential guidelines and the political dynamics of the moment. Carter's desire for a specific type of policy--one that led to deep cuts in U.S.-Soviet strategic arsenals--led the president to move beyond his own interagency process. He achieved this by getting more deeply involved in the decision process to guarantee a specific policy choice and by giving the NSC staff primary responsibility for developing the negotiating proposals to the exclusion of many officials from the departmental bureaucracies.

The standard interagency process for arms control. Decision making for SALT initially followed the standard pattern. The SCC was the committee charged with decision making on SALT even though Paul Warnke, the director of the Arms Control and Disarmament Agency (ACDA), an arm of the State Department, also served as the U.S. chief negotiator for SALT. A SALT working group and ad hoc working groups created under its rubric provided the working-level staffing for the SCC decisions on SALT These committees contained representatives from the state and defense departments, the ACDA, the Joint Chiefs of Staff, and the CIA. However, they were chaired by an NSC staff member. SALT policy was developed by Presidential Review Memorandum--2 (PRM-2), issued on January 24, 1977. It asked for agency studies of SALT options. Once the PRM had been finished and included a full accounting of agency and interagency views, it would be given to the SCC for consideration (Gottemoeller 1978; Brzezinski 1983, 51; Garthoff 1985, 801; Moens 1990, 66). Roger Molander, a member of the NSC staff Defense Coordination Cluster, chaired the study, and William Hyland, of the USSR/East Europe Cluster, provided expertise on the USSR. The other key players at this level were Leslie Gelb, assistant secretary of state for political-military affairs, and Walter Slocombe, principal deputy assistant secretary of defense for international security affairs and director of the Department of Defense SALT Task Force. With representation from the State Department, the Department of Defense, and the NSC staff playing key roles, policy was reviewed through the routines of the interagency process (Talbott 1979, 44-45). The results of PRM-2 were presented to the SCC on February 3, 1977.

Political environment and organizational dynamics. The results of the interagency review contrasted sharply with Carter's vision of SALT II. The fruits of PRM-2, as presented in February, represented policy inertia and incremental change. Three options, called the "Slocombe Triptych," had been devised. All three were based on the 1974 Vladivostok Accords, an "in-principle" outline of SALT II agreed on by the Ford administration and the Soviet leadership (Talbott 1979, 46-47).

Neither Carter nor most of his senior advisers were pleased with PRM-2. They preferred a SALT II proposal based on deep cuts in both U.S. and Soviet strategic arsenals. In a 1975 autobiography and in his inaugural address, Carter had called for "elimination of nuclear capability" and deep reductions in nuclear arms.(10) Carter believed deeply in reducing the danger of nuclear war (Carter 1975, 1982; Hyland 1987, 208). Vice President Walter Mondale, APNSA Brzezinski and Deputy APNSA Aaron, and Secretary of Defense Harold Brown were also disappointed with the Vladivostok-based options, though Brzezinski's and Brown's motivations for deep cuts were based on rolling back the Soviet nuclear buildup that increasingly threatened the U.S. ICBM fleet. For them, it was not a matter of the horror of nuclear war but the strategic position of the United States (Brzezinski 1983, 146; Talbott 1979, 50; Garthoff 1985, 804). Deep cuts would satisfy Carter's goal of beginning the process of reversing the arms race. For the others, it was a way of slashing into the growing Soviet ICBM force and solving the problem of U.S. ICBM vulnerability. This group also felt uncomfortable with the idea of completing the Ford-Kissinger negotiations (Talbott 1979, 50). For strategic and political reasons, they wanted something new, something bold.

In addition, the Carter administration negotiated SALT with a very large shadow looming over its shoulder. Senator Henry Jackson of Washington, chair of the Subcommittee on Arms Control and International Security of the Senate Foreign Relations Committee, proved to be a formidable presence. Jackson was a critic of the Nixon-Ford-Kissinger approach to SALT. He was held in high esteem within the Senate, respected enough that his approval or disapproval of Carter's arms control policy was nearly a prerequisite for passage of a SALT II treaty. Jackson and Carter met on February 4. Jackson made it clear that he was disappointed with the Vladivostok Accords; the administration must look for deeper cuts in the Soviet arsenal. Jackson was concerned about the vulnerability of U.S. ICBMs to Soviet attack, a problem he felt was only made worse by the Vladivostok approach. Jackson outlined his argument in a twenty-three-page memo to Carter on February 15 (cowritten by Jackson aide Richard Perle). Carter had requested the memo from Jackson, then passed it around to his advisers and used it as a reference point during discussions on SALT (Talbott 1979, 52-53; Garthoff 1985, 803; Perle 1987, 100; Caldwell 1991, 40). Deeper cuts in the strategic arsenal were not simply a matter of Carter's preferences or strategic calculation; the administration needed politically feasible ideas. Both the domestic and international political context pushed the president, the NSC staff, and the Department of Defense toward a new direction on SALT.

Secretary of State Cyrus Vance and Warnke, however, were skeptical of the deep-cuts idea. They felt that reductions in both the U.S. and USSR arsenals were a more long-term goal. For now, the United States should focus on sealing the deal begun at Vladivostok and then move on to deep cuts for SALT III (Vance 1983, 47-49; Moens 1990, 71). Their views were in direct contrast to the rest of the administration.

The first SCC meeting steered discussion toward two possibilities: Vladivostok-based options and deeper reductions. The study of these options fell to the SALT working group, primarily Slocombe, Hyland, and Gelb (Moens 1990, 67). Continued SCC meetings accomplished little. The basic dispute between advocates of the two approaches was not settled. Carter had wanted the interagency process to produce an analysis of both a specific approach to arms control and plans for implementing that approach. The interagency process failed to produce the policy desired by the president.

Presidential decision to alter decision making. President Carter used a direct method of changing the process. He essentially seized the decision from the interagency process. Shielding the policy from the dissenting bureaucracies until it was publicly announced on the eve of U.S.-Soviet talks, he guaranteed that his ideas would be presented in full as the initial U.S. proposal for SALT II.

A March 19, 1997, SCC meeting was elevated to an NSC meeting with a surprise appearance by the president. Carter arrived at the SCC meeting with the intention of pushing the principal decision makers toward a final decision that favored deep cuts. The results of this meeting, even its existence, were kept quiet within the administration. Hyland of the NSC staff was charged with producing a formal proposal under the supervision of Deputy APNSA Aaron. This meant that the proposal Vance was scheduled to bring to Moscow was created not by the interagency process or the negotiating team but by the NSC staff. Other members of the SALT working group as well as the departmental representatives of the SALT negotiating team were kept unaware of the parameters of the proposal that Vance was to give the Soviets. Hyland accompanied Vance to Moscow with the job of keeping the contents of the proposal under wraps (Kaiser and Marder 1977; Talbott 1979, 59-60, 63-64; Hyland 1987, 211-12). In an organizational sense, the NSC had the job of keeping the State Department, Department of Defense, and the Joint Chiefs of Staff uninformed.

Conclusion. The USSR rejected the deep-cuts proposal outright on March 30, 1977, and the negotiations fell back on the Vladivostok options.(11) Carter did not achieve his policy objectives, but he did achieve his policy process objectives. The interagency process had given him a "business-as-usual" arms control policy, a continuation of the Nixon-Ford-Kissinger negotiations. It was not as innovative as Carter desired, nor did it produce a policy that fit the new president's goals. Preferring a bold deep-cuts approach, Carter stopped the standard interagency process in mid-stride, enforced his own view of the policy, and then protected the policy from those who disagreed with it. This led to permanent changes in the decision-making process for SALT. The NSC staff took a greater role in the routine staffing of ideas from the Geneva negotiations and in directing those negotiations. Since the new policy was essentially the policy that Vance and Warnke had supported all along, their role was elevated. The strongest proponents of the deep-cuts ideas, the Department of Defense and the Joint Chiefs of Staffs, saw their role diminish. If the bureaucratic battle had been Brown versus Vance and Warnke, the USSR had settled the issue temporarily by refusing the deep-cuts idea. The general shape of decision making was one in which Gelb (state) and Hyland (NSC) were the key working-level players. Their actions were supervised more directly by the SCC. Carter had a greater role than before in supervising the decision process. Warnke's instructions during negotiations and Vance's instructions when he met with Soviet Ambassador Dobrynin or Soviet Foreign Minister Andrei Gromyko were given by Brzezinski (Garthoff 1985, 811; Brzezinski 1983, 166-67). The basic rules and procedures for NSC decision making had not changed, but the subtle channels of authority had.(12) This would remain the administration's standard policy and process throughout the SALT II negotiations as they dragged on for two more years until a treaty was produced in June 1979.

President Reagan and the Eureka Proposals of 1982

The decision process that led to the Reagan administration's first Strategic Arms Reduction Talks (START) proposal, announced on May 9, 1982, provides a stark illustration of why presidents manipulate their decision-making processes from time to time. Interagency gridlock on arms control placed the administration under severe political pressure. More than a year had passed since the Reagan team assumed power, and the administration had yet to initiate talks on strategic arms control with the USSR. In the absence of administration leadership, the American public and Congress developed their own plans for arms control in the form of a nuclear freeze. Reagan was faced with a dilemma. The political environment demanded some form of arms control, and the ideas in vogue at the time were antithetical to the administration's preferences. However, the standard interagency process seemed incapable of coming up with alternatives. The solution for the president was to allow the NSC staff to push the interagency process harder and to bypass that process completely when necessary. The Eureka proposals represent a compromise designed by the NSC staff outside of the interagency process.

The standard interagency process for arms control. An interagency group (IG) on arms control provided the basic working-level forum for policy making. It began considering strategic arms control in early 1982. Under the arms control IG were any number of subordinate working groups created and convened as needed. The group was chaired by Richard Burt, assistant secretary of state for political-military affairs, and reported to the SIG-FP. However, the IG functioned as if Burt and Richard Perle, assistant secretary of defense for international security policy, were co-chairs. Since strategic modernization had priority over arms control within the administration, Perle's role as the Department of Defense representative was elevated to an equal of Butt (Talbott 1982, 13-14).

Organizational dynamics. Within the arms control IG, two competing positions formed and solidified. A group of "hardliners" wanted to use throw weight as the unit of measurement for arms control, while a rival group of "pragmatists" was willing to settle on the number of warheads or launchers as the unit for measuring the strategic balance.(13) The hardliners were led by Perle and included Arms Control and Disarmament Agency Director Eugene Rostow, START negotiator Edward Rowny, and Intermediate Nuclear Force talks negotiator Paul Nitze. The Department of Defense was the institutional backbone for this group. The pragmatists were based in the State Department. Burt was the main bureaucratic battler but was backed up by Secretary of State Haig. It became clear as these debates within the arms control IG continued that the hardliners saw delay in the resumption of U.S.-Soviet strategic arms control negotiations as a preferred alternative to arms control based on the old SALT measures of launchers and/or warheads. Organizational and bureaucratic gridlock had tied the arms control IG in knots (Talbott 1982, 235-48; Gordon 1983; Thompson 1993, 52-53; Winik 1996, 123).

Political environment. While administration officials slugged it out in the arms control IG, Congress and the American people were busy developing their own policies on arms control. The administration had waited so long to restart arms control with the USSR that it had lost the ability to set the agenda. Congress and various public advocacy groups began calling for a freeze on development, testing, and deployment of nuclear weapons. This nuclear freeze movement had its origins in the late 1970s among nongovernmental organizations (American Friends Services Committee) and intergovernmental organizations (United Nations) and had been introduced in the U.S. Senate by Senators George McGovern (D-SD) and Mark Hatfield (R-OR) in 1979. Action on the freeze remained contained at the grassroots level or in state legislatures until the winter of 1982, when the administration's delay allowed it to grow toward critical mass. Resolutions in the House, sponsored by Edward Markey (D-MA), and the Senate, sponsored by Edward Kennedy (D-MA) and Hatfield, were gaining cosponsors (Cole and Taylor 1983; Waller 1987; Meyer 1990; Rochan and Meyer 1997). The nuclear freeze idea had begun to fill the arms control policy vacuum left by the administration's inability to reach consensus on a new proposal. The administration had to act to regain that agenda or lose control of its policy to a Congress ready to surge ahead as the administration stumbled. Whether the freeze itself could pass both the Senate and House is not relevant. The fact that approximately twenty-five different resolutions on limiting strategic weapons, including the freeze, were in the congressional pipeline attests to the administration's political dilemmas (Miller 1982).

Presidential decision to alter decision making. The administration publicly denounced the freeze, and congressional allies introduced their own proposal for a freeze at equal and reduced levels of strategic weaponry (Towell 1982; Waller 1987, 92). However, the crucial response for the administration was internal. APNSA Clark and Deputy APNSA Robert McFarlane began to get involved in the decision-making process. Deadlines and specific strategic guidelines were set for the new START IG (the arms control IG had been split into two committees--one on the INF talks and one on START). A final deadline for consensus on an opening START proposal was set for May 1, 1982. An NSC meeting on April 21 and a SIG-FP meeting on April 29 failed to resolve the issue. The basic disagreements remained between the hardliners and pragmatists. Up to this point, the Joint Chiefs of Staff had yet to take sides. However, they began to lean toward the launcher limits as proposed by the pragmatists. Their need to carry out U.S. nuclear war plans as embodied by the revised Reagan Single Integrated Operating Plan (SLOP) of 1982 led them to lean away from deep cuts in the number of warheads in the U.S. nuclear arsenal that they felt were the logical extension of the arguments for reduced throw weight. From their perspective, they needed perhaps more rather than fewer warheads to hit all the targets required by the SIOP.

This State Department--Joint Chiefs of Staff alliance still failed to settle the issue. An all-day START IG meeting on May 1 designed to produce a recommendation for an NSC meeting on May 3 resulted in what has been called the "consensus proposal." The proposal had two phases. Phase I called for deep cuts in strategic weaponry based on launcher and warhead reductions. Phase II supplemented these cuts with indirect limitations on throw weight. Though this seemed to have something for everyone, the hardliners balked at the idea. However, McFarlane had attended the meeting and introduced the proposal to the president on the morning of May 3; Reagan reacted favorably. Though the NSC meeting deteriorated into rancorous debate, Reagan and Clark gave McFarlane the job of developing the administration's official proposal based on the consensus proposal, NSDD-32 "U.S. Approach to START Negotiations." In an equally critical move, Butt, not Perle, was given the job of writing the initial draft of Reagan's May 9 speech at his alma mater Eureka College, where the policy was announced to the public (Talbott 1982, 246-71; Gelb 1982; Reagan Administration 1995 c; Reagan 1982).

Conclusion. The organizational dynamics within the administration led to policy gridlock and resulted in increased political pressures on the administration to develop meaningful arms control proposals. To counter the freeze, the administration modified its standard interagency process by allowing Clark and McFarlane to push the START IG and by ultimately allowing McFarlane to manufacture a consensus where none existed. Reagan codified that consensus in an NSDD and announced it publicly as the Eureka proposals. However, this was a temporary adjustment to the decision process. The specific details of the proposal that the administration brought to the negotiations with the USSR in June 1982 were hammered out by the START IG and then approved by the NSC on June 25, the day before the negotiations were to begin. The proposal itself illustrates a lack of consensus, rather than a meaningful compromise. It contained warhead and launcher limits and sublimits designed specifically to limit Soviet throw weight (Talbott 1982, 270-71). The process had been modified out of political necessity but had been returned to the standard interagency process once the political pressure had been eased by the president's May 9 speech.

President Bush's Arms Control Speech of September 1991

The Bush administration decision-making process reveals quite a bit about how and why presidents make changes to or bypass their own interagency machinery. In this case, Bush and his senior advisers found the standard interagency process too cautious, slow, too lacking in innovative options, and unable to produce policies favored by the president. Bush saw a political opportunity to move ahead on arms control as the USSR began to crumble in the aftermath of the coup against Gorbachev in late summer 1991. For this reason, in the case study presented here, as well as in other instances, Bush chose to manipulate the policy process, taking control of it, excluding officials, and holding the decision within a tight circle of advisers until he was ready to announce it to the public.

The standard interagency process for arms control. Though Bush was succeeding an administration in which he served as vice president, a major element of his approach to arms control was the question of how much the administration should try to "Bushify" the START treaty that had nearly been completed by the Reagan administration. For this reason, the incoming administration began with a comprehensive review of arms control, rather than simply starting from where the negotiations with the USSR had been left in 1988, despite the fact that a new round of START negotiations was scheduled to begin in February 1989, less than a month after the inauguration (Oberdorfer 1989).

The administration rescheduled the next round of START for June 1989 and began a comprehensive review of national security and arms control that lasted throughout the summer. The National Security Review on arms control, NSR-14, was chaired by Arnold Kanter, the NSC staff's senior director for defense policy and arms control. Other key contributors to the interagency review were Under Secretary of State for International Security Affairs Reginald Bartholomew and START negotiator Richard Burt. NSR-14 started from square one, considering a large range of strategic arms control proposals not included in the pending draft treaty (Gordon 1989a, 1989b; Rowny 1992, 214-15; Beschloss and Talbott 1993, 116). U.S. policy to open the START was eventually hammered out in a series of NSC meetings on June 6, 14, and 15, immediately preceding the first round of negotiations (Gordon 1989c, 1989d, 1989e; Toth 1989). Very little was actually accomplished at these meetings. The U.S. position on START was altered only slightly. For the purposes of this research, the key aspect of this decision is that it began with an interagency review and ended with an interagency decision within the formal structure of the NSC.

Organizational dynamics. By the fall of 1989, the Bush administration began what became almost standard operating procedure for it: moving outside the interagency process by creating small groups of officials who concentrated on a specific issue as a replacement for the departmental bureaucracies. In particular, two examples can be found in arms control: (1) the use of four officials--Robert Blackwill and Condoleezza Rice of the NSC staff and Robert Zoellick and Dennis Ross of the State Department--in preparation for the Malta Summit of December 1989 and (2) the use of a secret interagency committee dubbed the "Ungroup" (because all members were instructed to deny its existence) in the spring and summer of 1991 during the final negotiations on START (Oberdorfer 1998, 376; Beschloss and Talbott 1993,365; Baker 1995,660). However, to focus the research, this study only considers the post-START I decisions in the fall and winter of 1991-1992.

Political environment. The failed August 1991 coup against Soviet President Mikhail Gorbachev that ultimately led to the dissolution of the Communist Party of the Soviet Union, the collapse of the USSR itself, and the birth of fifteen post-Soviet nations was also the impetus for Bush's next decision on arms control. In the days immediately following the coup of August 20 to 23, Bush and APNSA Brent Scowcroft began to discuss a follow-up to START. The issue was never placed into the interagency process. It remained at the senior-most levels of the administration. Bush desired a rapid decision-making process; roughly a month passed between these first discussions and Bush's speech of September 27. Bush also wanted a bold approach to START II. In fact, Bush's speech reversed decades of U.S. nuclear policy in calling for an elimination of all U.S. tactical nuclear weapons as well as a new short-range attack missile (SRAM) and both U.S. ICBM modernization programs--the MX and mobile version of the small ICBM (Bush 1993). Bush had often been criticized as too cautious, pragmatic, or incremental in his policy making (Kegley 1989, 717; Peele 1992, 151; David 1996, 197; Snow and Brown 1997, 120-21). He delighted in confounding his critics with a surprise policy turnaround. It had taken the administration nearly two and half years to tie up the loose ends on Reagan's START treaty. Bush hoped to prove his critics wrong by staying ahead of the game as the USSR collapsed. Bush seized the political moment by using the opportunity given to him by the turmoil in the USSR to move quickly toward far-reaching arms control. There were also strategic considerations at play. Moving quickly on arms control while a willing partner--Gorbachev--still had centralized control of the USSR and its military assets allowed the administration to take advantage of the situation to enhance U.S. security.

Presidential decision to alter decision making. It would seem that in the aftermath of such sweeping change in the USSR in 1991 that the United States might need to completely reassess its foreign policy goals through potentially mammoth and comprehensive interdepartmental studies. The purpose of U.S. foreign policy since the 1940s--the containment of communism--was becoming obsolete over only a few short months. However, the Bush administration first moved in the opposite direction in the fall of 1991, toward top-heavy decision making and a notable lack of comprehensive study; Bush felt he could only achieve his objectives by moving outside his own interagency process. With history accelerating at a rapid pace, the Bush administration had to sprint to keep up, and the standard interagency process was simply too slow for this. In particular, gaining some basic understanding of START II while Gorbachev still had the power to commit the USSR to agreements and while the hardliners in the USSR were on the defensive was crucial. In addition, 1992 was an election year. Signing START II sometime in 1992 might help Bush's reelection chances. Political pressures demanded swift decisions, politically and strategically timely decisions. It also demanded policy options that could move arms control forward by leaps and bounds rather than cautious baby steps. Bush turned to the methods experience had taught him were the best--a small group of his friends making decisions in secret.

Bush's strategy to achieve his policy-making goals was to make decisions at a blistering pace, with strictly limited participation, while pushing for imaginative options and carefully timing the release of the information. Speed, secrecy, and surprise were the watchwords of the process. On the back porch of Bush's Walker's Point home in Kennebunkport, Maine, while vacationing during the third week in August, Bush and Scowcroft decided to move quickly to take advantage of the Soviet hardliners' weaknesses and to act before more changes occurred within the USSR. Secretary of Defense Dick Cheney and U.S. Ambassador to the USSR Robert Strauss came to Kennebunkport for consultations on the issue on August 27. Scowcroft asked Cheney to develop some bold proposals with the aid of the Joint Chiefs of Staff. Bush and Scowcroft had some additional brainstorming sessions on August 28 at breakfast and on Bush's speedboat off the coast of Maine (Yang 1991; Rosenthal 1991; Bush and Scowcroft 1998,545). During the first week of September, Bush, Cheney, and Scowcroft met again to discuss the issue. This was followed up by a meeting of Bush's informal small group of senior advisers on September 5 that included Bush, Secretary of State James Baker, Cheney, Scowcroft, Chairman of the Joint Chiefs General Colin Powell, Deputy APNSA Gates, Chief of Staff John Sununu, and Vice President Dan Quayle. Bush made it absolutely clear that he wanted creativity and new ideas for START II (Powell 1995, 541). Here Bush was managing both the content--innovative proposals only--and the participation in the process: knowledge of these discussions was held to the small group and some of their senior aides. The task of developing these proposals was given to Cheney and Powell. The two met with Scowcroft several times between September 9 and 17 to consider the issues. A formal presentation of possible options made to the small group on September 17 included a number of sweeping proposals, and the president approved several. Bush, Scowcroft, and Cheney worked on the draft together to maintain total secrecy on the contents and even existence of a follow-up to START. On the night of September 27, after only minimal leaks that something major was in the works, Bush gave an Oval Office address containing new arms control proposals. Bush's need to control the pace of the decision is evident in how quickly he pushed the administration--five weeks passed from musings by Bush and Scowcroft in Kennebunkport to Bush's speech of September 27. His need to fine-tune the timing of the public release of the policy was also a critical motivation for his decision-making style during this period. As the policy-making process came to a close in the last week of September, Bush was scheduled to give the address containing these new proposals at the National Defense University at Fort Leslie McNair in Washington. However, when some of the details were printed in the Washington Post that morning, Bush decided that to drown out the press coverage, he would have to go straight to the American people with a nationally televised Oval Office address (Yang 1991; Rosenthal 1991).

Conclusion. Gorbachev responded with his own set of proposals on October 5, and negotiations began on START II. By mid-December, the debate on a follow-up to Bush's September speech and responses to Gorbachev's October proposals was taking place. Ultimately, these resulted in Bush's January 28, 1992, State of the Union speech calling for removal of all multiple independently targeted reentry vehicles (MIRVs) and elimination of mobile systems from both sides' arsenals, as well as deep cuts in both sides' arsenals down to forty-seven hundred warheads each (Bush 1993). Bush had initially been cautious. He took a slow approach to policy making for START as he entered office. For this reason, the standard interagency process was the appropriate vehicle in which to carry decision making. However, in the fall of 1991, when Bush had specific policy-making ideas that needed to be developed quickly, the slow and incremental approach of the interagency process was unacceptable. He elevated decision making to the upper reaches of the administration, pushing it toward rapid and bold policies, which were kept under wraps until he was ready to reveal them. The START II agreement was not signed, however, until January 1993. Once the initial proposals were on the table, the Bush administration returned to a more typical interagency process to make decisions. It was a Bush-style interagency process--deep involvement of the president and his small group of advisers--but it was not the secretive and exclusive process that was the norm in the summer and fall of 1991. Bush's changes in the process were temporary modifications made to produce a specific policy option under unique political circumstances.

Revisiting the Model

In light of the case studies above, several aspects of Walcott and Hult's (1995) basic framework can be developed further. Their model is essentially dynamic. The dependent variables of structure emergence, stability of structure, and nature of structure (differentiation of units) are influenced by the political environment, presidential choice, and organizational dynamics. The independent variables represent initial and constant influences on a decision-making structure. This means that when examining changes in decision-making structure, numerous characteristics of decision making must be taken into consideration within each category of independent variables that Walcott and Hult identify. In the study of national security decision making, as described above, the organizational process, bureaucratic politics, and presidential management models help identify those variables. Table 1 summarizes the independent variables that influenced changes in the decision-making processes in the above case studies.

TABLE 1 Variables Considered in the Carter, Reagan, and Bush Case Studies

Political        Congressional influences     Carter, Reagan
environment
                 Public opinion               Reagan

                 Philosophical and
                 strategic differences
                 between the beliefs of the
                 dominant coalition within
                 an administration and
                 vocal and politically
                 active members of Congress
                 and the public               Reagan

                 Loss of control of policy
                 agenda; alternative
                 agendas from outside the
                 executive branch gaining
                 currency                     Reagan

                 Rapidly changing
                 international situations     Bush

Presidential     Level of agreement between
choice           presidential policy
                 preferences and that of
                 advisers and bureaucracies   Carter

                 Strong presidential
                 beliefs                      Carter

                 Presidential desire for
                 carefully orchestrated
                 policy presentations         Carter, Bush

                 Presidential desire to
                 determine and manipulate
                 the pace at which policy
                 should be made               Reagan, Bush

                 Presidential desire to
                 defy common criticisms of
                 his leadership               Bush

                 Level of presidential
                 involvement in determining
                 policy choices               Carter, Reagan, Bush

                 Extent to which
                 administration policies
                 deviate from existing
                 policies                     Carter, Reagan, Bush

Organizational   Organizational location of
dynamics         policy process-which
                 department or committee is
                 charged with producing a
                 decision                     Carter

                 Organizational location of
                 directive or
                 speech-drafting process      Reagan

                 Philosophical and
                 strategic differences
                 between the beliefs
                 of senior advisers and
                 organizations                Carter, Reagan

                 Officials excluded at
                 various stages of the
                 process                      Carter, Reagan, Bush

What these variables suggest about changes in presidential decision making is addressed in the concluding section of this article. However, this list reemphasizes the basic assumptions of Walcott and Hult (1995) and this article: understanding the staffing and changes in staffing for presidential decision making necessitates research into system levels, state levels, and individual levels of analysis. Seeking independent variables within the political environment, organizational dynamics, and presidential choice, respectively, accomplishes this. Figure 1 also illustrates how these variables are related.

In addition, the variables in Table 1 suggest some additions to the theories of Walcott and Hult (1995). Several commonalities or perhaps "themes" cut across each category developed by Walcott and Hult. It might be best to describe them as contextual factors that shape many of the variables that Walcott and Hult focus on. These contextual factors and their addition to Walcott and Hult's independent variables will be used to develop the initial framework. The discussion of each variable will include a research hypothesis to test the impact of the variable.

First, political environment, presidential choice, and organizational dynamics need to be seen as a function of time. Walcott and Hult's (1995) independent variables--emergence of structure, stability of structure, and nature of structure--include time to a certain extent, but the notion of time needs to be emphasized. Decision-making processes are by their nature dynamic and unstable, as described above. To truly understand decision making, the changes in variables over time need to be examined within a single administration and across administrations. The impression a scholar might receive of a specific administration's decision-making process will depend on when that process is examined. For example, studies of the Kennedy administration that judge its decision process by examining the Cuban missile crisis will produce a skewed impression of its standard national security machinery.

Comparative case studies of presidential decision-making adaptations could search for patterns of change that might exist within several administrations or patterns of change that continue across administrations. The existence of a similar pattern in a number of administrations would suggest that the pressures placed on any administration and the subsequent administrative reactions are essentially the same for all presidencies. Patterns found across administrations might present a larger picture of cycles of change (centralization and decentralization, for example) with a frequency that lasts for more than one administration. Are changes in decision making linear, cyclical, or somewhat random? Are there patterns specific to Democratic and Republican administrations? Are these patterns similar for different issues? These types of studies focus on the dependent variable of stability; do national security decision-making processes evolve in an identifiable and unique way (Newmann 1999)? To do justice to these questions would call for the tracing of decision-making structure and process for several issues across several administrations.

Second, the role of ideology and ideological congruence both within an administration and between an administration, Congress, and the public is evident in all three categories of variables defined by Walcott and Hult (1995). Real philosophical differences over the nature of international affairs (idealism or realism) or strategic issues (what constitutes an adequate deterrent) carry an impact that is felt within the political environment, presidential choice, and organizational dynamics. This is similar to the way Hult and Walcott (1990, 5041) define controversy over goals (preferred policy outcomes) and technologies (means to achieve policy outcomes). Controversy exists when two or more actors are certain about their views but their views are different. However, a specific focus on ideology highlights the fault lines within the domestic polity of the United States.

For example, the nature of a pluralistic society such as the United States results in political parties that have distinct ideological wings. The Reagan administration is an illustration of an administration populated with members of the moderate and conservative wings of the Republican Party. The rivalry between those wings was more contentious in the 1980s, as the conservative faction overtook the moderates, than at any time since the 1950s. Similarly, the Brzezinski-Vance feud in the Carter administration is often considered to be a reflection of the Democratic Party's struggle to rebuild its views on foreign policy after the Vietnam War (Rosati 1987). These historical ideological contests provide some context for political environment, presidential choice, and organizational dynamics. A second example might be provided by the nature of presidential elections. Presidents often believe that they have a mandate to govern. Is it a mandate that stretches across all issues a president cares about? Again, the Reagan administration provides an example. The president had greater success implementing his mandate in the realm of economic policy. His mandate for shutting down U.S.-Soviet arms control was less well received. A president who wins a plurality rather than a majority, such as Nixon or Clinton, could have an even more difficult time implementing policy. This relates to decision making in the sense that presidents may misperceive the political environment, resulting in greater uncertainty on presidential choices and a free-for-all style of organizational dynamics. An even larger problem could stem from a lack of consensus within U.S. public opinion. Defining the U.S. national interest during the periods immediately following the Vietnam War and the cold war seems inherently more difficult than at times when threats to the United States are both more stable and more obvious.

A research agenda based on these issues would examine the definitions of the U.S. national interest in two ways: (1) level of consensus within an administration and (2) level of consensus between an administration and Congress and the public. Examinations of national security decision making during periods of high or low consensus would help identify whether there is a relationship between the extent of consensus within government and society and the stability of decision making within an administration. Lack of consensus may make more of a difference with some issues than others.

Third, when studying national security decision making, the international system needs to have a special place in each category of independent variables suggested by Walcott and Hult (1995). The definition of U.S. national interests and the security threats to the United States are influences on every aspect of decision making. When it comes to national security and foreign policy, the study of U.S. decision making and the current study of international relations meet head on. The dominant neorealist paradigm of international relations essentially factors foreign and national security policy decision making out of the field of international politics (Waltz 1979, 1996; Elman 1996). It is based on the notion of the structure of the international system as the independent variable. Neorealism argues that nation-state behavior is a response to the international system based on rational choice; all nations are simply utility maximizers, and formal models can be used to map out their behavior.(14) Supporters of neorealist theories might advise Walcott and Hult to jettison the domestic political component of their political environment category while eliminating presidential choice and organizational dynamics entirely.

Obviously, the analysis in this article disagrees. However, some way to rejoin the subfield of international politics with the study of decision making that goes beyond rational choice models to recapture the richness and complexity of decision making needs to be found. A possible method of doing this is to relax the strictures of rational choice decision but accept the premise that changes in the international system are the impetus for foreign and national security policy changes. The variables discussed within the case studies and illuminated in the Walcott and Hult (1995) framework help account for the type of reaction a nation has to changes in the international system, the amount of time it takes for a nation to react, and the permanence of those changes. Die-hard system-level scholars might see all of Walcott and Hult's variables save the international system as intervening variables. However, if scholars of international relations accept that non-system-level factors can be legitimate independent variables, future research could include more attention to the ways in which domestic politics, organizational dynamics, and presidential choice affect the international system (Snyder 1991; Katzenstein 1996; Christensen 1996). For example, the inability of the Reagan administration to reach consensus had a clear influence on the nuclear freeze movement's ability to gain momentum. Figure 1 contains "shadow arrows" illustrating potential feedback loops in which changes within the executive branch affect the domestic and international political environment, as well as potential feedback in which changes within the domestic environment influence the international environment.

In sum, a refinement of Walcott and Hult (1995) in light of these case studies of national security decision making suggests that three overarching contextual factors play a larger role than other influences in determining the emergence, stability, and nature of decision-making structures. They influence each of the independent variables described by Walcott and Hult.(15) In adapting the model for future research, these contextual factors should be defined before the details of the independent variables are explored. They will place the details in a larger political context and help bring more depth to the study.

Conclusion

This article is intended to bring further understanding to the way presidents make modifications in their national security decision processes. In each of the case studies above, the president has deliberately changed aspects of this process in an effort to achieve specific policy goals. Two important issues are raised: (1) the causes of those changes in decision making and (2) the relationship between process and policy.

Examination of the decision processes in the Carter, Reagan, and Bush cases reveals a different independent variable for each change in decision making. Carter's desire for a new approach to arms control led him to adjust the decision-making process to achieve that end--the interagency process was temporarily abandoned, and officials who might object to the policy were excluded. In the Reagan case, the administration's failure to move on arms control led to the rise of an alternative agenda based in Congress that threatened the administration's basic national security goals. The temporary seizing of policy by the NSC staff represents the president's need to combat that congressional agenda with one of his own, an agenda that the standard interagency process seemed incapable of developing. Bush and some of his key aides saw the aftermath of the coup in the USSR as a pathway to deeper cuts in nuclear weaponry. The timing of any proposal was crucial, and the standard interagency process was considered to be incapable of developing a new policy in sufficient time. The decision process was modified to make sure the new policy would be ready for public airing in time to take advantage of strategic and political opportunities.

All three of these cases exhibit learning in the sense that the president decided that his standard interagency process was no longer the optimal method of making policy. However, each administration exhibits a different impetus for drawing this conclusion. The adaptations themselves could be categorized based on the cause of change. This reiterates Walcott and Hult's (1995) notion that there are multiple independent variables. Carter's decisionmaking modifications were influenced by presidential choice and the domestic political environment. These modifications can be considered innovative; Carter's need for a bolder arms control policy was based primarily on his desire to be different from his predecessors. He was instituting policy change on his own initiative. Reagan's administrative changes were essentially reactive. The administration was influenced by a domestic political environment that seemed to be turning against it. Its arms control policies were considered to be inadequate; policy change and decision-making change were made as a reaction to shifts in the politics of arms control. The Bush changes were mainly a function of the international political environment and presidential choice, with perhaps a nod to the domestic political considerations of a coming election year. These decision adaptations can be classified as opportunistic to differentiate them from the Carter and Reagan adaptations. The Bush speech of September 1991 was both innovative and reactive. The combination of motivations and the added factor of time pressure (exploiting a window of opportunity) felt by the administration make the case significantly different from Carter's innovation and Reagan's reaction.

This brings up the notion of the relationship between process and policy. Though the motivations for making changes in the decision-making process were different in each case, one similarity is crucial. Each president adjusted the process because he felt (or felt he had learned) that the standard interagency process was inadequate to the task of providing him with the policy he felt he needed at the moment. Decision-making structures and procedures are put in place to serve the president. When they no longer do so, the president will find another method of making decisions. Further research in this area might be able to develop some theoretical decision rules that presidents and senior officials have used in deciding when to adapt decision-making procedures.

All three cases illustrate a relationship between the process and the timing of the policy. Despite the name of the discipline of political science, practitioners might argue that politics is more art than science. Presidents have a feel for policy, and that sensitivity is partly a function of the timing of an issue. Agendas may be set, but there are limited opportunities to see those agendas enshrined into policy (Light 1982). Often presidents must act quickly before a perceived opportunity disappears. These limited opportunities have been called "policy windows" (Kingdon 1984, 174-80). These policy windows are instances when contentious issues are settled, dormant agendas spring to life, or new ideas swiftly mature. But these new or renewed policies must be generated from a process. The three cases above illustrate a presidential judgment that the policy needs of the moment could not be satisfied by the standard interagency process. None of the presidents abandoned their interagency systems or made deep changes in their NSC committee structures and processes. Reagan and Bush made temporary adjustments, and even Carter's permanent change was slight. This suggests that presidents come to realize that they need to make decisions at different speeds and with different levels of effort. The interagency process operates at a single speed and requires the difficult task of consensus building among a group of advisers with strong opinions and stronger egos. Placing every policy into the crucible of the interagency process is inadequate. Some opportunities would be lost as policy windows close while committees debate. Innovations, swift reactions to political developments, and seizing opportunities may be instances when the standard interagency process is simply not enough. Certain policy needs require new decision styles.

This is a reversal of the important study of whether different styles of policy process will produce different policies (Burke and Greenstein 1989). Here policies with differing characteristics require a president to use different styles of policy process. As noted above, Chandler's (1962) dictum that "structure follows strategy" holds true for these cases. The national security decision-making structure was modified with the specific intention of giving it the ability to develop a specific policy preferred by the president and/or to prevent the interagency process from precluding such an ability. The above is an example of top-down decision making. The routine decision making of the interagency process can be seen as bottom-up decision making. In the standard interagency process, the senior levels of the administration may ask very broad and general questions, such as, What are the best ways of achieving U.S. national security goals through arms control? The NSC committees and departmental bureaucracies are tasked with answering it through the standard interagency process described above. However, in the arms control cases studied here, the president and his advisers already had answers--specific policies they favored. In this sense, what they want from the interagency process are refinements of the policy and its implementation, analyses that make all the political and strategic implications clear and, perhaps, arguments that will legitimize the policy. Essentially, the president is saying, Here is what I want to do. You are the experts. Show me how to do it. Such a task is significantly different from what is asked of the interagency process in the routine conception of bottom-up decision making.

A normative question concerning the nature of the changes in the national security process remains at issue: is modifying the standard interagency process or adapting the routine decision structure a wise way to make decisions, or can this lead to poorly staffed and insufficiently developed policies? The answer is beyond the scope of this article, but it is a key question for further research. A preliminary answer is twofold. First, whether or not making ad hoc changes to the interagency process in mid-administration is advisable is irrelevant in some respects. These types of changes are inevitable. The nature of presidential decision making is that presidents make decisions in whatever manner they desire, regardless of how they intended to make decisions when they first entered office. The standard interagency process designed for new administrations is only a theory. Incoming presidents have not been president before. The decision-making structure represents hypotheses about what administration resources need to be brought to bear for decision making. Presidents will discover whether their initial hypotheses are true. Generally, there are periods of adjustment and transition and modifications made to both committee structures and procedures. One of the marked tendencies in foreign policy decision making is to centralize administrative procedures toward advisers--both line and staff--who are closer (in both physical proximity and political goals) to the president (Newmann 1999). This is evident in the rise of the NSC staff to the center of the foreign policy process. Despite all the resources of the departmental bureaucracies and the coordinating mechanisms of the interagency process, presidents have felt the need to create their own foreign policy bureaucracies. The entire executive office of the presidency is a replica of the executive branch that owes its allegiance to the presidency rather than to the departments or agencies that have ties to both the executive and legislative branches. As Henry Kissinger (1979, 31) stated, "Presidents listen to those advisers whose views they think they need, not to those who insist on a hearing because of the organization chart." The same is true for committees and staffing procedures; presidents will make policy in their own way.

Second, further research on whether changes in the standard decision-making procedures represent a potential danger to high-quality decisions might consider the following. The full interagency process was designed for inclusivity of executive departments, diversity of opinion, and coordination of policy. It is the style of decision a president sees as best before he is pressured and burdened by the realities of life inside the Oval Office. Changing the method of decision on the run as a response to immediate political needs at first sounds like a recipe for making poor decisions, perhaps by excluding diverse opinions or poorly coordinating implementation. For example, the Iran-Contra case in the Reagan years is an example of a change in decision-making practices that led to policy disaster (Draper 1991; Walsh 1994). However, President Nixon's opening to China is an example of a change in policy making that led to significant and far-reaching strategic developments (Garthoff 1985; Foot 1997). One key difference between the two is that the Iran-Contra initiatives were never vetted through the interagency process once the initial political need for a breakthrough ended. The opening to China was made through a change in the decision-making process but was moved back into the standard processes once the policy initiative had become reality.

A hypothesis for future research could be the following: as long as the interagency process is not completely abandoned, there may be little harm done to standard decision making and the ability for national security processes to function smoothly. In the cases above, temporary modifications of the decision structure were necessary to achieve major policy breakthroughs (Carter), to break the gridlock of the departmental rivalry when bureaucratic disagreements over small issues may jeopardize the larger issues on which there is agreement (Reagan), or to take advantage of swiftly unfolding political developments (Bush). Given that one of the most difficult presidential tasks is getting the career bureaucrats to follow the policies of the temporary occupant of the White House, changing the decision-making process temporarily or making slight alterations to the national security decision process is not only an important tool of the presidency but also part of the president's growth in office. In all three cases, though the standard decision process was pushed aside, it was not destroyed. As long as the interagency structures continue to function, they will have their turn at policy analysis and development. If the USSR had accepted Carter's deep-cuts proposal, the standard interagency process would have been part of the policy development for negotiations that would have turned a diplomatic breakthrough into a detailed arms control treaty. Following the May 9 speech, the Eureka proposal was vetted through the START IG. The Bush proposals of September 1991 led to a START II agreement in January 1993. The standard interagency process was at the heart of turning these proposals and the high-level negotiations of senior advisers into a detailed implementation plan. The modifications represent additions to the standard decision-making model as presidents attempt to control both policy and process.

(1.) These system-level theories often consider the decision-making process within the U.S. executive branch to be irrelevant to the outcome of national security policy. Policy is a reaction to international systemic factors; therefore, the way that policy is generated is of no interest to those examining the international system as an independent variable (see Waltz 1979; Keohane 1986; Baldwin 1993). There have been exceptions (see Hammond 1986; Kozak and Keagle 1988; Bender and Hammond 1992; Welch 1992; Allison and Zelikow 1999). However, in most cases, the focus of research on decision making has been centered on rational choice or cognitive models (see Geva and Mintz 1997).

(2.) Analyses that focus on the president's power to manipulate the foreign policy process often were written as direct criticisms of the bureaucratic paradigm.

(3.) Walcott and Hult (1995, 14-16) identify the following different managerial styles: hierarchy, collegial-competitive, collegial-mediative, collegial-consensual, adjudicative, adversarial, and market.

(4.) This recognizes that changes in the international environment can lead to changes in the domestic environment.

(5.) This structure was outlined in "Presidential Directive/NCS-2" (PD-2) in Carter Administration (1977). Details on the operation of this structure can be found in Bonafede (1977, 1598-1601), Korb (1979, 115-19), and Rosati (1987, 182).

(6.) The administration claimed that there were twenty-five senior interagency groups (SIGs) and fifty-five interagency groups (IGS); however, others have tallied the SIG- and IG-level committees at nearly two hundred. See, respectively, National Security Decision Directive 276, "National Security Interagency Process," June 6, 1987, as printed in Simpson (1995, 808), and Prados (1991, 463).

(7.) National Security Council policy coordinating committees (NSC/PCCs) were developed for the following: Europe; USSR; Latin America; East Asia; Africa; Near East/South Asia; Defense; International Economics; Intelligence; Arms Control; Counterterrorism; Special Activities; Refugees; International Oceans, Environment, and Science Affairs; Resources for International Affairs Programs; Emergency Preparedness/Mobilization Planning; Nonproliferation Policy; Technology Transfer Policy; and National Security Telecommunications. See "NSD-1, Reorganization of the National Security Council," in Bush Administration (1995a), and "NSD-10, List of Additional Policy Coordinating Committees," in Bush Administration (1995b).

(8.) Within the Carter administration, this included assigning the issue to either the Policy Review Committee (PRC) or the Standing Coordination Committee (SCC). If the Presidential Review Memorandum (PRM) was placed under the PRC, there would be one lead department. Competition for that role was intense. Not only would the chosen department have the lead role in the decision process, but it would also have the responsibility (or privilege) of writing the final draft of the study that would eventually be considered in the PRC (Drew 1978, 101). Under the Reagan administration, a battle usually ensued over which SIG or IG would lead the study. Often this led to the creation of new SIGs and IGs. Within the Bush administration, the NSC/Deputies Committee usually decided which NSC/PCC had jurisdiction over an issue.

(9.) Carter's decision making fell victim to the clash between Vance's idealism and Brzezinski's realism. The Reagan administration had a similar split between conservative hardliners and more pragmatic realists. The Bush administration proved to be a much more ideologically homogeneous group of cautious and pragmatic realists. On Carter, see Rosati (1987) and Smith (1986). On Reagan, see Kirschten (1983), Barnett (1983, 224-25), Destler (1983, 122), Garthoff (1994, 45), and Cannon (1991, 305-6). The agreement within the Bush administration is best revealed through a look at their views on the end of the cold war in 1989 (see Woodward 1991; Baker 1995; Powell 1995; Gates 1996; Oberdorfer 1998; Bush and Scowcroft 1998).

(10.) In a preinauguration meeting with the Joint Chiefs of Staff (JCS), Carter asked the JCS about the possibility of a "minimum deterrent" force, just enough nuclear weapons to provide the United States with a retaliatory response to any Soviet first strike. During the discussion, the number of missiles mentioned was 200 to 250 missiles and reportedly even included a "Blue Water" option in which the United States relied on submarine-launched ballistic missiles (SLBMs) alone (Brzezinski 1983, 157; Walsh and Wilson 1977).

(11.) The USSR appreciated continuity more than boldness in negotiations. It also had strategic reasons for not wanting to cut into a strategic arsenal that had it nearing parity or achieving superiority, depending on the analysis. Vance actually came to Moscow with a fallback position based on the Vladivostok Accords. An opponent of deep cuts and a loser in the intra-administration battle, Vance had secured this fallback position--the "deferral option"--at an NSC meeting on March 22 (Talbott 1984, 62; Vance 1983, 52).

(12.) Back channels were also opened up between Carter and his senior advisers on the U.S. side and Dobrynin and Gromyko on the Soviet side. These provided an alternate method of communicating positions or floating trial balloons. They would keep the discussion at the senior level, bypassing the Geneva delegations and any bureaucratic disagreements at that level that might prompt leaks.

(13.) Hardliners refers to a faction in the administration who saw the USSR as the source of all international threats. They viewed detente as a mistake and arms control as a fool's burden. The pragmatists saw the USSR as a threat but preferred to use old-style balance of power politics, such as detente and arms control, to contain it. Throw weight refers to the overall weight that a missile can carry. Soviet missiles were larger than their U.S. counterparts and therefore had much greater throw weight. The issue becomes important in the context of the deployment of multiple independently targeted reentry vehicles (MIRVs), warheads that can hit different targets. The hardliners saw the Soviet throw weight advantage as a way for the USSR to add more and more MIRVs to their missiles as a matter of policy or as an option in the middle of a crisis. They therefore wanted to limit the overall throw weight of both sides to prevent the Soviets from adding warheads. Launcher or warhead limits of the type favored by the pragmatists would still leave the USSR with the capability to add warheads to its fleet of large missiles if it desired, even if these missiles were not normally deployed with the additional warheads.

(14.) A major debate about the merits of rational choice models versus more traditional models that include historical, cultural, leadership, and decision-making factors is under way in the field of international relations (see Bueno de Mesquita and Morrow 1999; Martin 1999; Niou and Ordeshook 1999; Powell 1999; Zagare 1999; Walt, 1999a, 1999b).

(15.) Presidential character might be a fourth, but that is beyond the scope of this article.

References

Allison, Graham. 1969. Conceptual models and the Cuban missile crisis. American Political Science Review 63 (3): 698-718.

--. 1971. Essence of decision. Boston: Little, Brown.

Allison, Graham, and Morton Halperin. 1972. Bureaucratic politics: A paradigm and some policy implications. World Politics 24 (Suppl.): 40-79.

Allison, Graham, and Philip Zelikow. 1999. Essence of decision. Rev. ed. New York: Addison-Wesley.

Anderson, Martin. 1990. Revolution. Stanford, CA: Hoover Institution Press.

Art, Robert. 1968. The TFX decision: McNamara and the military. Boston: Little, Brown.

--. 1973. Bureaucratic politics and American foreign policy: A critique. Policy Sciences 4 (4): 467-90.

Baker, James. 1995. The politics of diplomacy. New York: Putnam.

Baldwin, David, ed. 1993. Neorealism and neoliberalism. New York: Columbia University Press.

Ball, Desmond. 1974. The blind men and the elephant: A critique of bureaucratic politics theory. Australian Outlook 28 (1): 71-92.

Barnett, Lawrence I. 1983. Gambling with history. Garden City, NY: Doubleday.

Bender, Jonathan, and Thomas Hammond. 1992. Rethinking Allison's models. American Political Science Review 86 (2): 301-22.

Beschloss, Michael, and Strobe Talbott. 1993. At the highest levels. Boston: Little, Brown.

Bonafede, Dom. 1977. Brzezinski--Stepping out of his backstage role. National Journal 9 (42): 1596-1601.

Brzezinski, Zbigniew. 1983. Power and principle. New York: Farrar, Straus, Giroux.

--. 1987-1988. The NSC's midlife crisis. Foreign Policy, Winter, 81-95.

Bueno de Mesquita, Bruce, and James D. Morrow. 1999. Sorting through the wealth of notions. International Security 24 (2): 56-73.

Burke, John P. 1992. The institutional presidency. Baltimore: Johns Hopkins University Press.

Burke, John P., and Fred Greenstein. 1989. How presidents test reality. New York: Russell Sage.

Burns, Tom, and G. M. Stalker. 1961. The management of innovation. London: Tavistock.

Bush, George. 1987. Looking forward. New York: Doubleday.

--. 1993. Address before a joint session of Congress on the State of the Union, January 28, 1992. In Public papers of the presidents of the United States, book 1. Washington, DC: Government Printing Office.

Bush Administration. 1995a. NSD-1: Reorganization of the National Security Council. In National security directives of the Reagan and Bush administrations, edited by C. Simpson, 892-93, 903-5. Boulder, CO: Westview.

--. 1995b. N SD-10: List of additional policy coordinating committees. In National security directives of the Reagan and Bush administrations, edited by C. Simpson, 906. Boulder, CO: Westview.

Bush, George, and Brent Scowcroft. 1998. A world transformed. New York: Knopf.

Caldwell, Dan. 1991. The dynamics of domestic politics and arms control. Columbia: University of South Carolina Press.

Campbell, Colin. 1986. Managing the presidency. Pittsburgh, PA: University of Pittsburgh Press.

Cannon, Lou. 1991. President Reagan: The role of a lifetime. New York: Touchstone.

Carter, Jimmy. 1975. Why not the best? Nashville, TN: Broadman.

--. 1982. Keeping faith. New York: Bantam.

Carter Administration. 1977. Presidential directive/NSC-2 (PD-2), National Security Council System.

Chandler, Alfred. 1962. Strategy and structure. Cambridge, MA: MIT Press.

Child, John. 1972. Organizational structure, environment, and performance: The role of strategic choice. Sociology 6 (1): 1-22.

Christensen, Thomas J. 1996. Useful adversaries. Princeton, NJ: Princeton University Press.

Cole, Paul M., and William J. Taylor, Jr. 1983. The nuclear freeze debate: Arms control issues for the 1980s. Boulder, CO: Westview.

David, Charles-Phillipe. 1996. Who was the real George Bush? Foreign policy decision making under the Bush administration. Diplomacy and Statecraft 7 (1): 197-220.

Davis, Vincent. 1979. The president and the national security apparatus. In Defense policy and the presidency: Carter's first years, edited by Sam Sarkesian. Boulder, CO: Westview.

Destler, I. M. 1972a. Comment: Multiple advocacy: Some limits and costs. American Political Science Review 66 (3): 786-95.

--. 1972b. President's bureaucrats and foreign policy. Princeton, NJ: Princeton University Press.

--. 1977. National security advice to US presidents: Some lessons from thirty years. World Politics 29 (2): 143-76.

--. 1980. A job that doesn't work. Foreign Policy 38:8048.

--. 1983. The evolution of Reagan foreign policy. In The Reagan presidency: An early assessment, edited by Fred Greenstein, 117-58. Baltimore: Johns Hopkins University Press.

Draper, Theodore. 1991. A very thin line. New York: Simon & Schuster.

Drew, Elizabeth. 1978. A reporter at large: Brzezinski. New Yorker 54 (11): 90, 95-118, 121-30.

Dumbrell, John. 1993. The Carter presidency. Manchester, UK: Manchester University Press.

Elman, Colin. 1996. Horses for courses: Why not neorealist theories of foreign policy? Security Studies 6 (1): 7-53.

Endicott, John. 1984. The National Security Council: Formalized coordination and policy planning. In National Security Policy, edited by Robert Pfaltzgraff and Uri Ra'anan, 177-200. Hamden, CT: Archon.

Fenno, Richard. 1959. The president's cabinet. Cambridge, MA: Harvard University Press.

Foot, Rosemary. 1997. The practice of power. New York: Oxford University Press.

Foreign policy by committee: Can it really work? 1977. US News and World Report, February 21, p. 27.

Garthoff, Raymond. 1985. Detente and confrontation. Washington, DC: Brookings Institution.

--. 1994. The great transition. Washington, DC: Brookings Institution.

Gates, Robert. 1996. From the shadows. New York: Touchstone.

Gelb, Leslie. 1982. US called ready to propose limits on nuclear arms. New York Times, May 8, p. 1.

--. 1989. Who makes foreign policy? New York Times, February 3, p. A30.

George, Alexander. 1971. The case for multiple advocacy in making foreign policy. Paper presented at the annual meeting of the American Political Science Association, September, Chicago.

--. 1972a. The case for multiple advocacy in making foreign policy. American Political Science Review 66 (3): 751-85.

--. 1972b. Rejoinder. American Political Science Review 66 (3): 791-95.

--. 1980. Presidential decision making in foreign policy. Boulder, CO: Westview.

Geva, Nehemia, and Alex Mintz, eds. 1997. Decision making in war and peace: The cognitive-rational debate. Boulder, CO: Lynne Rienner.

Gordon, Michael. 1983. Administration debates arms cuts with Congress as well as the Soviets. National Journal 15 (32): 1624-29.

--. 1989a. State department weighs plan for arms control focus. New York Times, February 9, p. A13.

--. 1989b. US is considering a tactical shift on strategic arms. New York Times, April 16, p. 1.

--. 1989c. Bush rejects pressure to soften antimissile policy. New York Times, June 9, p. A3.

--. 1989d. Pentagon offers plans to help smooth arms talks. New York Times, June 16, p. A6.

--. 1989e. US places priority on cuts in troops, not missiles. New York Times, June 19, p. A11.

Gottemoeller, Rose. 1978. Evolution of the US organizational setup for dealing with SALT. Santa Monica, CA: RAND.

Hall, David K. 1975. The "custodian-manager" of the policy making process. In Commission on the Organization of the Government for the Conduct of Foreign Policy (Murphy Commission), vol. 2, 100-118. Washington, DC: Government Printing Office.

Halperin, Morton. 1974. Bureaucratic politics and foreign policy. Washington, DC: Brookings Institution.

Halperin, Morton, and David Halperin. 1983-1984. The Key West key. Foreign Policy 52:114-28.

Halperin, Morton, and Arnold Kanter. 1973. The bureaucratic perspective. In Readings in American foreign policy, edited by M. Halperin and A. Kanter, 1-42. Boston: Little, Brown.

Hammond, Paul. 1961. Organizing for defense. Princeton, NJ: Princeton University Press.

Hammond, Thomas. 1986. Agenda control, organizational structure, and bureaucratic politics. American Journal of Political Science 30 (2): 379-420.

Hart, John. 1995. The presidential branch. 2d ed. Chatham, NJ: Chatham House.

Heclo, Hugh. 1991. The changing presidential office. In The managerial presidency, edited by James P. Pfiffner. Pacific Grove, CA: Brooks/Cole.

Hermann, Margaret G., and Thomas Preston. 1994. Presidents, advisers, and foreign policy: The effects of leadership style on executive arrangements. Political Psychology 15 (1): 75-96.

Hess, Stephen. 1988. Organizing the presidency. 2d ed. Washington, DC: Brookings Institution.

Hill, Dilys M., and Phil Williams. 1990. The Reagan presidency: Style and substance. In The Reagan presidency, edited by Dilys M. Hill, Raymond Moore, and Phil Williams. London: Macmillan.

Hilsman, Roger. 1959. The foreign policy consensus: An interim research report. Journal of Conflict Resolution 3 (4): 361-82.

Hoffman, David. 1989. President scales back National Security Council. Washington Post, February 3, p. A8.

Hult, Karen M., and Charles Walcott. 1990. Governing public institutions. Pacific Grove, CA: Brooks/Cole.

Huntington, Samuel. 1961. The common defense. New York: Columbia University Press.

Hutchings, Robert L. 1997. American diplomacy and the end of the cold war. Baltimore: Johns Hopkins University Press.

Hyland, William. 1987. Mortal rivals. New York: Random House.

Johnson, Richard T. 1974. Managing the White House. New York: Harper & Row.

Kaiser, Robert, and Murray Marder. 1977. In pursuit of a SALT II agreement. Washington Post, 11 April, A1.

Katzenstein, Peter J. 1996. Cultural norms and national security. Ithaca, NY: Cornell University Press.

Kegley, Charles. 1989. The Bush administration and the future of American foreign policy: Pragmatism or procrastination. Presidential Studies Quarterly 19 (4): 717-32.

Keohane, Robert, ed. 1986. Neorealism and its critics. New York: Columbia University Press.

Kingdon, John. 1984. Agendas, alternatives, and public policies. New York: HarperCollins.

Kirschten, Dick. 1982. Clark emerges as a tough manager, not a rival to the secretary of state. National Journal 14 (29): 1244-48.

--. 1983. Inner circle speaks with many voices, but maybe that's how Reagan wants it. National Journal 15 (22): 1100-103.

Kissinger, Henry A. 1979. White House years. Boston: Little, Brown.

Kittle, Robert. 1982. New force in Reagan's foreign policy. US News and World Report 92 (14): 30.

Kohl, Wilfred. 1975. The Nixon-Kissinger foreign policy system and US-European relations: Patterns of policy making. World Politics 28 (1): 1-43.

Korb, Lawrence. 1979. National security organization and process in the Carter administration. In Defense policy and the presidency: Carter's first years, edited by Sam Sarkesian. Boulder, CO: Westview.

Kozak, David, and James Keagle, eds. 1988. Bureaucratic politics and national security. Boulder, CO: Lynne Rienner.

Krasner, Stephen. 1972. Are bureaucracies important (or Allison Wonderland). Foreign Policy 7:159-79.

Lake, David, and Robert Powell, eds. 1999. Strategic choice and international relations. Princeton, NJ: Princeton University Press.

Lawrence, Paul, and Jay Lorsch. 1967. Organizations and environment. Cambridge, MA: Harvard University Press.

Light, Paul C. 1982. The president's agenda. Baltimore: Johns Hopkins University Press.

Mandelbaum, Michael. 1990-1991. The Bush foreign policy. Foreign Affairs 70 (1): 5-22.

Martin, Lisa L. 1999. The contributions of rational choice. International Security 24 (2): 74-83.

Melbourne, Roy. 1983. The odyssey of the NSC. Strategic Review 11 (3): 51-64.

Meyer, David S. 1990. A winter of discontent. New York: Praeger.

Miller, Judith. 1982. Arms pact revival is being promoted. New York Times, May 4, p. A9.

Moens, Alexander. 1990. Foreign policy under Carter. Boulder, CO: Westview.

Mulcahy, Kevin V. 1986. The secretary of state and the national security adviser: Foreign policy making in the Carter and Reagan administrations. Presidential Studies Quarterly 16 (2): 280-99.

Neustadt, Richard. 1990. Presidential power and the modern presidents. New York: Free Press.

Newmann, William. 1999. The pattern of foreign policy decision making: Developing an evolutionary model. Ph.D. diss., Pittsburgh, PA: Graduate School of Public and International Affairs, University of Pittsburgh.

Niou, Emerson M. S., and Peter C. Ordeshook. 1999. Return of the Luddites. International Security 24 (2): 84-96.

Oberdorfer, Don. 1989. Bush team plans early review of military, arms talks policy. Washington Post, January 17, p. A20.

--. 1998. From the cold war to a new era. Rev. ed. Baltimore: Johns Hopkins University Press.

Pearlmutter, Amos. 1974. The presidential political center and foreign policy. World Politics 27 (1): 87-106.

Peele, Gillian. 1992. The constrained presidency of George Bush. Current History 91 (564): 151-5.

Perle, Richard. 1987. Arms control. In Staying the course: Henry M. Jackson and national security, edited by Dorothy Fosdick. Seattle: University of Washington Press.

Porter, Roger. 1980. Presidential decision making: The economic policy board. Cambridge, MA: Cambridge University Press.

Powell, Colin. 1995. My American journey. New York: Random House.

Powell, Robert. 1999. The modeling enterprise and security studies. International Security 24 (2): 97-106.

Prados, John. 1991. The keepers of the keys. New York: Morrow.

Reagan, Ronald. 1982. Address at commencement exercises at Eureka College in Illinois, May 9, 1982. In Public papers of the presidents of the United States, vol. 1, 585. Washington, DC: Office of the Federal Register, National Archives and Records Service.

Reagan Administration. 1995a. National security decision directive 1, "National Security Council directives." In National security directives of the Reagan and Bush administrations, edited by C. Simpson, 19. Boulder, CO: Westview.

--. 1995b. National security decision directive 2, "National Security Council structure." In National security directives of the Reagan and Bush administrations, edited by C. Simpson, 20-28. Boulder, CO: Westview.

--. 1995c. National security decision directive 33, "US approach to START negotiations." In National security directives of the Reagan and Bush administrations, edited by C. Simpson, 118-20. Boulder, CO: Westview.

Reedy, George. 1970. The twilight of the presidency. New York: New American Library.

Rochan, Thomas IL, and David S. Meyer, eds. 1997. Coalitions and political movements. Boulder, CO: Lynne Rienner.

Rockman, Bert. 1988. The style and organization of the Reagan presidency. In The Reagan legacy, edited by Charles O. Jones. Chatham, NJ: Chatham House.

--. 1991. The leadership style of George Bush. In The Bush presidency: First appraisals, edited by Colin Campbell and Bert Rockman. Chatham, NJ: Chatham House.

Rosati, Jerel. 1987. The Carter administration's quest for global community. Columbia: University of South Carolina Press.

Rosenthal, Andrew. 1991. Arms plan germinated in back porch sessions. New York Times, September 29, p. A14.

Rourke, Francis. 1972. Bureaucracy and foreign policy. Baltimore: Johns Hopkins University Press.

Rowny, Edward. 1992. It takes one to tango. Washington, DC: Brassey's.

Schieffer, Bob, and Gary Paul Gates. 1991. The acting president. New York: Dutton.

Schilling, Warner, Paul Hammond, and Glenn Snyder. 1962. Strategy, politics, and defense budgets. New York: Columbia University Press.

Senate Committee on Armed Services. 1985. Defense organization: The need for change. Staff report, United States Senate, October 15.

Senate Committee on Foreign Relations. 1980. The national security adviser: Role and accountability. United States Senate, April 17.

Shoemaker, Christopher. 1989. The National Security Council staff: Structure and functions. Association of the U.S. Army Land Warfare Paper No. 3. Arlington, VA: Institute of Land Warfare.

Simpson, Christopher, ed. 1995. National security directives of the Reagan and Bush administrations. Boulder, CO: Westview.

Smith, Gaddis. 1986. Morality, reason, and power. New York: Hill & Wang.

Smith, Hedrick. 1981. Haig is given broad policy power, but less than he initially thought. New York Times, February 27, p. 1.

Smith, Perry. 1970. The Air Force plans for peace: 1943-1945. Baltimore: Johns Hopkins University Press.

Snow, Donald, and Eugene Brown. 1997. Beyond the water's edge. New York: St. Martin's.

Snyder, Jack L. 1991. Myths of empire: Domestic politics and international ambition. Ithaca, NY: Cornell University Press.

Spencer, Donald. 1988. The Carter implosion. New York: Praeger.

Szanton, Peter. 1980. Two jobs, not one. Foreign Policy 38:89-91.

Szulc, Tad. 1981. The vicar vanquished. Foreign Policy 43:173-86.

Talbott, Strobe. 1979. Endgame. New York: Harper & Row.

--. 1982. Deadly gambits. New York: Knopf.

Thompson, James D. 1967. Organizations in action. New York: McGraw-Hill.

Thompson, Kenneth. 1990. Interview with Jimmy Carter, "contemporary national and international issues." In The Carter presidency, edited by K. Thompson. Lanham, MD: University Press of America.

--. 1993. Interview with Caspar Weinberger, "Reagan and international arms agreements." In Foreign policy in the Reagan administration, edited by K. Thompson. Lanham, MD: University Press of America.

Thompson, Robert J. 1992. Contrasting models of White House staff organization: The Eisenhower, Ford, and Carter experiences. Congress and the Presidency 19 (2): 113-36.

Toth, Robert. 1989. Bush exploring bold cuts in long range arms. Los Angeles Times, June 9, p. 1.

Towell, Pat. 1982. Reagan cites Soviet power: Embraces nuclear reduction and says talks may begin soon. Congressional Quarterly 40 (14): 725-26.

Tower Commission. 1987. Report of the president's Special Review Board (Tower Commission Report). New York: Times Books and Random House.

Vance, Cyrus. 1983. Hard choices. New York: Simon & Schuster.

Walcott, Charles E., and Karen M. Hult. 1995. Governing the White House. Lawrence: University Press of Kansas.

Waller, Douglas C. 1987. Congress and the nuclear freeze. Amherst: University of Massachusetts Press.

Walsh, Edward, and George C. Wilson. 1977. Carter to get study on A-deterrence. Washington Post, January 28, p. A3.

Walsh, Lawrence. 1994. Iran-Contra: The final report. New York: Times Books.

Walt, Stephen. 1999a. Rigor or rigor mortis? Rational choice and security studies. International Security 23 (4): 5-48.

--. 1999b. A model disagreement. International Security 24 (2): 115-30.

Waltz, Kenneth N. 1979. Theory of international politics. New York: Random House.

--. 1996. International politics is not foreign policy. Security Studies 6 (1): 54-57.

Weinraub, Bernard. 1989. Bush backs plans to enhance role of security staff. New York Times, February 2, p. A1.

Weisman, Steven. 1982. Clark is staking out his turf as security adviser. New York Times, January 19, p. A22.

Welch, David A. 1992. The organizational process and bureaucratic politics paradigms: Retrospect and prospect. International Security 17 (2): 112-46.

Winik, Jay. 1996. On the brink. New York: Simon & Schuster.

Wohlstetter, Roberta. 1962. Pearl Harbor: Warning and decision. Stanford, CA: Stanford University Press,

Woodward, Bob. 1991. The commanders. New York: Simon & Schuster.

Woodward, Joan. 1958. Management and technology. London: Her Majesty's Stationary Office.

Yang, John E. 1991. Bush plan emerged after failed coup. Washington Post, September 28, p. A23.

Zagare, Frank C. 1999. All mortis, no rigor. International Security 24 (2): 107-13.

William Newmann is assistant professor of political science and public administration at Virginia Commonwealth University. He has published articles and book chapters regarding the Eisenhower administration's national security decision making, U.S.-Soviet/Russian relations, and the Association of South East Asian Nations.