¹ Hopkins, Raymond F. and Mansbach, Richard W., Structure and Process in International Politics (New York: Harper and Row 1973), 100Google Scholar.

² Young, Oran R., ”Interdependencies in World Politics,” International Journal, xxiv (Autumn 1969), 726CrossRefGoogle Scholar.

³ Hinsley, F. H., Power and the Pursuit of Peace (Cambridge, England: Cambridge University Press 1967), 13Google Scholar.

⁴ See the discussion in Claude, Inis L., Swords Into Plowshares (4th ed.; New York: Random House 1971), 250Google Scholar–56.

⁵ Claude, Inis L., ”The OAS, the UN, and the United States,” International Conciliation, No. 547 (March 1964), 11Google Scholar. On the Chapter VIII arrangements see also Russell, Ruth B., A History of the United Nations Charter (Washington: Brookings Institution 1958Google Scholar); and Goodrich, Leland M., Hambro, Edward, and Simons, Anne P., Charter of the United Nations (3rd rev. ed.; New York: Columbia University Press 1969Google Scholar), esp. 354–69.

⁶ See the discussion in Claude (fn. 4), 115–17. The Churchill-U.S. differences are detailed in Russell (fn. 5), 105 ff.

⁷ Studies directly concerned with the management of interstate security disputes on a comparative organizational basis include: Claude (fns. 4 and 5); Etzioni, Minerva, The Majority of One (Beverly Hills: Sage 1970Google Scholar); Ernst B. Haas, Collective Security and the Future International System (University of Denver, Graduate School of International Studies, Monograph Series in World Affairs, 1967–1968, V); Haas, Ernst B., ”The United Nations and Regionalism,” International Relations, iii (November 1970Google Scholar); Haas, E. B., Butterworth, R. L., and Nye, J. S., Conflict Management by International Organizations [hereafter referred to as Conflict Management] (Morristown, N.J.: General Learning Press 1972Google Scholar); selections (esp. by Lynn Miller) in Falk, Richard A. and Mendlovitz, Saul, eds., Regional Politics and World Order (San Francisco: W. H. Freeman 1973Google Scholar); Nye, Joseph S., Peace in Parts (Boston: Little, Brown 1971Google Scholar); and publications of the UNITAR project on the relationship between the UN and regional intergovernmental organizations, esp. Robertson, A. H., The Relations Between the Council of Europe and the United Nations (New York: UNITAR Regional Study No. 1, 1972CrossRefGoogle Scholar); Andemicael, Berhanykun, Peaceful Settlement Among African States: Roles of the United Nations and the Organization of African Unity (New York: UNITAR PS No. 5, 1972Google Scholar); and Pechota, Vratislav, The Quiet Approach: A Study of the Good Offices Exercised by the United Nations Secretary-General in the Cause of Peace (New York: UNITAR PS No. 6, 1972Google Scholar).

⁸ See, for example, Kissinger's claim that the UN is ”uniquely situated to foster and to anchor in the hearts of men” a ”comprehensive, institutionalized peace encompassing all nations” (text of speech delivered to the UN General Assembly on September 24, 1973, as reported in the New York Times, September 25, 1973, p. 18).

⁹ This term is from Alker, Hayward R. Jr., and Greenberg, William J., ”The UN Charter: Alternate Pasts and Alternate Futures,” in Fedder, E. H., ed., The United Nations: Problems and Prospects (St. Louis: University of Missouri, Center for International Studies, 1971Google Scholar).

¹⁰ Haas, ”The United Nations …” (fn. 7). Haas's figures are higher than those presented in part IV of the present study because he includes collective defense organizations like NATO and SEATO in his survey.

¹¹ ”By compatibility we mean that the relationship between two organizations is such that the activities of one do not undermine those of the other and vice versa. The extent of compatibility of regional organizations with a universal body should be judged according to the conflicts between the two and the degree to which such conflicts are minimized. Universal organizations and compatible regional organizations complement each other; there is an interest in maintaining harmonious relations in the sharing of international tasks.” Etzioni (fn. 7), 18.

¹² Note that this point refers to broader systems characteristics than simply ”back-up” capabilities. It encompasses a major element of what Landau refers to as an ”energy-transfer model”; in particular, the characteristic of a living system ”that it is 'self-regulating,' acting always to keep both its internal state and external environment stable (its regulatory processes are manifold, complex, and redundant, and serve to protect against threat).” Landau, Martin, ”Linkage, Coding, and Intermediacy: A Strategy for Institution-Building,” Journal of Comparative Administration, 11 (February 1971), 403Google Scholar.

¹³ See Claude (fn. 4), 102–17.

¹⁴ See Nye (fn. 7), esp. 129–30. Nye quotes Nyerere as an example: ”Certain regional and ideological associations have an advantage over the UN. As a means of preventing or settling disputes, talking is more productive and certainly easier, the greater the general feeling of sympathy and friendship among the participants.”

¹⁵ Claude (fn. 4), 103.

¹⁶ Haas, ”Collective Security …” (fn. 7).

¹⁷ See the appendix to Conflict Management (fn. 7), for the complete codebook that was used.

¹⁸ This power ranking was obtained from the Cox-Jacobson scale, reported in Cox, Robert W. and Jacobson, Harold K., The Anatomy of Influence (New Haven: Yale University Press 1973), 437Google Scholar–43.

¹⁹ The coding procedures are described fully in Conflict Management (fn. 7). For those variables requiring judgmental scoring, panels of experts were set up for each organization, and each panel (except two) included at least one scholar who had published studies on the topic and one practicing diplomat from the organization in question. Two members of the panel were able to work face-to-face, while the others responded by mail. This constituted the first round of coding, during which disagreements were resolved (not compromised) by further research and appeal back to the panel members. The second round of coding involved intensive independent studies of each organization and further research into each case, with coding being done on a face-to-face basis. Th e cases during the second round were slightly different from those during the first round; during the first, however, inter-coder agreement was greater than 85%, and judgmental scores from the first round were changed during the second only if documentary evidence showed them to be incorrect.

²⁰ See Lingoes, James C., ”The Multivariate Analysis of Qualitative Data,” Multi-variate Behavioral Research, iii (January 1968), 61–94CrossRefGoogle Scholar, and The Guttman-Lingoes Non-metric Program Series (Ann Arbor: Mathesis Press 1973), esp. chap. 19. The ”alignment” variable categories were ordered as follows: 1 = members of opposing cold-war blocs; 2 = both parties unaligned; 3 = one party aligned, one not; 4 = members of same cold-war bloc. The ”type of issue” variable categories were ordered as follows: 1 = interstate cold war; 2 = internal cold war; 3 = internal ”other”; 4 = interstate ”other”; 5 = colonial. See Conflict Management (in. 7), for a fuller explication of the

²¹ The coefficient of reproducibility for this scale was 97.3%. This scale of management impact was computed differently from the ”success” scale reported in Conflict Management (fn. 7). Th e differences between the two, however, are irrelevant for present purposes; a few cases are shifted from one category to the next, but cases that scored as ”failures” on the ”success” scale are, of course, scored as ”none” on the impact scale.

²² The factor loadings were squared (preserving the original signs), multiplied by the standardized variables, and summed for each case. This technique was used in a recent article by Beck, Paul Allen, ”Environment and Party: The Impact of Political Demographic County Characteristics on Party Behavior,” American Political Science Review, Vol. 68 (September 1974), 1229CrossRefGoogle Scholar–44. For basic considerations important in composite estimates techniques, see Rummel, R. J., Applied Factor Analysis (Evanston: Northwestern University Press 1970), 441Google Scholar–42. In some respects the procedure employed here could be considered to reify factor loadings by treating different weightings with spurious precision. Consequently three other scales, corresponding to those reported in the text, were calculated by taking only the variables that loaded strongly on a factor and simply summing their values to score each case. The correlations between the original and alternative scales were .97, .91, .92 for the massive violence, great-power conflict, and intractability scales respectively.

²³ If the redundancy model were accurate, there would be only a random relationship between a conflict's characteristics and the organization to which it is submitted. It should be recalled that we are talking in terms of revealed capabilities; an artificial intelligence model like the Alker-Greenberg-Christensen one might be able to decide whether regional organizations have greater potentials, but the present analysis can deal only with the jurisdictional capabilities that have been demonstrated. Cf. Alker, Hayward R. Jr. and Christensen, Cheryl, ”From Causal Modeling to Artificial Intelligence: The Evolution of a UN Peace-Making Simulation,” in Smoker, Paul and Laponce, Jean, eds., Experimentation and Simulation in Political Science (Toronto: University of Toronto Press 1972Google Scholar); and Alker and Greenberg (fn. 9).

²⁴ The seven regional cases that are excluded as being perhaps atypical are the following: the Nigerian civil war, the Congo civil war, the Yemen civil war, the Tutsi-Hutu conflicts in Rwanda and Burundi, and the 1965 Dominican Intervention, all of which scored too high on massive violence; the 1954 Guatemala coup and the 1960–62 problems of Cuba-OAS relations, in both of which the scores were too close to superpower cold-war confrontations.

²⁵ The Biafran case might at first appear to have been an exception, but Andemicael reports: ”The United Nations Secretary-General had initially explored possibilities for exercising his good offices, and then decided to encourage the OAU to assist in the quest for a peaceful solution of the problems” (fn. 7), 35. Pechota presents essentially the same analysis, distinguishing between the political and humanitarian considerations in the conflict (fn. 7), 38–39.

²⁶ On the ”fig leaf” role, see Nye (fn. 7), 135.

²⁷ This difference, of course, might have been due to the fact that quite different regression models are needed for UN and regional management—that is, that the two sets of organizations represent essentially different management systems. That possibility was checked using the Chow test, and it was found not to hold; it thus seems that we are dealing with a case of different emphases within essentially the same system. For the technique involved in this test, see Fisher, Franklin M., ”Tests of Equality Between Sets of Coefficients in Two Linear Regressions: An Expository Note,” Econometrica, xxxviii (March 1970), 361CrossRefGoogle Scholar–66.

²⁸ These cases were the 1960–62 tensions between the Dominican Republic and the United States and Venezuela, and the 1965 intervention in the Dominican Republic. Our scoring of the former appears at first to conflict with Nye's; he found ”a general anti-Trujillo consensus” in this case (fn. 7), 143. His coding, however, seems to have focused very tightly on the tensions between Venezuela and the Dominican Republic; ours takes the Venezuelan claims more as an opportunity for the U.S. (and a few others) to get the OAS involved in the internal Dominican situation. From this perspective we credit the OAS with helping to stop hostilities because Trujillo changed his policies, and with helping to abate the conflict, because it held back from the degree of intervention that the U.S. really wanted. Cf. Slater, Jerome, The OAS and United States Foreign Policy (Columbus:Ohio State University Press 1967), 183–216Google Scholar.

²⁹ Guatemala, 1954; Lebanon/Jordan, 1958; Kuwait, 1961; Haiti, 1963; Panama, 1964.

³⁰ Such a shift would not be likely to improve the degree of management impact on a case. The weak differences in organizational impact noted in Table 4 hold for the full universe of cases: the U N manages cases less frequently than regional organizations do, but its impact tends to be stronger. If impact is scaled from 1 to 4 (1 = no impact, 4 = settlement), the mean U N impact on the cases that it managed was 2.8, compared to 2.5 for regional organizations.

³¹ On the notion of ”steering,” see Deutsch, Karl W., The Nerves of Government (New York: Free Press 1966Google Scholar), 169 ff. On the notion of ”directive” capacities, see Eckstein, Harry, ”Authority Patterns: A Structural Basis for Political Inquiry,” American Political Science Review, Vol. 67 (December 1973), 1142CrossRefGoogle Scholar–61.

³² The question of whether IO conflict management activities have had a cumulative impact on international politics is more fully examined in Butterworth, , ”Moderating Interdependence,” Ph.D. diss. (University of California, Berkeley 1973Google Scholar).

³³ The years 1945–71 can be divided into four system periods defined by the degree of polycentrism that was apparent in U N voting on security issues: unipolar (1945–47); bipolar (1948–55); tricentric (1956–62); multicentric (1963–71). See Haas, ”Collective Security” (fn. 7) ; cf. Hayward R. Alker, Jr. and Bruce Russett, World Politics in the General Assembly (New Haven: Yale University Press 1965), 20–22. The rate of management impact and its degree both declined in the recent period but the total number of disputes increased, suggesting that increasing pluralism in the international system has been associated with an increasing practical irrelevance of these organizations to security problems.

³⁴ Reference here is to the kind of process described by Claude (fn. 5), whereby the UN can be used to keep regional agencies ”honest.”