¹ See, for example, the recommendation that an adversary system be adopted at various echelons within the State Department, in Task Force VII, “Stimulation of Creativity,” Diplomacy for the 70's: A Program of Management Reform for the Department of State. Department of State Publication 8551, series 143. December 1970. U.S. Government Printing Office, Washington, D.C.

A laboratory exploration of the possible advantages for administrators of an adversary proceeding between technical experts is reported by Fairley, William and Mosteller, Frederick, “Trial of an Adversary Hearing: Public Policy in Weather Modification” (draft), Harvard University, 05 10, 1971Google Scholar.

² For a discussion and evaluation of the judicial model from the standpoint of decision-making theory, see Mason, Richard O., Dialectics in Decision-Making: A Study in the Use of Counterplanning and Structured Debate in Management Information Systems, Ph.D. dissertation, Business Administration, University of California, Berkeley, 1968, pp. 81–4Google Scholar.

³ See, for example, de Rivera, Joseph H., The Psychological Dimension of Foreign Policy (Columbus, Ohio: Charles E. Merrill, 1968), pp. 61–4, 209–11Google Scholar; also, Jervis, Robert, “Hypotheses on Misperception,” World Politics, 20 (04, 1968), 464–5CrossRefGoogle Scholar; Kennedy, R. F., Thirteen Days (New York: Norton, 1969), p. 112Google Scholar; Holsti, Ole R., Crisis, Escalation, War (Montreal and London: McGill-Queen's University Press, 1972)CrossRefGoogle Scholar.

⁴ James C. Thomson, Jr. provides an insightful and sobering account of the “domestication” of devil's advocates in Vietnam policy making. (“How Could Vietnam Happen? An Autopsy,” Atlantic Monthly [04, 1968] pp. 47–53.Google Scholar) The efficacy of a devil's advocate is firmly discounted also by George Reedy, the one-time press secretary to PresidentJohnson, , in The Twilight of the Presidency (N.Y.: World, 1970), p. 11Google Scholar.

I have dealt more fully with the uses and limitations of devil's advocates in U.S. foreign policy making in “Stress Adaptation in Political Decision-making: The Individual, Small Group, and Organizational Contexts,” in Coping and Adaptation, ed. Coelho, George V., Hamburg, David A., and Adams, JohnGoogle Scholar (forthcoming).

⁵ Even less can be said about the merits and drawbacks of different foreign policy-making models than about alternative prescriptive models of industrial management, which continue to be the subject of controversy among social psychologists despite considerable research efforts to clarify their effectiveness. For an incisive analytical summary of this controversy, useful in several respects for the present discussion, see Vroom, Victor H., “Industrial Social Psychology,” in Handbook of Social Psychology, ed. Lindzey, Gardner and Aronson, Elliot, 2nd edition, Vol. 5 (Reading, Mass: Addison-Wesley Publishing Company, 1970) esp. 227–240Google Scholar.

⁶ For detailed description of President Nixon's reorganization of the NSC see, for example, Davis, Vincent, “American Military Policy: Decision-making in the Executive Branch,” Naval War College Review (05 1970), pp. 4–23Google Scholar; Johnson, Robert H., “The National Security Council: The Relevance of Its Past To Its Future,” Orbis, 13 (Fall, 1969), 709–735Google Scholar; Kolodziej, E. A., “The National Security Council: Innovations and Implications,” Public Administration Review, 29 (11.–12. 1969), 573–585CrossRefGoogle Scholar; Thayer, Frederick C., “Presidential Policy Processes and ‘New Administration’: A Search for Revised Paradigms,” Public Administration Review, 31 (09/10 1971) 552–561CrossRefGoogle Scholar; and particularly Destler, Irving M., Presidents, Bureaucrats and Foreign Policy (Princeton, N.J.: Princeton University Press, 1972)Google Scholar.

For accounts of the historical evolution of the NSC see, for example, Clark, Keith C. and Legere, Lawrence J., eds., The President and the Management of National Security (N.Y.: Praeger, 1969)Google Scholar; Hammond, Paul Y., Organizing for Defense (Princeton: Princeton University Press, 1961)CrossRefGoogle Scholar; SenatorJackson, Henry M., ed., “Jackson Subcommittee Papers on Policy-Making at the Presidential Level,” The National Security Council (N.Y.: Praeger, 1965)Google Scholar; Falk, Stanley, “National Security Council under Truman, Eisenhower, and Kennedy,” Political Science Quarterly, 79 (09., 1964), 403–434CrossRefGoogle Scholar.

⁷ As, for example, in the policy planning for the SALT talks. See Orr, Samuel C., “Defense Report/National Security Council Network Gives White House Tight Rein over SALT Strategy,” National Journal, 3 (04 24, 1971), 877–886Google Scholar; and Leacocos, John P., “Kissinger's Apparat,” Foreign Policy, No. 5 (Winter 1971–1972), p. 19Google Scholar.

⁸ Leacocos, p. 5.

⁹ Cyert, Richard M. and March, James G., Behavioral Theory of the Firm (Englewood Cliffs, N.J.: Prentice-Hall, 1963), pp. 116–120Google Scholar. As Axelrod, Robert (Conflict of Interest [Chicago: Markham, 1970, p. 123])Google Scholar points out, Cyert's and March's observation regarding internal bargaining and other adaptive devices on the part of the subunits of an organization “leaves open the question of how the upper levels of the organization supervise the sub-units. …” This is not surprising, however, since the Cyert-March theory is not a prescriptive management theory.

¹⁰ This criticism of the State Department was tacitly accepted by its Task Force VII, “Stimulation of Creativity,” in noting that “the lack of a system for subjecting policy to the challenge of an adversary view” within the Department was “a major weakness in the Department's organization ….” (Diplomacy for the 70's, p. 294).

¹¹ Nixon, Richard, U.S. Foreign Policy for the 1970's: A New Strategy for Peace, A report to the Congress, 02 18, 1970 (Washington, D.C.), p. 22Google Scholar. Henry Kissinger had earlier noted and deplored the play of bureaucratic politics in policy making in “Domestic Structure and Foreign Policy,” DAEDALUS, 95 (Spring, 1966), 503–529Google Scholar.

¹² Alain C. Enthoven and K. Wayne Smith take note of the fact that the NSC in the Nixon Administration has been made “the focal point for reviewing alternatives and formulating broad national security policy.” They add that the NSC staff under Kissinger “appears to be trying to do for the President at the national level what the Systems Analysis office did for the Secretary of Defense.” (How Much Is Enough? Shaping the Defense Program, 1961–1969 [N.Y.: Harper & Row, 1971], p. 332Google Scholar).

¹³ Destler credits the NSSM device with a particularly important contribution to Nixon's decision to renounce use of biological weapons. “Only the Army really wanted them, and, within the Army, mainly the branch specializing in them. But … the issue never got to the President for his review because neither the Secretary of Defense nor the Secretary of State was willing to pay the bureaucratic political price of recommending a change” (Destler, Presidents, Bureaucrats, and Foreign Policy: The Politics of Organizational Reform, Ch. 5).

¹⁴ This is evidently the case, for example, with regard to the Arms Control and Disarmament Agency. See Orr, “Defense Report” and Leacocos, , “Kissinger's Apparat,” p. 19Google Scholar.

¹⁵ This is implied in several accounts of the Cambodian decision. See Smith, Hedrick, “Cambodian Decision: Why President Acted,” N.Y. Times, 06 30, 1970Google Scholar; Maxey, David, “How Nixon Decided to Invade Cambodia,” Look, 08 11, 1970Google Scholar.

¹⁶ For a general discussion of the desirability of keeping specialists carrying out search, analysis, and choice close together within an organization, and also of the conditions under which it is “rational” to separate them, see Downs, Anthony, Inside Bureaucracy, (Boston: Little, Brown & Co., 1967), pp. 179–188CrossRefGoogle Scholar. (Downs uses the term “evaluation” for what is here referred to as “choice.”)

¹⁷ This point is also emphasized from a somewhat different standpoint by Destler in Presidents, Bureaucrats, and Foreign Policy.

¹⁸ Destler, Ch. 5.

¹⁹ Quoted by Viorst, Milton, “William Rogers Thinks Like Richard Nixon,” New York Times, Magazine, 02 27, 1972Google Scholar. On the NSC's Washington Special Actions Group, see also Leacocos, “Kissinger's Apparat,” pp. 7–8.

²⁰ Destler believes the failure to do so is one of the major deficiencies of the Nixon Administration's organization of foreign policy making. The creation of the Planning and Coordination Staff in the State Department in 1969 has evidently not been as successful in this respect as had been hoped. In mid 1971 a new system of evaluating country programs (Policy Analysis Resource Allocation) was installed within State having as its goal to give State more weight in the implementation of foreign policy (Leacocos, , “Kissinger's Apparat,” p. 10Google Scholar).

²¹ Orr, , “Defense Reports,” p. 881Google Scholar.

²² Leacocos, p. 23.

²³ Johnson, Robert H., “National Security Council,” p. 720Google Scholar.

²⁴ See the discussion by Joseph L. Bower, who argues for a synthesis of normative and descriptive theory on the grounds that “both normative and descriptive content are prerequisites for prescription” (“Descriptive Decision Theory from the ‘Administrative’ Viewpoint,” in Bauer, Raymond A. and Gergen, Kenneth J., eds., The Study of Policy Formation [N.Y.: Free Press, 1968])Google Scholar.

²⁵ There is no reference to decision making, for example, in Coser's, Lewis A.The Functions of Social Conflict (Glencoe: Free Press, 1956)Google Scholar. Similarly, there is little direct interest in the quality of organizational decision making in William A. Gamson's otherwise stimulating analysis of the ways in which authorities can control discontented partisans seeking change (Power and Discontent, Homewood, Ill.: Dorsey Press, 1968Google Scholar).

²⁶ A number of such studies have been conducted by N. R. F. Maier and his associates over a period of many years; they are reproduced and collated in Maier's, Problem-solving and Creativity in Individuals and Groups (Belmont, Calif.: Brooks/Coles, 1970)Google Scholar. See also Hoffman, R. L., “Conditions for Creative Problem Solving,” Journal of Psychology, 52 (10., 1961) 429–444CrossRefGoogle Scholar; Vroom, Victor H., Grant, Lester D., and Cotton, Timothy S., “The Consequences of Social Interaction in Group Problem Solving,” Organizational Behavior and Human Performance, 4 (02, 1969), 77–95CrossRefGoogle Scholar; Vroom, “Industrial Social Psychology,” and Mason, Dialectics in Decision-Making.

While the simplified structure of these laboratory experiments precludes their findings from being directly transferable to governmental decision making, relevant structural and leadership variables have been studied, and some of the emergent propositions are consistent with impressionistic observations of critical variables that affect the performance of political decision-making groups in real life.

²⁷ As usually employed by organizational theorists such as Herbert Simon, James March, Richard Cyert, and others, “search” refers to the processes of obtaining and sharing relevant information, and of identifying and inventing alternative options; “analysis” (or “evaluation”) refers to the processes of examining relationships among available information and evaluating the relative appropriateness of alternative options with reference to stated objectives and values; and “choice” refers to the procedures and decision rules followed in choosing from among the alternative options.

²⁸ Bower, Joseph L., “The Role of Conflict in Economic Decision-Making Groups: Some Empirical Results,” Quarterly Journal of Economics, 79 (05, 1965), 263–277CrossRefGoogle Scholar. For this study Bower simulated 58 capital budgeting committees, each of which was staffed by three Harvard MBA candidates. The groups were divided into “rational teams” and “foundations,” the main difference between them being that all individuals in the first type of group had identical utility functions whereas members of the second type of group did not. One of the limitations of Bower's research study and of his findings stems from the fact that he employed and varied only two decision rules—unanimity and majority rule. Thus, Bower did not study the effect of a decision rule that is especially pertinent for government decision making, namely when the leader of the group (chief executive) makes the decision after group discussion, debate, bargaining, etc. Bower recognized that severe conflict within a group can have a debilitating effect on the probability of its reaching an agreement, but both of his two types of groups were evidently leaderless groups. The possibility suggests itself that groups with informal or authoritative leaders, operating with the third decision rule mentioned here, might succeed in reducing the debilitating effect of intra-group conflict on the probability of reaching agreement without, it should be added, also reducing the value of intra-group conflict for its performance of the problem-solving task.

²⁹ Graham T. Allison has brilliantly codified and explicated much of the previous literature on bureaucratic politics by writers such as Lindblom, Neustadt, Schilling, Hammond, Huntington, Hilsman. The fullest and most recent report of Allison's, work is his Essence of Decision: Explaining the Cuban Missile Crisis (Boston: Little, Brown & Co., 1971)Google Scholar. (An earlier summary of his Ph.D. dissertation was published in APSR, September, 1969, under the title, “Conceptual Models and the Cuban Missile Crisis.”)

Important new insights into bureaucratic aspects of foreign policy making are being developed in a current study by Halperin, Morton H., Bureaucratic Politics and Foreign Policy, (Washington, D.C.: Brookings Institution, 03 1970 [draft manuscript])Google Scholar.

A number of organizational theorists have discussed the phenomenon of bureaucratic or administrative “politics” and ways in which internal conflicts over policy may be resolved or regulated. For a brief summary see Feldman, Julian and Kanter, Herschel E., “Organizational Decision-Making,” in Handbook of Organizations, ed. March, James G. (Chicago: Rand McNally, 1965, pp. 638–641.)Google Scholar

Recent case studies of bureaucratic politics include Steinbruner, John D., “The Mind and Milieu of Policy-makers: A Case Study of the MLF,” Ph.D. dissertation, Department of Political Science, Massachusetts Institute of Technology, 02 1968Google Scholar; Neustadt, Richard, Alliance Politics, (N.Y.: Columbia University Press, 1970)Google Scholar; Halperin, Morton H., “The Decision to Deploy the ABM: Bureaucratic Politics in the Pentagon and White House in the Johnson Administration,” paper presented at APSA meetings, 09., 1970Google Scholar; Sigal, Leon V., “The ‘Rational Policy’ Model and the Formosa Straits Crises,” International Studies Quarterly, 14, (06, 1970), 121–156CrossRefGoogle Scholar; and Williamson, Samuel R. Jr. “‘Stumbling’ Toward War: Organizational Structure and the Austro-Hungarian Experience, 1912–1914,” paper delivered at the APSA annual meetings, 09, 1971Google Scholar. Earlier case studies of “administrative” or “bureaucratic” politics are listed by Allison in his bibliography.

³⁰ Although most research on bureaucratic politics has confined itself to descriptive case studies and explanatory theory, often implying that little could be done to eliminate or control its worst excesses, some specialists have recently turned to a cautious exploration of ways in which the chief executive might master it. See Halperin, Morton H., “Why Bureaucrats Play Games,” Foreign Policy (05, 1971), pp. 70–90Google Scholar, and particularly his recent proposal for restructuring the defense policy-making system to improve the quality of advocacy and the ability of the President to benefit from it (“The President and the Military,” Foreign Affairs, 50 [01, 1972], 310–324CrossRefGoogle Scholar). See also Graham T. Allison and Morton H. Halperin, “Bureaucratic Politics: A Paradigm and Some Policy Implications,” World Politics (forthcoming).

³¹ See, for example, Rowen, Henry S., “Bargaining and Analysis in Government,” in The Administrative Process and Democratic Theory, ed. Gawthrop, Louis C. (Boston: Houghton Mifflin Co., 1970), pp. 31–37Google Scholar. Similarly, William Capron offers the insightful observation that centralized management tools such as program budgeting systems, if correctly applied, can yield a more productive and effective decentralized decision-making structure (The Impact of Analysis on Bargaining in Government,” reprinted in Gawthrop, , 354–371Google Scholar).

³² To these three conditions Hilsman appended an additional proviso: “that the circumstances make no demand for speed,” and he suggested the usefulness of distinguishing among “crisis policy,” “program policy,” and “anticipatory policy.” Hilsman, Roger, “The Foreign Policy Consensus: An Interim Research Report.” Journal of Conflict Resolution, 3 (12, 1959), 361–382CrossRefGoogle Scholar. Hilsman's more recent views on improving the policy-making machinery (in To Move a Nation, esp. pp. 564–576, and The Politics of Policy Making in Defense and Foreign Affairs [N.Y.: Harper & Row, 1971], Ch. 10Google Scholar) restate and elaborate the ideas he presented in 1959.

³³ This includes a number of different things: access to the President or other senior officials and the ability to use their confidence and trust as a bargaining asset; responsibility for implementation of policies decided upon, which amplifies one's voice in policy making; the ability to go outside the executive branch itself to secure powerful allies in Congress, among specialized foreign policy elites, in the media, etc.

³⁴ Time-compressed, stress-producing international crises are particularly likely to strain the workings of multiple advocacy even while making such advocacy more important than ever for obtaining a balanced, multisided examination of options. I shall discuss this further when I take up the various types of malfunctions the foreign policy-making process is prone to suffer. More generally, a substantial literature has accumulated on the effects of crisis on the rationality of organizational decision making. A particularly useful and current review of laboratory and field studies of the effects of stress is provided by Ole R. Holsti, Crisis, Escalation, War.

³⁵ It should be noted that the balancing of actor resources is relevant also in forming ad hoc policy planning groups and interdepartmental task forces. The staffing of these ad hoc groups should include not merely representatives of the different agencies but, in order to assure effective multiple advocacy, individuals with adequate weight, competence, bargaining and persuasive skills. Sensitivity to these considerations is evident also in the staffing of some of the NSC-centered interdepartmental committees in the Nixon Administration. (See, for example, Orr, , “Defense Report,” p. 881.Google Scholar)

³⁶ As these observations imply, the full complexity and variability of the relationship between analysis and bargaining is not adequately captured in organizational theory. The fundamental proposition advanced by March, James G. and Simon, Herbert A. (Organizations, [N.Y.: Wiley, 1965], p. 131)Google Scholar, i.e., that rational, analytic decision-making processes take precedence over bargaining only when the actors in the system share the same operational goals, may be generally valid; but it leaves open the important question of how bargaining and analysis can best be combined or, at least, reconciled to improve policy performance. As system analysts have suggested, at the very least, analysis can usefully moderate bargaining processes and clarify for each advocate the relative utility of different possible compromises before he accepts one of them. Beyond that, a competent analytical staff reporting directly to the executive can act as an important constraint on the partisan analyses of advocates and on the bargaining among them.

³⁷ For assistance in elaboration of these roles, I am indebted to Flanagan, Scott C., Aborted Democratization: The Inter-War Crisis in Japan, draft Ph.D. dissertation, Political Science Department, Stanford University, 1971Google Scholar.

³⁸ Lindblom's earlier views on the advantages of the bargaining model appear in his “Bargaining: The Hidden Hand in Government,” The RAND Corporation, RM-1436-RC (22 February, 1955). A substantial broadening and fuller exposition of the bargaining model, which Lindblom renamed “partisan mutual adjustment,” appears in The Intelligence of Democracy: Decision-Making Through Mutual Adjustment (N.Y.: Free Press, 1965)Google Scholar. Interesting refinements and additions appear in his most recent statement, The Policy-Making Process (Englewood Cliffs, N.J.: Prentice-Hall, 1968)Google Scholar.

³⁹ This point was made by Charles Lindblom in a personal communication (October 12, 1971); he adds that “in some forms of partisan mutual adjustment, to be specific, the mutual adjustment is achieved by mutually adaptive moves of distant parties who do not in any way communicate with each other and indeed may not even know each other or be able to identify each other.” In any form of mutual adjustment, “the adjustments are made serially by persons or parties that react to preceding moves of other parties and are in turn reacted to by successive moves of further parties—all in circumstances in which no person or party sees the earlier moves of the other parties as anything but part of the environment to which he is reacting and in which it may not be at all conspicuous.” Types of partisan mutual adjustment and the position of negotiation therein are indicated in Lindblom's, Intelligence of Democracy, pp. 33–4Google Scholar.

⁴⁰ For a incisive critique of these and other assumptions of pluralist theory, see Schick, Allen, “Systems Politics and Systems Budgeting,” Public Administration Review, 29 (03–04 1969), 139–150CrossRefGoogle Scholar.

⁴¹ It is to precisely these dangers that proponents of centralized management tools such as systems analysis and program budgeting have called attention in diagnosing the deficiencies and risks of pure bargaining models of policy making. Thus, as Rowen, Henry “Bargaining and Analysis,” p. 35Google Scholar has observed: “There are not only wide differences in the bargaining power of the ‘firms’ (bureaucratic policy actors), this bargaining power is not necessarily highly correlated with the information or the power to take relevant action to accomplish objectives with a high degree of efficiency.”

On occasion Lindblom himself has briefly acknowledged the possible deficiencies and poor performance of the bargaining model. For example, in “Bargaining: The Hidden Hand in Government,” p. 24, he acknowledged that “carried to decision-making processes in which a strong hierarchical element is desirable, bargaining in the wrong place at the wrong time accounts for some of the worst aspects of American government.” He has had little to say, however, about how to avoid bad variants and ensure good variants of the bargaining model. Morerecently, in The Policy-Making Process, pp. 33–4, while continuing to note the limits of policy analysis, Lindblom recognizes the contributions that even “partisan analysis” can make to clarification and improvement of policy.

⁴² I have discussed various orientations to “politics” and political conflict held by earlier presidents and the strategies they employed to cope with bureaucratic politics in the executive branch in “Stress Adaptation in Political Decision-making.”

⁴³ The advice Neustadt's book offers a president by no means exhausts the possibilities available to him. In a cogent explication and critical appreciation of Presidential Power, Peter Sperlich argues that Neustadt overstates the President's need to rely exclusively on bargaining and persuasion to influence others and notes that the chief executive has other sources of authority and influence, such as organizational ideology and norms, at his disposal for this purpose. Sperlich also calls attention to the danger of “overload and breakdown” if a president attempts to follow Neustadt's advice too literally. (Sperlich, , “Bargaining and Overload: An Essay on Presidential Power,” in The Presidency, ed. Wildavsky, Aaron, [Boston: Little, Brown & Co., 1969], pp. 168–192.Google Scholar) On the chief executive's ability to use symbolic manipulation, not merely commands, to influence subordinates and other actors in the political system, see also Edelman, Murray, The Symbolic Uses of Politics (Urbana: University of Illinois Press, 1964)Google Scholar.

⁴⁴ This has inevitably made some political scientists uneasy and led to the view that Presidential Power is too Machievellian. A useful commentary along these lines is provided by Bluhm, William T., Theories of the Political System (Englewood Cliffs, NJ.: Prentice-Hall, 1965)Google Scholar.

President Kennedy himself is reported to have been made uneasy at suggestions that he was modeling his presidency along the lines of Neustadt's prescriptions. See Schlesinger, Arthur M. Jr.A Thousand Days (Boston: Houghton Miffln Co., 1965), pp. 678–9Google Scholar; Sorensen, Theodore C., Kennedy (N.Y.: Harper & Row, 1965), p. 389Google Scholar.

More generally, faith in a strong president and in the virtues of presidential power has been called into question by a number of political scientists in recent years. See Cronin, Thomas E., “The Textbook Presidency and Political Science,” paper, delivered at American Political Science Association meetings, 09, 1970Google Scholar; and Andrews, William G., “The Presidency, Congress, and Constitutional Theory,” paper, presented at American Political Science Association annual meetings, 09, 1971Google Scholar.

⁴⁵ On this point see Allison, Essence of Decision, Chs. 4 and 5.

⁴⁶ In Problem-Solving Discussions and Conferences: Leadership Methods and Skills (San Francisco: McGraw-Hill, 1963)Google Scholar, N. R. F. Maier outlines nine basic principles for leading a problem-solving group in order to enhance the quality of its “solution” or decision. Although employing a terminology that differs somewhat from that employed by organizational theorists, Maier's principles can be translated into the multifunction decision framework of “search”-“evaluation”-“choice,” and many of his principles are applicable to the management of small political decision-making groups. He emphasizes, for example, the value of (a) turning situations that are prematurely structured for “choice” among immediately obvious options into a fuller discussion of the “problem”; (b) preventing the “search” for options to be inhibited by too early or too dominant a preoccupation with “evaluation”; (c) requiring the group to obtain a “second solution” (i.e., an iterative approach to problem solving) to encourage issues to be raised that were previously overlooked or suppressed; (d) encouraging and protecting the expression of minority opinions, and utilizing disagreement to create the conditions for creative problem solving; etc.

In separate seminar papers, two of my students, Mike Garner and Craig Cotora, applied Maier's principles in order to compare and evaluate Kennedy's management of the policy-making process in the Bay of Pigs and Cuban missile crises. The results indicate that Kennedy “violated” most of Maier's principles in making the Bay of Pigs decision while acting in consonance with them in the Cuban missile case. A particularly incisive comparison of Kennedy's handling of top-level policy deliberations in the two cases, with reference to small-group dynamics that affect the performance of advisory groups, is presented in Irving Janis, Victims of Groupthink (forthcoming).

⁴⁷ Although the importance of the “generalist” has often been emphasized by experienced policy makers, the role does not appear to be well defined. Also, it seems to have received little formal attention in studies of group problem solving. As Victor Thompson suggests, one should not restrict one's view of the ideal generalist to a person who is broad-gauged without being, or having been, an expert of some kind. Rather, “generalist” should be thought of as a special kind of relationship—that is, a role that even the expert can play vis-à-vis subjects on which others in the group, not he, are specialists. But for an expert to play the role of generalist he must be willing to violate the tacit agreement, often entered into by experts, not to challenge each other's expertise. The essence of the generalist's role is to ask the questions his very ignorance, naivete and different perspective allow him to ask of experts, questions which might never occur to the experts or whose relevance would not be evident to them until pointed out. (Thompson, , Bureaucracy and Innovation [University, Alabama: University of Alabama Press, 1969], pp. 12–14Google Scholar; see also Specialists and Generalists: A Selection of Readings, U.S. Congress, Senate, Subcommittee on National Security and International Operations of the Committee on Government Operations [Washington, D.C., Government Printing Office, 1968].)Google Scholar For various reasons I question the assumption that the role of “generalist” in problem-solving groups can be played adequately by the political leader or chief decision maker.

⁴⁸ These remarks about the “collegial” style that dominated the deliberations of the ExCom draw on Sorensen, Theodore C., Kennedy, pp. 679–680, 684–686Google Scholar; and Abel, Elie, The Cuban Missile Crisis, pp. 47, 52, 69–73, 79, 86–88Google Scholar. Similar observations are to be found in other sources on the Cuban missile crisis. See also my case study of the Cuban missile crisis in George, A. L., Hall, David K., and Simons, William E., The Limits of Coercive Diplomacy: Laos, Cuba, Vietnam (Boston: Little, Brown & Company, 1971)Google Scholar.

⁴⁹ Less pronounced forms of the collegial model of policy discussions are also cultivated in hierarchical systems. Johnson, Robert H. (“National Security Council,” p. 727)Google Scholar describes efforts to achieve some of the advantages of this approach within the NSC system under Eisenhower.

⁵⁰ A selective bibliography appears in Clark and Legere, President and Management of National Security.

⁵¹ Bennis, Warren G., “Post-Bureaucratic Leadership,” Transaction (07–08 1969), pp. 44–61Google Scholar. See also Bennis, W. G. and Slater, Philip E., The Temporary Society (N.Y.: Harper & Row, 1968)Google Scholar; and Toffler, Alvin, Future Shock (N.Y.: Random House, 1970), Ch. 7Google Scholar, “Organizations: The Coming Adhocracy.”

Charles Perrow's concept of “multiple leadership” is similar in some ways to the notion of policy making via multiple advocacy (Perrow, , “The Analysis of Goals in Complex Organizations,” American Sociological Review, 26 [12, 1961], 854–866)CrossRefGoogle Scholar.

⁵² Bennis, , “Post-Bureaucratic Leadership,” p. 45Google Scholar. Bennis recognizes that one task will be the control of conflict within such organizational forms. It will be necessary, he believes, to develop ways of developing “group synergy” and “collaborative cultures.” The new concept of leadership required will include (what we also stress here) “practical theories of intervening and guiding these systems, theories that encompass methods for seeding, nurturing, and integrating individuals and groups” (p. 51).

⁵³ Wilensky, Harold, Organizational Intelligence (N.Y.: Basic Books, 1967)Google Scholar. For a more general discussion of six types of “failures or pathologies” to which every political decision system is prone, see Deutsch, Karl W., The Nerves of Government (N.Y.: Free Press, 1963), Ch. 13Google Scholar, “The Self-Closure of Political Systems.” Breakdowns in presidential advisory systems are discussed and illustrated by Neustadt, Presidential Power, and by Wildavsky, Aaron, Dixon-Yates: A Study in Power Politics, (New Haven: Yale University Press, 1962)Google Scholar.

⁵⁴ Thus, for example, Wilensky, pp. 174–181, lists a number of “organizational defenses against informational pathologies,” and Anthony Downs identifies strategies (such as redundancy, use of counter-biases, eliminating the middlemen, developing distortion-proof messages) for reducing or avoiding distortion in hierarchical communication within the organization. (Inside Bureaucracy [Boston: Little, Brown, 1967], pp. 118–127Google Scholar). See also Bobrow, Davis B., “International Indicators,” paper presented at annual meeting of the American Political Science Association, 09, 1969Google Scholar; and Webb, Eugene J., “Individual and Organizational Forces Influencing the Interpretation of Indicators,” (unpublished ms.) Stanford University, 04, 1969Google Scholar.

⁵⁵ von Neumann, John, “The General and Logical Theory of Automata,” in The World of Mathematics, ed. Newman, James R. (N.Y.: Simon and Schuster, 1956). Vol. IV, 2085–2086Google Scholar. (Quoted by Landau, Martin, in “Redundancy, Rationality, and the Problem of Duplication and Overlap,” Public Administration Review, 29 [07–08, 1969], 346–358.)CrossRefGoogle Scholar Stimulated by mathematical demonstrations that redundancy can help create a system that is more reliable than any of its parts, Landau argues that duplication and overlap in administrative agencies are not necessarily signs of waste and inefficiency, though he admits the task remains to distinguish between inefficient redundancies and those that are constructive and reinforcing.

⁵⁶ The empirical base for this part of the multiple advocacy theory is, therefore, admittedly confined to a special (though very important) class of foreign policy decisions. A broader empirical base would be desirable in order to define more clearly the scope of this prescriptive theory and to identify the types of policy problems and conditions under which it is most applicable.

⁵⁷ It should be clear that the study is concerned with possible malfunctions of the policy-making process, and not other kinds of “errors” or “flaws,” such as psychological aberrations, man-machine errors, etc. To this end, I have tried to confine my definition of “malfunctions” to instances in which the policy-making process failed to replicate one or another component of the rational model of decision making, i.e., did not succeed in reproducing fully within the organizational process the functional equivalent of the formal rational model.

Neither does the present study attempt to present in any systematic way an explanation for a particular malfunction as does, for example, Irving Janis's theory of “groupthink,” which he applies to some of the same historical cases.

⁵⁸ Jeffrey Race has examined a number of psychological and organizational processes that seem to lead American policy makers to systematically suppress certain ways of thinking about revolutionary movements. (In “American Intervention Abroad: Systematic Distortions in the Policy-Making Process,” paper given at American Political Science Association annual meetings, 09, 1970.Google Scholar)

⁵⁹ Relevant here is the familiar distinction in political science between “decisional output” and “policy outcome.” The latter is usually determined not merely by the quality of the policy decision made earlier (i.e., the “decisional output”) but by other factors and subsequent events, some of which could probably not have been anticipated by policy makers in reaching their decision.

⁶⁰ A useful discussion is provided by Paige, Glenn D., The Korean Decision, (N.Y.: Free Press, 1968), Ch. 12Google Scholar, “Evaluation.”

⁶¹ For a useful discussion of “effectiveness,” which is somewhat broader than the criterion of “efficiency,” see N. R. F. Maier, Problem Solving and Creativity, Ch. 1; Vroom, , “Industrial Social Psychology,” p. 229Google Scholar; and, particularly, Vroom, Victor H. and Yetton, Philip W., “Normative and Descriptive Models of Participation in Decision Making,” paper prepared for the Joint U.S.-Soviet seminar on Organization Design, Kiev, USSR, 06 1971Google Scholar. Common to these discussions is the operationalization of “effectiveness” in terms of (1) the objective quality or rationality of the decision; (2) the acceptance of the decision by subordinates who will have to execute it; (3) the amount of time required to make the decision. A conflict or latent tension between quality and acceptance is recognized, as is the desirability of resolving it in different ways, depending on circumstances. Thus, particularly germane to this paper is Vroom's recognition, with which I agree, that “quality will be weighted heavily in strategic decisions which are not reversible …,” p. 239.

⁶² Quoted by Drucker, Peter F., The Effective Executive (N.Y.: Harper & Row, 1966), p. 148Google Scholar.

⁶³ Indeed, as in psychological experiments on decision making in crisis situations, fewer alternatives for dealing with the expected damage are likely to be identified and evaluated under these circumstances. Recent research of this kind is reported by Robinson, James A., Hermann, Charles F., and Hermann, Margaret G., “Search Under Crisis in Political Gaming and Simulation,” in Theory and Research on the Causes of War, ed. Pruitt, Dean G. and Snyder, Richard C. (Englewood Cliffs, N.J.: Prentice-Hall, Inc., 1969), pp. 80–94Google Scholar. The authors report results of simulation experiments that indicate that “under crisis, both the search for alternative courses of action and the actual number of such alternatives considered by the political decision-makers are reduced” (p. 91).

They recognize, however, that earlier experiments have shown that “moderate stress produces creative decision-making, including search, and induces more search and innovation than either absence of stress or presence of intensive stress.” (On this point see also Wilensky, , Organizational Intelligence, p. 76ffGoogle Scholar. and Holsti, Crisis, Escalation, War, Chapter 1.) In fact, as in the Cuban missile crisis, when decision makers are alert to the possibility that a crisis atmosphere may disrupt rational decision making they may be able to design or introduce ad hoc procedures to safeguard against these tendencies.

⁶⁴ Geyelin, Philip, Lyndon B. Johnson and the World (New York: Praeger, 1966), pp. 244–5Google Scholar.

⁶⁵ According to one source, which is very critical of McGeorge Bundy generally, Bundy favored U.S. intervention in the Dominican crisis. If so, it must have made it difficult to combine the role of policy adviser with that of “protecting the President's options” (Halberstam, David, “The Expensive Education of McGeorge Bundy,” Harper's [07, 1969], p. 23Google Scholar).

⁶⁶ Some of the major accounts on the Dominican intervention, in addition to Geyelin's book, are; Martin, John Bartlow, Overtaken by Events (New York: Doubleday, 1966)Google Scholar; Draper, Theodore, The Dominican Revolt (N.Y.: Commentary, 1968)Google Scholar; Szulc, Tad, Dominican Diary (N.Y.: Delacorte Press, 1965)Google Scholar; Evans, Rowland and Novak, Robert, Lyndon B. Johnson: The Exercise of Power (N.Y.: The New American Library, 1966)Google Scholar; Kurzman, Dan, Santa Domingo: Revolt of the Damned (N.Y.: Putnam, 1965)Google Scholar; Center for Strategic Studies, Georgetown University, Dominican Action—1965: Intervention or Cooperation? (Washington, D.C.: Georgetown University Press, 1966)Google Scholar; Johnson, Haynes and Gwertzman, Bernard M., Fulbright: The Dissenter (N.Y.: Doubleday, 1968)Google Scholar; Slater, Jerome, Intervention and Negotiation (N.Y.: Harper & Row, 1970)Google Scholar; in addition, I have drawn on two seminar papers on the Dominican intervention by Stanford University graduate students, Robert Service and Robert G. Weinland.

⁶⁷ In a recent detailed analysis of available sources, Robert F. Randle concedes that Dulles's reaction to Radford's initial proposal for U.S. air strikes to relieve pressure on Dienbienphu is not known, but offers the interpretation that Dulles did not really favor the plan (Geneva 1954 (Princeton: Princeton University Press, 1969), pp. 65, 97Google Scholar). Randle's interpretation of Dulles's attitude contradicts that of other scholars; but the question of Dulles's position on the matter is not critical for the analysis in this paper.

⁶⁸ Ridgway's claim, however, cannot be easily verified on the basis of existing sources, which do not pinpoint the date of its completion and submission closely enough to judge at what point it entered into policy making. On the other hand, knowledge of Ridgway's strong opposition to intervention and of the fact that he was preparing a detailed study to back his position may have had some impact even though, as seems to be the case, the report was formally submitted in late April after Eisenhower made his decision not to intervene. More generally, as Stanford University student Stuart Baskin emphasizes in a recent seminar paper (December, 1971), the level of information and analysis on the situation throughout the government was remarkably poor both before and during the crisis, with the belated exception of Ridgway's report. It must be said, therefore, that while multiple advocacy was unusually vigorous and effective in this case, none of the participants in the policy debate possessed an adequate analytical basis for his statements.

See also Ridgway, Matthew B., Soldier (N.Y.: Harper & Row, 1956)Google Scholar; also “What Ridgway Told Ike—War in Indochina Would be Tougher than Korea,” U.S. News & World Report, 06 25, 1954Google Scholar; Taylor, Maxwell D., The Uncertain Trumpet (N.Y.: Harper & Row, 1959), p. 24Google Scholar; Gurtov, Melvin, The First Vietnam Crisis (N.Y.: Columbia University Press, 1967), pp. 125–6, 136, 146Google Scholar; The Pentagon Papers (N.Y.: N.Y. Times, Bantam Books, 1971), Ch. 1Google Scholar; The Senator Gravel Edition: The Pentagon Papers (Beacon Press, 1971), Vol. IGoogle Scholar. (I am also indebted to Jane Howland and Peter Fitzpatrick for earlier research on this case, in which many other sources were also consulted.)

⁶⁹ “National Security Council,” p. 723. Others who have called attention to the subterfuge of “straw man” options include Destler, Presidents, Bureaucrats, and Foreign Policy, Ch. 5; Leacocos, , “Kissinger's Apparat,” p. 24Google Scholar; Yost, Charles, “On Affairs at State,” N.Y. Times, 05 30, 1971Google Scholar.

⁷⁰ Of the many accounts and sources available in addition to The Pentagon Papers, see particularly Thomson, James C. Jr., “How Couid Vietnam Happen?” (Atlantic Monthly, 1968)Google Scholar; Geyelin, Philip, Lyndon B. Johnson and the World (N.Y.: Praeger, 1966)Google Scholar; Pfeffer, Richard M., ed., No More Vietnams? (N.Y.: Harper & Row, 1968)Google Scholar; Hoopes, Townsend, The Limits of Intervention (N.Y.: David McKay, 1969)Google Scholar; Cooper, Chester, The Lost Crusade: America in Vietnam (N.Y.: Dodd, Mead & Co., 1970)Google Scholar; Kalb, Marvin and Abel, Elie, Roots of Involvement (N.Y.: Norton, 1971)Google Scholar; Johnson, Lyndon B., The Vantage Point (N.Y.: Holt, Rinehart and Winston, 1971), Ch. 6Google Scholar.

⁷¹ Thomson, , “How Could Vietnam Happen?” (p. 52)Google Scholar goes so far as to assert that Johnson's advisers converged on the option of bombing North Vietnam despite his request for all the options. On the other hand, other sources suggest that on more than one occasion Johnson gave his advisers clear indications that he did not intend to lose South Vietnam.

⁷² Ball avoids discussion of his devil's advocate role in his book, The Discipline of Power (Boston: Little, Brown & Co., 1968)Google Scholar. His participation in the policy debate in November, 1964, is documented in the Pentagon Papers and he has referred to it publicly on a number of occasions.

Vice-President Humphrey is reported to have made a last-ditch attempt, following the VC attack on Pleiku in February, 1965, to prevent the bombing of North Vietnam. According to Townsend Hoopes, former undersecretary of the Air Force, Humphrey's views were received at the White House “with particular coldness, and he was banished from the inner councils for some months thereafter, until he decided to ‘get back on the team’” (Hoopes, , Limits of Intervention, p. 31Google Scholar).

⁷³ There were, of course, other important decisional premises as well. Leslie H. Gelb attaches importance, among other things, to the belief that loss of South Vietnam would have a variety of severe domestic political consequences (“Vietnam: How the System Worked,” Foreign Policy [Summer, 1971], 140–167Google Scholar). Daniel Ellsberg places central emphasis on Gelb's point in his own interpretation, “The Quagmire Myth and the Stalemate Machine,” Public Policy, 19 (Spring 1971) 217–274Google Scholar.

⁷⁴ Taking note of this, Richard C. Snyder and Glenn D. Paige emphasize that the definition of a situation itself can, in effect, exclude consideration of any but a single course of action; see “The United States Decision to Resist Aggression in Korea,” Administrative Science Quarterly, 3 (12, 1958) 245Google Scholar; Truman, Harry S., Memoirs: Years of Trial and Hope, Vol. II, Garden City: Doubleday & Co., 1956, pp. 332ffGoogle Scholar; Paige, Glenn D., The Korean Decision, pp. 98–99, 113–115, 124–6, 148–9, 174Google Scholar; George, Alexander L., “American Policy-making and the North Korean Aggression,” World Politics, 7 (01., 1955), 209–32CrossRefGoogle Scholar.

⁷⁵ Paige, Glenn D., The Korean Decision, N.Y.: The Free Press, 1968; p. 136, 164–5, 255, 257, 260–1, 300–1Google Scholar. Kennan's hesitation over use of U.S. ground troops, however, should not be exaggerated into making him, as some revisionist writers have tried to do, into an opponent of U.S. intervention. In his memoirs Kennan states that from the onset of the Korean crisis he felt that the United States would have to react with all necessary force to repel the attack and to expel North Korean forces from South Korea. He also favored prompt steps to protect Formosa from falling into Communist hands on the ground that “two such reverses coming one on the heel of the other could easily prove disastrous to our prestige and to our entire position in the Far East” (Memoirs, 1925–1950, [Boston: Little, Brown & Co., 1967], p. 486Google Scholar).

⁷⁶ Kennan, pp. 487–96; also Higgins, Trumbull, Korea and the Fall of MacArthur (N.Y.: Oxford University Press, 1960), pp. 78–79Google Scholar; Osgood, Robert E., Limited War (Chicago: University of Chicago Press, 1957), pp. 183–84Google Scholar.

⁷⁷ In a laudatory account of the way in which Truman organized and managed the process of foreign policy making, Dean Acheson credits him with infusing an adversary process into N.S.C. meetings similar to that of the law court. It is clear that what Acheson endorses in this respect is the formal, orderly variant of the adversary process—one which worked to Acheson's advantage because Truman acknowledged and leaned on his special competence in foreign affairs—and not the unstructured and unpredictable variants of multiple advocacy associated with the game of bureaucratic politics. Acheson's distaste for unstructured, informal variants of multiple advocacy—in which he enjoyed less influence—emerges clearly from his critical account of the workings of the ad hoc Executive Committee of the N.S.C. at the time of the Cuban missile crisis, in which Acheson participated at President Kennedy's invitation (Acheson, , Present at the Creation, [N.Y.: Norton, 1969], pp. 733–737Google Scholar; see also Acheson's, review of Kennedy's, R. F.Thirteen Days in Esquire, 02 1969Google Scholar).

⁷⁸ Anderson, Patrick, The Presidents' Men (Garden City, N.Y.: Doubleday, 1968), p. 116Google Scholar.

⁷⁹ Geyelin, , LBJ and the World, p. 162Google Scholar. A detailed analysis of the workings of the bureaucratic politics process leading up to the final confrontation of the M.L.F. advocates with President Johnson, based partly on interviews, is provided in Steinbruner, John D., “The Mind and Milieu of Policy-makers: A Case Study of the M.L.F.” Ph.D. dissertation, Department of Political Science, M.I.T., 02 1968Google Scholar. For this portion of the case study, however, Steinbruner relies on Geyelin's account, which he regards as accurate.

⁸⁰ Geyelin, p. 170. Other accounts represent McGeorge Bundy intervening not as a neutral watchdog of the President's options but as a concealed advocate against the M.L.F. who had bided his time. (Anderson, Patrick, The Presidents' Men, pp. 271—2Google Scholar; Halberstam, David, “The Expensive Education of McGeorge Bundy,” Harper's, 07 1969, p. 29.Google Scholar)

⁸¹ Geyelin, p. 162.

⁸² Neustadt, Richard, Presidential Power (N.Y.: Wiley, 1960), p. 145Google Scholar. Neustadt makes use in his account of Martin Lichterman's correspondence and an interview with Acheson, reported in Lichterman, , “To the Yalu and Back,” in American Civil-Military Decisions, ed. Stein, Harold (University of Alabama Press, 1963), p. 602Google Scholar.

⁸³ Neustadt's and Lichterman's accounts of these events were qualified some years later both by Dean Acheson and General J. Lawton Collins, Army Chief of Staff at the time. (Acheson, , Present at the Creation, pp. 466–68, 754Google Scholar; Collins, , War in Peacetime: The History and Lessons of Korea [Boston: Hough ton Mifflin Co., 1969], pp. 202, 205–217.Google Scholar) Neustadt's interpretation was questioned earlier also by McLellan, David S. in his “Dean Acheson and the Korean War,” Political Science Quarterly, 83 (03 1968), 16–39CrossRefGoogle Scholar.

McLellan agrees that Acheson and other top-level advisers did labor under the “dread of Chinese involvement” in early November. He notes, however, that at two important policy meetings on November 21, 1950, Acheson displayed relative optimism that MacArthur could accomplish his mission. McLellan is perplexed by this marked change of mood. He fails to see that it can be traced to the Chinese disengagement on the battlefield that took place after the high-point of Washington's (and MacArthur's) anxiety on November 9. When day after day passed without further combat or major contact with Chinese forces, the acute anxiety American leaders had felt in early November gradually declined. It is evident from McLellan's account of the November 21 policy meetings—the content of which is disclosed for the first time by McLellan—that Acheson had now fully overcome the concern he and others had felt earlier in the month about the riskiness of the military strategy MacArthur was bent on pursuing. The seeming disappearance of the Chinese forces and the continued quiescence of the battlefield evidently strengthened hopes in Washington that MacArthur might succeed after all in repelling Chinese intervention. Having considered the need for a change in presidential directives to MacArthur earlier in November without acting to bring it about, Acheson at a meeting with his staff on November 21 expressed the belief that the directive to MacArthur should not be changed until the General had had a chance to probe the situation and push forward his planned offensive. The fact of this change of mood in Washington was carefully withheld by administration witnesses during the Congressional hearings on MacArthur's dismissal in 1951. Such a disclosure would have been damaging to the administration, which represented itself as having remained deeply concerned about MacArthur's military operations up to the point at which the Chinese launched their all-out offensive in late November.

⁸⁴ On this general point see also Holsti, O. R., Crisis, Escalation, War, p. 21Google Scholar.

⁸⁵ Sorensen, Theodore C., Kennedy (N.Y.: Harper & Row, 1965), p. 302Google Scholar. Cf. also the detailed account of flaws in decision making in the Bay of Pigs case in Arthur Schlesinger, Jr., A Thousand Days, Chapter 10; and Bowles, Chester, Promises to Keep (N.Y.: Harper & Row, 1971), p. 326Google Scholar.

⁸⁶ Anderson, , The Presidents' Men, pp. 264–5Google Scholar. At the same time, the creation of the “situation room” paved the way for the NSC staff's extensive involvement in the internal activities of the departments and agencies. (Clark, K. C. and Legere, Lawrence J., eds., The President and the Management of National Security, [N.Y.: Praeger, 1969], pp. 73–4Google Scholar.)

⁸⁷ A conscious attempt by Washington to develop an alternative source of information on events in the Dominican Republic was undertaken finally on April 29 after the U.S. decision to intervene overtly had been made and was being implemented. This was the Presidential mission of John Bartlow Martin.

⁸⁸ Geyelin, , LBJ and the World, p. 244–5Google Scholar.

⁸⁹ Sorenson, pp. 304, 306; Schlesinger, ch. 10.

⁹⁰ Among the numerous sources consulted for this case the following were particularly useful: Schlesinger; Sorensen; Hilsman, Roger, To Move A Nation (N.Y.: Doubleday & Co., 1967)Google Scholar; Chester Bowles, Promises to Keep; Galbraith, John K. and Lewis, Anthony, “The World Through Galbraith's Eyes,” A Tape Recorded Conversation, N.Y. Times Magazine, 12. 18, 1966Google Scholar; Stempel, John D., “Policy/Decision-Making in the Dept. of State: Vietnamese Problem, 1961–1965,” Ph.D. dissertation, Political Science Department, University of California, Berkeley, 1965Google Scholar; Ball, George W., Discipline of Power, pp. 333–4Google Scholar; Reston, James, The Artillery of the Press, pp. 25–6Google Scholar; Ellsberg, Daniel, “The Quagmire Myth and the Stalemate Machine,” Public Policy, 19 (Spring, 1971), 217–274Google Scholar; Schlesinger, Arthur Jr., “Eyeless in Indochina,” New York Review of Books, 10 21, 1971Google Scholar; The Pentagon Papers. The research assistance of Jane Howland and David K. Hall on this case is gratefully acknowledged.

⁹¹ Schlesinger, , A Thousand Days, pp. 250, 255Google Scholar; Sorensen, p. 305.

⁹² Cutler, Robert, No Time for Rest (Boston: Little, Brown & Co., 1966), pp. 315–316Google Scholar. A useful, up-to-date review of the development of the NSC and the variations in its importance, organization and performance is provided in Clark and Legere, President and Management of National Security (see especially Chapter 4 by Richard M. Moose and Chapter 10 by John Ponturo). Useful materials are also contained in Patrick Anderson, The Presidents' Men; Johnson, Robert H., “The National Security Council: The Relevance of its Past to its Future,” Orbis, 13 (Fall, 1969), 709–735Google Scholar; Paul Y. Hammond, Organizing for Defense; SenatorJackson, Henry M., ed., The National Security Council (N.Y.: Praeger, 1965)Google Scholar.

⁹³ Patrick Anderson emphasizes the basic duality that affected Bundy's performance as Special Assistant. He suggests that Bundy was aware of the latent role-conflicts in the position and tried to give priority to his task of keeping the President's choices open. (The Presidents' Men, p. 270). Material on Bundy's alleged tendency to shift from arbiter to advocate on a number of occasions is reported in the highly critical appraisal offered by David Halberstam (“The Very Expensive Education of McGeorge Bundy,” pp. 23, 29, 34–35.) Irving Destler's Presidents, Bureaucrats, and Foreign Policy helps clarify to some extent the role interpretations and performance of Bundy and his successors. David K. Hall is undertaking a comprehensive study of the development and performance of role tasks associated with the Executive Secretary and the Special Assistant within the NSC system as part of his Ph.D. dissertation in the Political Science Department, Stanford University.

⁹⁴ Robert Cutler, pp. 315–316.

⁹⁵ A former long-term member of the NSC staff notes that the individuals Eisenhower selected to be his Special Assistant for National Security Affairs “were not, with the single exception of his last appointee, Gordon Gray, experts in foreign affairs or defense policy. They were the President's agents in operating the NSC system, but they were not, by and large, independent sources of power and advice” (Johnson, Robert H., “National Security Council,” p. 716Google Scholar).

⁹⁶ See, for example, Hoopes, Townsend, Limits of Intervention, pp. 59–61, 116, 123Google Scholar.

⁹⁷ Destler, Presidents, Bureaucrats, and Foreign Policy, Ch. 5.

⁹⁸ Personal communication (1971). Cronin considers this and related problems in his recent study of White House “staffers” and their relations with department officials. “‘Everybody Believes in Democracy Until He Gets to the White House …’: An Examination of White House-Department Relations,” Law and Contemporary Problems, 35 (Summer, 1970), 573–625CrossRefGoogle Scholar (published July 1971 at the School of Law, Duke University).

I am very much in agreement, as this paper indicates, with the suggestion made earlier by Cronin that political scientists should give more attention to examining the roles and perspectives of members of the presidential advisory system. (Cronin, T. E., “Political Science and Executive Advisory Systems,” in Cronin, T. E. and Greenberg, S. D., eds., The Presidential Advisory System [N.Y.: Harper & Row, 1969], pp. 321–335.Google Scholar)

⁹⁹ Suggestive in this respect is the systematic effort that management specialists are making to draw on available empirical evidence to specify the set of rules which should be used by an executive in determining the form and amount of participation by subordinates to be used in different classes of situations. (Vroom, and Yetton, , “Normative and Descriptive Models of Participation in Decision-Making,” pp. 8–9Google Scholar; for a fuller statement see these authors' forthcoming book, Leadership and Decision-Making [ms.].)