¹ Claude, Inis L. Jr, Power and International Relations (New York: Random House, 1962)Google Scholar; Guliek, Edward, Europe's Classical Balance of Power (Ithaca, N.Y.: Cornell University Press, 1955)Google Scholar; Kaplan, Morton A., System and Process in International Politics (New York: John Wiley, 1957)Google Scholar; Liska, George, Nations in Alliance: The Limits of Interdependence (Baltimore: Johns Hopkins University Press, 1962)Google Scholar; Morgenthau, Hans J., Politics among Nations (New York: Alfred A. Knopf, 1973)Google Scholar; Waltz, Kenneth N., Theory of International Politics (Reading, Mass.: Addison-Wesley, 1979)Google Scholar; Wright, QuincyA Study of War, 2d ed. (Chicago: University of Chicago Press, 1965).Google Scholar There have been some efforts to develop formal models of the balance of power. See, e.g., Niou, Emerson M. S., Ordeshook, Peter C., and Rose, Gregory F., The Balance of Power: Stability in International Systems (Cambridge: Cambridge University Press, 1989)CrossRefGoogle Scholar; Harrison Wagner, R., “The Theory of Games and the Balance of Power,” World Politics 38 (July 1986)Google Scholar; Zinnes, Dina, “An Analytical Study of the Balance of Power Theories,” Journal of Peace Research 4 (1967).CrossRefGoogle Scholar But these models have not been tested against empirical reality. Some recent empirical studies that support the balance-of-power hypothesis that equality of power tends to lead to war include Ferris, Wayne H., The Power Capabilities of Nation-States (Lexington, Mass.: Lexington Books, 1973)Google Scholar; David Singer, J., Bremer, Stuart A., and Stuckey, John, “Capability Distribution, Uncertainty, and Major Power War, 1820” in Russett, Bruce M., ed., Peace, War, and Numbers (Beverly Hills, Calif.: Sage, 1972)Google Scholar; Siverson, Randolph M. and Tennefoss, Michael R., “Power, Alliance, and the Escalation of International Conflict, 1815–1965,” American Political Science Review 78 (December 1984).CrossRefGoogle Scholar For an excellent review of the balance-of-power theory, see Levy, Jack S., “The Causes of War: A Review of Theories and Evidence,” in Tetlock, Philip E. et al., eds., Behavior, Society, and Nuclear War, vol. 1 (Baltimore: Johns Hopkins University Press, 1989).Google Scholar

² Gilpin, Robert, War and Change in World Politics (Cambridge: Cambridge University Press, 1981)CrossRefGoogle Scholar; Modelski, George, “The Long Cycle of Global Politics and the Nation State,” Comparative Studies in Society and History 20 (April 1978)CrossRefGoogle Scholar; Organski, A. F. K., World Politics, 2d ed. (New York: Alfred A. Knopf, 1968)Google Scholar; Organski, A. F. K. and Kugler, Jacek, The War Ledger (Chicago: University of Chicago Press, 1980)Google Scholar; Thompson, William R., On Global War (Columbia: University of South Carolina Press, 1988).Google Scholar Although Gilpin, Modelski and Thompson, and Organski differ on many points—what constitute capabilities, the way differential growth rates are operationalized, how to understand motivation and the behavior of the dominant state and the challenging power, and the definition of great power war— the similarities among them are nevertheless profound. They all focus on how changes in the distribution of capabilities drive large wars, and all consider power transitions to be critical, central moments in history.

² Mesquita, Bruce Bueno de and Lalman, David, “Empirical Support for Systemic and Dyadic Explanations of International Conflict,” World Politics 41 (October 1988)CrossRefGoogle Scholar; idem, War and Reason (New Haven: Yale University Press, 1992); Houweling, Henk and Siccama, Jan G., “Power Transitions as a Cause of War,” Journal of Conflict Resolution 32 (March 1988)CrossRefGoogle Scholar; idem, “Power Transitions and Critical Points as Predictors of Great Power War,” Journal of Conflict Resolution 35 (December 1991); Kim, Woosang, “Power, Alliance, and Major Wars, 1816–1975,” Journal of Conflict Resolution 33 (June 1989)CrossRefGoogle Scholar; idem, “Alliance Transitions and Great Power War,” American Journal of Political Science 35 (November 1991); Kim, Woosang and Morrow, James D., “When Do Shifts in Power Lead to War?” American Journal of Political Science (forthcoming); Organski and Kugler (fn. 2).Google Scholar

⁴ Levy, Jack S., “Alliance Formation and War Behavior: An Analysis of the Great Powers,” Journal of Conflict Resolution 25 (December 1981)CrossRefGoogle Scholar; idem, War in the Modern Great Power System, 1495–1975 (Lexington: University Press of Kentucky, 1983); Modelski, George and Thompson, William R., Seapower in Global Politics, 1494–1993 (Seattle: University of Washington Press, 1988)CrossRefGoogle Scholar; idem, “Long Cycles and Global War,” in Midlarsky, Manus I., ed., Handbook of War Studies (Boston: Unwin Hyman, 1989)Google Scholar; Thompson (fn. 2).

⁵ An exception is Thompson's study on the power transition theory. Although his research does not employ a rigorous statistical analysis, he examines the trends of relative naval capabilities among the principal contenders in the international system to test a key power transition hypothesis. See Thompson, William R., “Succession Crises in the Global Political System: A Test of the Transition Model,” in Bergesen, Albert, ed., Crises in the World System (Beverly Hills, Calif.: Sage, 1983)Google Scholar; and idem (fn. 2).

⁶ Bueno de Mesquita and Lalman (fn. 3, 1988, 1992); Houweling and Siccama (fn. 3, 1988, 1991); Kim (fn. 3, 1989, 1991); Kim and Morrow (fn. 3); Organski and Kugler (fn. 2).

⁷ Organski (fn. 2), 345–46.

⁸ Kugler, Jacek and Organski, A. F. K., “The Power Transition: A Retrospective and Prospective Evaluation,” in Midlarsky (fn. 4)Google Scholar; Organski (fn. 2); Organski and Kugler (fn. 2).

⁹ Organski also argues that once all states are industrially advanced, the circumstances would presumably once again preclude the possibility of transition struggles, since “great and sudden shifts in national power” would be less likely. Therefore, he suggests, we need a “new” theory for the future. See Organski (fn. 2).

¹⁰ Thucydides, , The Peloponnesian War, trans. Warner, Rex (Harmondsworth, England: Penguin Books, 1972)Google Scholar; Gilpin (fn. 2); idem, “The Theory of Hegemonic War,” Journal of Interdisciplinary History 18 (Spring 1988); Levy (fn. 4, 1981, 1983); idem, “Theories of General War,” World Politics 37 (April 1985); Modelski and Thompson (fn. 4, 1988, 1989); Thompson (fnn. 2, 5). Levy, Modelski and Thompson, Toynbee, and Wallerstein hold that the modern nation-state system emerged around 1500. Russett and Starr note otherwise, that many scholars date the modern nation-state from the Peace of Westphalia (1648). See Russett, Bruce M. and Starr, Harvey, World Politics: The Menu for Choice (San Francisco: W. H. Freeman, 1981)Google Scholar; Levy (fn. 4, 1981, 1983); Modelski and Thompson (fn. 4, 1988); Toynbee, Arnold J., A Study of History, vol. 9 (New York: Oxford University Press, 1954)Google Scholar; Wallerstein, Immanuel, The Politics of the World Economy (Cambridge: Cambridge University Press, 1984).Google Scholar

¹¹ Altfeld, Michael, “The Decision to Ally: A Theory and Test,” Western Political Quarterly 37 (December 1984)CrossRefGoogle Scholar; Iusi-Scarborough, Grace and Mesquita, Bruce Bueno de, “Threat and Alignment Behavior,” International Interaction 14 (1988)Google Scholar; Kim (fn. 3, 1991); Levy, Jack S., “Declining Power and the Preventive Motivation for War,” World Politics 40 (October 1987)CrossRefGoogle Scholar; Morrow, James D., “Alliances and Asymmetry: An Alternative to the Capability Aggregation Model of Alliances,” American Journal of Political Science 35 (November 1991)CrossRefGoogle Scholar; Benjamin A. Most and Harvey Starr, “International Relations Theory, Foreign Policy Substitutability, and ‘Nice’ Laws,” World Politics 36 (April 1984).

¹² Kim (fn. 3, 1991).

¹³ This argument might also be applied to cases of great power war in general, where the rapidly growing great power challenges the declining stronger rival (in those cases, the declining stronger rival is not the dominant power but a great power). The hegemonic war arguments seem to be special cases of power transitions in general. For the original arguments about power transition, power parity, and the rate of growth, see Organski (fn. 2); and Organski and Kugler (fn. 2).

¹⁴ The role of alliance is viewed differently in the revised power transition theory and in the balance-of-power theory. Balance-of-power theorists argue that alliances are made in response to a perceived threat and are indispensable to maintaining a balance of power and hence to preserving the peace. See, e.g., Guliek (fn. 1); Liska (fn. 1); and Walt, Stephen M., The Origins of Alliances (Ithaca, N.Y.: Cornell University Press, 1987).Google Scholar On the contrary, in the revised power transition framework, approximate equality of power for the challenger relative to that of the dominant nation, which results from such internal and external efforts of changes in power as internal growth and alliance formation, may be seen as a precursor of war.

¹⁵ Mesquita, Bruce Bueno de, “Pride of Place: The Origins of German Hegemony,” World Politics 43 (October 1990)CrossRefGoogle Scholar; Bueno de Mesquita and Lalman (fn. 3, 1992); Howard, Michael, The Causes of Wars (Cambridge: Harvard University Press, 1983), 18–20Google Scholar; Kennedy, Paul M., The Rise and Fall of the Great Powers (New York: Random House, 1987), chap. 5Google Scholar; Kugler and Organski (fn. 8); Organski (fn. 2); Organski and Kugler (fn. 2).

¹⁶ Organski (fn. 2), 366; emphasis in original.

¹⁷ The list of great powers in this analysis is the same as Modelski and Thompson's list of global powers. Modelski and Thompson emphasize naval capabilities: to be considered a great, or “global,” power in their terms, a nation-state must control a minimum of 10% of the global-reach naval capabilities and demonstrate a genuine oceanic scale of naval activities. For details about their definition of global power, see Modelski and Thompson (fn. 4, 1988), 97–99; or Thompson (fn. 2), 47–49. Modelski, and Thompson's list is somewhat different from Levy's list (fn. 4, 1983).Google Scholar Whereas Levy includes Austria-Hungary (1648–1815), Sweden (1648–1721), the Ottoman Empire (1648–1699), and Prussia (1740–1815) among the great powers, Modelski and Thompson do not, because those states did not have the necessary naval capabilities. I use Modelski and Thompson's definition of great power for two reasons. First, it is because of data availability. I develop a new national capability data set for the period from 1648 to 1815 using data based on army size, population, and sea power; sea power data for Austria, Sweden, the Ottoman Empire, and Prussia are not readily available. Second, Modelski and Thompson's definition is more globally oriented than Levy's, which is more Eurocentric.

¹⁸ Organski and Kugler (fn. 2).

¹⁹ Organski and Kugler (fn. 2) believe that for war to break out a long period of time is required to produce sufficient changes in the power distributions between possible adversaries. They argue for a period of approximately twenty years before each war.

²⁰ Disputes between the two groups of nations are decomposed into all possible dyads, a procedure that has both advantages and disadvantages. On the one hand, it increases the number of cases and avoids aggregating nations whose actions may be independent. On the other hand, it also isolates dependent decisions and increases the effects of random error in the measurement of the independent variables. For example, there are ten dyads in the period from 1800 to 1815. The behavior of some nations in those dyads was not independent across dyads. Cases that are not statistically independent can be problematic in statistical tests. The significance tests used here should be viewed as a heuristic device to suggest the relative strength of associations. Even given these problems, decomposing multilateral disputes is not novel. This procedure is comparable to the approaches of Houweling and Siccama (fn. 3, 1988, 1991); Kim (fn. 3, 1991); and Organski and Kugler (fn. 2). Statistical analyses of arms race disputes also often decompose multilateral disputes. See, e.g., Morrow, James D., “A Twist of Truth: A Reexamination of the Effects of Arms Race on the Occurrence of War,” journal of Conflict Resolution 33 (September 1989)CrossRefGoogle Scholar; and Wallace, Michael D., “Arms Races and Escalation: Some New Evidence,” Journal of Conflict Resolution 23 (March 1979).CrossRefGoogle Scholar

²¹ Organski and Kugler (fn. 2); Gilpin (fn. 2); and Modelski and Thompson (fn. 4, 1988). Gilpin speaks of hegemonic wars, Modelski and Thompson of global wars, and Organski and Kugler of major wars. For an excellent summary of differences in their definitions, see Levy (fn. 10, 1985).

²² Organski and Kugler (fn. 2).

²³ Unlike Gilpin and Modelski and Thompson, Organski and Kugler consider the Seven Years' War as a major war by their various criteria. During the Seven Years' War, first of all, great powers participated on each opposing side. Second, that war was as severe as previous wars: total battle deaths of the War of the League of Augsburg in nine years were 680,000, for an average of 75,555 dead per year; for the War of Spanish Succession the number over twelve years was 1,251,000, for an average annual level of death of 104,250; in the Seven Years' War total deaths in eight years were 992,000, averaging 124,000 per year. Thus, the severity of battle deaths of the Seven Years' War reached a higher level than those of the two previous wars, the War of the League of Augsburg and the War of Spanish Succession. This comparison is based on the data in Levy (fn. 4, 1983), 77–92.

²⁴ Levy includes the War of Austrian Succession (1739–48) in his list of the general wars, but it is not included in any of the three war lists in Table 3. Organski and Kugler do not indude the War of Austrian Succession, however, because by their criteria it was not as severe as earlier wars, the War of the League of Augsburg or the War of Spanish Succession. Total battle deaths for the nine years of the War of the Austrian Succession were 359,000, averaging 39,888 per year. See Levy, (fn. 4, 1983), 77–92Google Scholar; and idem (fn. 10).

²⁵ Bruce Bueno Mesquita, de, The War Trap (New Haven: Yale University Press, 1981)Google Scholar; Gilpin (fn. 2); Organski and Kugler (fn. 2); Singer, Bremer, and Stuckey (fn. 1); Taylor, A. J. P., The Struggle for Mastery in Europe, 1848–1918 (Oxford: Oxford University Press, 1954).Google Scholar

²⁶ See Mahan, Alfred, The Influence of Sea Power upon History, 1660–1783 (New York: Hill and Wang, 1890)Google Scholar; and Modelski and Thompson (fn. 4, 1988, 1989). Military, economic, and demographic indicators are important components of estimates of national capabilities. My capability index does not include direct economic indicators. This could be a shortcoming, although naval power was probably highly correlated with wealth. Although the current available data allow the construction of only a crude indicator, I hope that this measurement will soon be modified to include some economic indicators. The new index proposed here is the only composite national capability index for the pre-1816 period and at least is more comprehensive than Modelski and Thompson's naval capability data alone. Their sea power data are the only capability data used in other empirical analyses for the pre-1815 period.

²⁷ Sorokin, Pitirim A., Social and Cultural Dynamics (New York: American Book Company, 1937), 3:543–77.Google Scholar These rough estimates of army size are compared with other historical sources. See, e.g., Dupuy, Ernest R. and Dupuy, Trevor N., The Encyclopedia of Military History: From 3500 B.C. to the Present, 2d rev. ed. (New York: Harper and Row, 1977)Google Scholar, Langer, William L., An Encyclopedia of World History (Boston: Houghton Mifflin, 1948).Google Scholar I am in the process of developing more reliable data for army size by including other available sources as well. These sources include Kennedy (fn. 15); Parker, Geoffrey, Europe in Crisis, 1598–1648 (Brighton: Harvester Press, 1980)Google Scholar; Rasier, Karen A. and Thompson, William R., War and State Malting: The Shaping of the Global Powers (Boston: Unwin Hyman, 1989)Google Scholar; Ropp, Theodore, War in the Modern World (Durham, N.C.: Duke University Press, 1959).Google Scholar The data for army size based on Sorokin's work do not seem to be very different from the newly revised data based on the sources listed above. For example, I have revised the national capability measure for France and compared it with the French capability measure used in this analysis. The correlation coefficient for these two is very high (r=.97).

²⁸ Babuscio, Jack and Dunn, Richard M., European Political Facts, 1648–1789 (New York: Facts on File, 1984)CrossRefGoogle Scholar; Cook, Cris and Paxton, John, European Political Facts, 1789–1848 (New York: Facts on File, 1981)CrossRefGoogle Scholar; McEvedy, Colin and Jones, Richard, Atlas of World Population History (Middlesex, England: Penguin Books, 1978).Google Scholar

²⁹ Modelski and Thompson (fn. 4, 1988).

³⁰ The same weighing procedure is used in the Correlates of War (cow) project's national capability index for 1816 to 1970. This index and that of the cow project differ, as follows: whereas mine is based on three indicators, the cow index is based on six indicators (army size, military expenditure, energy consumption, iron/steel production, total population, and urban population). For details about the cow index, see Singer, Bremer, and Stuckey (fn. 1).

³¹ See David Singer, J. and Small, Melvin, “Formal Alliances, 1815–1939: A Quantitative Description,” Journal of Peace Research 1 (1966).Google Scholar Singer and Small focus on formal military alliances and classify them into three categories. These are defense pacts, in which the signatories agree to come to each other's mutual defense in case any one of them is attacked; neutrality or nonaggression pacts, in which the signatories agree not to declare war against each other in the event that a third nation declares war against one of them; and entente, in which the signatories agree to consult each other about possible coordinated action in the event one of them is attacked by a third party.

³² Alliances identified by any of the three leading historical sources (Albrecht-Carrié, Dupuy and Dupuy, and Langer) are included in this alliance data set. However, most of the alliances mentioned by one source are also identified by the other two and/or by other important sources. Albrecht-Carrié, René, The New Encyclopedia Britannica, 15th ed., s.v. “European diplomacy and wars, c. 1500–1914”Google Scholar; Dupuy and Dupuy (fn. 27); Langer (fn. 27). Other important sources include Babuscio and Dunn (fn. 28); Cook and Paxton (fn. 28); Levy(fn. 4, 1981).

³³ Albrecht-Carrié (fn. 32); Babuscio and Dunn (fn. 28); Dupuy and Dupuy (fn. 27); Langer (fn. 27). In a subsequent paper, I explain in more detail how these new national capability and alliance data are developed and provide these raw data.

³⁴ The “utility index” comes from Bueno de Mesquita's expected utility research program. It is a measure of the degree to which the policies pursued by two nations are congruent and is based on the military alliance patterns of nations, which may be considered the single most important foreign policy dimension. Bueno de Mesquita uses Singer and Small's formal alliance data to create his utility index; that is, his utility index, U_ki or U_kj, for the period from 1816 to 1970 is based on four categories (defense pacts, neutrality or nonaggression pact, entente, and no alliance). With these four categories, a four-by-four contingency table for every possible dyad in every year from 1816 to 1975 is constructed; the similarity of alliance portfolios (or the utility index) is measured by calculating Kendall's tau-b score for each four-by-four table. A more detailed description of the utility index can be found in Bueno de Mesquita (fn. 25). As described above, when calculating a utility index for the period from 1648 to 1815, I do not distinguish three types of alliances. Instead, I use only the two categories alliance and no alliance.

³⁵ Sabrosky, Alan N., “Interstate Alliances: Their Reliability and the Expansion of War,” in David Singer, J., ed., The Correlates of War: II (New York: Free Press, 1980).Google Scholar

³⁶ In this paper, therefore, “allies” does not necessarily mean nations that have a bilateral formal treaty. Allies refers to nations that have shared values (or a common alliance pattern) even if they are not formally allied with each other.

³⁷ Altfeld, Michael and Mesquita, Bruce Bueno de, “Choosing Sides in Wars,” International Studies Quarterly 23 (March 1979).CrossRefGoogle Scholar

³⁸ For more details about the concept and measure of dissatisfaction, see Organski (fn. 2); and Kim (fn. 3, 1991).

³⁹ Kim (fn. 3, 1991).

⁴⁰ Kim (fn. 3, 1989, 1991); Kim and Morrow (fn. 3); Kugler and Organski (fn. 8); Organski and Kugler (fn. 2).

⁴¹ Kim (fn. 3, 1991); Kim and Morrow (fn. 3).

⁴² Bueno de Mesquita and Lalman suggest that it is not dissatisfaction but the “large stake” of a system-transforming dispute that is an important factor influencing the probability of major war. See Bueno de Mesquita and Lalman (fn. 3, 1992), chap. 6.

⁴³ By changing the value of the independent variable from 0.15 to 1.00, we get an approximately 0.34 increase in the probability of war, holding other variables constant: p = 1/(1 + e^−1.0*3.139) − 1/(1 + e ^{−0.15*3.139}) = 0.9585 − 0.6156 = 0.3429. The relative adjusted national capability ratio = 0 is impossible because the value 0 means that the adjusted capability score of the weaker great power is 0. The minimum value of relative power ratio in this data is about 0.15 and the maximum value is 1.00. So we can change the value of the alliance equality variable from 0.15 to 1.00 and obtain about 0.34 increase in the probability of war, holding other variables constant.

⁴⁴ Kim (fn. 3, 1989, 1991); Kim and Morrow (fn. 3).

⁴⁵ Thompson (fn. 5).

⁴⁶ Houweling and Siccama (fn. 3, 1988, 1991); Kim (fn. 3, 1989, 1991); Kim and Morrow (fn. 3); Organski and Kugler (fn. 2).

⁴⁷ Organski and Kugler (fn. 2), 63.

⁴⁸ Modelski's long-cycle phases of “world power, deconcentration, delegitimation, and global war” also suggest the deterministic view that cataclysmic wars are out of human control. Although Gilpin notes that “peaceful adjustment of the systemic disequilibrium” is not impossible, he also seems to argue that hegemonic wars are not easily avoidable. For more details, see Gilpin (fn. 2); Modelski (fn. 2); Thompson (fn. 2).