1 Farrar, L. L. Jr., “The Limits of Choice: July 1914 Reconsidered,” in Small, Melvin and Singer, J. David (eds.), International War: An Anthology and Study Guide (Homewood, 111.: Dorsey Press, 1985), 251–52.Google Scholar

2 The classic expositions from sociology and social psychology may be found in Coser, Lewis A., The Functions of Social Conflict (New York: Free Press, 1956);Google ScholarReuck, Anthony de and Knight, Julie, Conflict in Society (Boston: Little, Brown, 1966);Google ScholarKluckhohn, Clyde, Mirror for Man (Greenwich, Conn.: Fawcett Books, 1960);Google ScholarSimmel, Georg, Conflict and the Web of Group Affiliations (Glencoe, 111.: Free Press, 1955);Google Scholar and Sherif, Mustafa Sherif and C. W., Groups in Harmony and Tension (New York: Harper, 1955).Google Scholar

3 Among the most prominent assertions are those of Quincy Wright, A Study of War (Chicago: University of Chicago Press, 1965), 1016;Google ScholarPubMedHaas, Ernst B. and S., AllenWhiting, Dynamics of International Relations (New York: McGraw-Hill, 1956), 62;Google Scholar and Rosecrance, Richard N., Action and Reaction in World Politics: International Systems in Perspective (Boston: Little, Brown, 1963), 304, 306.Google Scholar

4 Rummel, R. J., “Dimensions of Conflict Behavior Within and Between Nations,” General Systems Yearbook 8 (1963), 1–50.Google Scholar

5 Stohl, Michael, “The Nexus of Civil and International Conflict,” in Gurr, Ted Robert (ed.), Handbook of Political Conflict: Theory and Research (New York: Free Press, 1980), 325.Google Scholar

6 Jonathan Wilkenfeld's highly influential studies include “Domestic and Foreign Conflict Behavior of Nations,” Journal of Peace Research 5 (1968), 55–69Google Scholar; “Some Further Findings Regarding the Domestic and Foreign Conflict of Nations,” Journal of Peace Research 7 (1969), 147–56;Google ScholarZinnes, Dina and Wilkenfeld, Jonathan, “An Analysis of Foreign Conflict Behavior of Nations,” in Hanrieder, Wolfram (ed.), Comparative Foreign Policy: Theoretical Essays (New York: David McKay, 1971);Google ScholarWilkenfeld, Jonathan, “Models for the Analysis of Foreign Conflict Behavior of States,” in Russett, Bruce M. (ed.), Peace, War and Numbers (Beverly Hills: Sage, 1972);Google Scholar and Wilkenfeld, and Zinnes, , “A Linkage Model of Domestic Conflict Behavior,” in Wilkenfeld, (ed.), Conflict Behavior and Linkage Politics (New York: David McKay, 1973).Google Scholar

7 See Collins, John, “Foreign Conflict Behavior and Domestic Disorder in Africa,” in Wilkenfeld, (ed.), Conflict Behavior and Linkage Politics;Google Scholar and Onate, Andres, “The Conflict Interactions of the People's Republic of China, 1950–1970,” Journal of Conflict Resolution 18 (1974), 578–94.CrossRefGoogle Scholar

8 A more extensive discussion of these issues is contained in James, Patrick, “Conflit et cohésion: nouveau regard sur la littérature savant,” Études Internationales 17 (1986), 621–33.CrossRefGoogle Scholar

9 Coser, The Functions of Social Conflict, 88, 92.

10 Blainey, Geoffrey, The Causes of War (London: Macmillan, 1973), 81.CrossRefGoogle Scholar

11 Ward, Michael D. and Widmaier, Ulrich, “The Domestic-Foreign Conflict Nexus: New Evidence and Old Hypotheses,” International Interactions 9 (1982), 78.CrossRefGoogle Scholar

12 Ibid.

13 Brecher, Michael and James, Patrick, Crisis and Change in World Politics (Boulder: Westview Press, 1986), 22.Google Scholar

14 Foran extended discussion of this definition, see Brecher, Michael, “Toward a Theory of International Crisis Behavior: A Preliminary Report,” International Studies Quarterly 21 (1977), 39–74.CrossRefGoogle Scholar

15 Only the principal adversaries are listed for each crisis. Other states also may perceive a crisis (or become involved in some way), but these additional actors do not concern the model of the conflict nexus that is being presented here.

16 Some of the role designations may appear unusual given the political logic of externalization. For example, in cases 1,18 and 90, the much stronger defender would appear to have been the aggressor. However, the coding used is appropriate because each initiator tried to change the status quo in an area of basic concern to the defender. Moral considerations deliberately are left out of the formulation, hence the existence of some counterintuitive role designations.

17 This description follows International Crisis Behaviour (ICB) Project Case Summary #195, in Michael Brecher and Jonathan Wilkenfield, Handbook on International Crises, forthcoming.

18 The internal conflict data are aggregated on an annual basis, so it is impossible to measure change over the exact one-year period preceding the outbreak of an international crisis. However, the current approach does give an approximate measurement of the trend in domestic conflict over the pre-crisis phase.

19 The notation used for expected value corresponds to that of Bruce Bueno de Mesquita, The War Trap (New Haven: Yale University Press, 1981).Google ScholarPubMed

20 Rigorous derivations of these categories appear in Gurr, Ted Robert, Why Men Rebel (Princeton: Princeton University Press, 1970);Google Scholar and Hibbs, Douglas A. Jr., Mass Political Violence: A Cross-National Causal Analysis (New York: Wiley, 1973).Google Scholar While some of the studies relying upon factor analysis also produced wide-ranging components of internal and external conflict, they derived factors statistically and then correlated them. The current approach is different because civil war and protest behaviour are hypothesized to show specific linkages to crisis outcomes.

21 Ward and Widmaier (“The Domestic-Foreign Conflict Nexus”) also considered the use of riots to represent domestic conflict. However, they discovered that data on riots exhibited reliability problems across the two volumes of the World Handbook and therefore excluded that variable from the measurement of civil war. For definitions of protest demonstrations, political strikes, armed attacks, and deaths from domestic violence, see Taylor, Charles Lewis and Jodice, David H. (eds.), World Handbook of Political and Social Indicators (3rd ed.; New Haven: Yale University Press, 1981).Google Scholar

22 The use of percentages rather than logarithms would create a bias in the other direction. For example, in percentage terms a change from 1 to 2 would outweigh an increase from 10 to 15 (100% compared to 50%). By contrast, in logarithmic terms the respective scores generated by equation (1) would be 0.30 and 0.78. Another alternative—weighting the scores by population size—would have approximately the same effect as the logarithmic transformation. However, it also would require the incorporation of further data, thus increasing the risk of using unreliable cross-national statistics.

23 This analysis is borrowed from Babbie, Earl, The Practice of Social Research (Belmont, Calif.: Wadsworth, 1983), 371Google Scholar.

24 de Mesquita, Bueno, The War Trap, 46–49.Google Scholar

25 Bueno de Mesquita has made his data available for many of the 133 cases; the other utility scores have been calculated as described in The War Trap, 109–18.

26 The primary source for GNP data is the United Nations Statistical Office, Statistical Yearbook (New York: UN Department of Social Affairs, 1948-1975).Google ScholarPubMed

27 The War Trap, 105.

28 Organski, A. F. K. and Kugler, Jacek, The War Ledger (Chicago: University of Chicago Press, 1980), 67.Google Scholar

29 Ibid., 68.

30 William B. Moul, “Measuring the Balances of Power: A Look at Some Numbers,” paper presented at the annual meeting of the Canadian Political Science Association, Winnipeg, 1986.

31 The zero cases occur only between allies with identical alliance commitments. Thus even small positive or negative values are regarded as fundamentally different from scores of zero.

32 These rules of categorization are discussed at length in Small, Melvin and Singer, J. David, Resort to Arms: International and Civil Wars, 1816–1980 (Beverly Hills: Sage, 1980), 31–61.Google Scholar There are a few points of disagreement with COW concerning the wars that are included in Table 1. China Civil War (1948), Indonesian Independence (1948) and War of Attrition (1969) are here counted as wars even though they are not in the Singer-Small compilation. The first case is regarded as suitable because the US and Nationalist China were involved in the crisis and each was a nation-state, with the transition for the People's Republic into recognized status being difficult to pinpoint. When the military operations of the Dutch are taken into consideration, the Indonesian case is consistent with the conventional meaning of warfare, even though the casualty levels sustained are not easy to specify. An exception also is made for the War of Attrition because the historical record demonstrates that the participants considered it to be a war in the common sense of the word.

33 Probit is a “particular nonlinear specification of the probability model” for a dichotomous dependent variable. It is based on the cumulative normal distribution function. See Aldrich, John H. and Nelson, Forrest D., Linear Probability, Logit, and Probit Models (Beverly Hills: Sage, 1984), 34–37.CrossRefGoogle Scholar

34 Useful explanations of probit analysis can be found in SPSS, User's Guide (2nd ed.; New York: McGraw-Hill, 1986);Google ScholarPubMed and Aldrich and Nelson, Linear Probability, Logit, and Probit Models.

35 Ibid., 57–59.

36 Seven cases have been dropped from the analysis because of missing data, including two wars: Korean War I (1950), Ethiopia/Somalia (1960), Congo I: Katanga (1960), Mali Federation (I960), Congo II (1964), Guinea Regime (1965) and War In Angola (1975).

37 Brecher and Wilkenfield, Handbook, Case Summary #203.

38 Brecher and Wilkenfield, Handbook, Case Summary #194.

39 Brecher and Wilkenfield, Handbook, Case Summary #115.

40 Lawson's conclusions about Syria's 1976 intervention in Lebanon are consistent with several of the arguments made in the present study. See Lawson, Fred H., “Syria's Intervention in the Lebanese Civil War, 1976: A Domestic Conflict Explanation,” International Organization 38 (1984), 451–80.CrossRefGoogle Scholar