Alliance Behavior in Balance of Power Systems: Applying a Poisson Model to Nineteenth-Century Europe

McGowan, Patrick J.; Rood, Robert M.

doi:10.2307/1958395

Cited by 49 publications

(20 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The third category of statistical methods used in international relations are Poisson process models (see Richardson, 1944;McGowan and Rood, 1975). These methods are more closely applicable to the special nature of event count data, but they have been used in only very simple ways.…”

Section: Empirical Methods In International Relationsmentioning

confidence: 99%

“…For example, in part of their analysis McGowan and Rood (1975) use one period of one hundred years to study the pattern of alliance formation and its fit to a Poisson distribution. Cramer's theorem essentially allows them to partial out and then ignore variation in the expected number of events over the years.…”

Section: Empirical Methods In International Relationsmentioning

confidence: 99%

“…These help display the essential features of the data and methods but do not attempt to address every sophisticated substantive argument in the literature. For a sampling of studies on intekmtional alliances, see singer and Small (1968) Wallace (1973), McGowan and Rood (1975), Siverson and Tennefoss (1984) McDonald and Rosecrance (1985), and Walt (1985).…”

Section: Hurdle Modes Of Instability and International Alliancesmentioning

confidence: 99%

See 2 more Smart Citations

Event Count Models for International Relations: Generalizations and Applications

King¹

1989

International Studies Quarterly

175

View full text Add to dashboard Cite

International relations theorists tend to think in terms of continuous processes. Yet we observe only discrete events, such as wars or alliances, and summarize them in terms of the frequency of occurrence. As such, most empirical analyses in international relations are based on event count variables. Unfortunately, analysts have generally relied on statistical techniques that were designed for continuous data. This mismatch between theory and method has caused bias, inefficiency, and numerous inconsistencies in both theoretical arguments and empirical findings throughout the literature. This article develops a much more powerful approach to modeling and statistical analysis based explicitly on estimating continuous processes from observed event counts. T o demonstrate this class of models, I present several new statistical techniques developed for and applied to different areas of international relations. These include the influence of international alliances on the outbreak of war, the contagious process of multilateral economic sanctions, and reciprocity in superpower conflict. I also show how one can extract considerably more information from existing data and relate substantive theory to empirical analyses more explicitly with this approach. IntroductionWhereas most theories in the social sciences attempt to explain underlying continuous processes, we generally observe only finite numbers of discrete events. Bertram M. Gross writes: "[Tlhe world or my part of it is seen as an ongoing stream of events in time . . . Facts and process are separated into discrete elements only by human analysis . . . Change-whether rapid or slow, hidden or open-is continuous" (Gross, 1968:262). For example, influence among political actors, the continuing allocation of resources, constituency representation, and other aspects of politics can all be described as unobserved continuous processes that generate observed discrete events. A legislator probably represents constituents in varying degrees continually in all aspects of his or her work, but most observers cannot record much more than roll call votes and compare them to occasional polls of constituent opinion. The constant trade flows between nations are important features of economic cooperation, but an Authur'c S u t p : h1) thanks to,Jim Alt, Bill Brunclage, Kancy Burns,,John Freeman, Kelsuke Iicla, Lisa Martin, and Bob Powell for helpful comments and discussions, and to Lisa Martin for her data on economic sanctions. An earlier \ersion of this paper. was presented on a panel at the 1988 Midwest Political Science ..\ssociation meetings; I appreciate the suggestions from participants at that panel, particularl) J o h n hldl-ich and Dan \Vood. Thanhs also to the National Science Foundation for research support (KSF #SES-87-22715). 1989 International Studies Association 124Event Count L\40dels fo7 Inte7natzoizal Relatzons analyst might only have a list of major treaties or quarterly summaries of economic activities. In the U.S., presidential-congressional relations continuou...

show abstract

Section: Empirical Methods In International Relationsmentioning

confidence: 99%

Section: Empirical Methods In International Relationsmentioning

confidence: 99%

Section: Hurdle Modes Of Instability and International Alliancesmentioning

confidence: 99%

See 1 more Smart Citation

Event Count Models for International Relations: Generalizations and Applications

King¹

1989

International Studies Quarterly

175

View full text Add to dashboard Cite

show abstract

“…The domain for each observation may be a cross-section, a time interval, or even the cell of a contingency table. The number of visible uses of military force initiated by the United States in each six-month interval (Stoll, 1984), the number of presidential vetoes per year (Rohde and Simon, 1985), the frequency of formal and informal military alliances (Russett, 1971;McGowan and Rood, 1975), and the annual number of presidential appointments to the Supreme Court (King, 1987) are examples of time series counts. Examples of cross-sectional event count studies include the number of coups d'etat in each black African state (Johnson, Slater, and McGowan, 1984) and the number of political activities engaged in and reported by Soviet emigres (Di Franceisco and Gitelman, 1984).l *My thanks to Chris Achen, Jim Alt, Neal Beck.…”

Section: Introductionmentioning

confidence: 99%

Variance Specification in Event Count Models: From Restrictive Assumptions to a Generalized Estimator

King¹

1989

American Journal of Political Science

253

145

View full text Add to dashboard Cite

This paper discusses the problem of variance specification in models for event count data. Event counts are dependent variables that can take on only nonnegative integer values, such as the number of wars or coups d'etat in a year. I discuss several generalizations of the Poisson regression model, presented in King (1988), to allow for substantively interesting stochastic processes that do not fit into the Poisson framework. Individual models that cope with, and help analyze, heterogeneity, contagion. and negative contagion are each shown to lead to specific statistical models for event count data. In addition. I derive a new generalized event count (GEC) model that enables researchers to extract significant amounts of new information from existing data by estimating features of these unobserved substantive processes. Applications of this model to congressional challenges of presidential vetoes and superpower conflict demonstrate the dramatic advantages of this approach.

show abstract

“…Virtually all authors (Coleman, 1964;Feller, 1968;Hayes, 1973;McGowen & Rood, 1975;Midlarsky, 1970) who make use of any of these distributions make use of the x2 distribution as a test of fit. In all sets of occurrences within this article, the Poisson and either the contagious Poisson or the contagious binomial are employed.…”

Section: F+;mentioning

confidence: 99%

Epidemiology of domestic military intervention: Evaluation of contagion as an explanatory concept

Laemmle¹

1977

Syst. Res.

View full text Add to dashboard Cite

One of the most perplexing questions in politics is the effect of the international or macrosystem environment upon the activity of either regional subsystems or the basic components of the macropolitical system, the individual national political systems. One of the behavioral areas where it is crucial to detect the effect of the macrosystem is in the area of military policy, particularly the action taken by national military forces against host or governing institutional structures. This type of action, known as domestic military intervention, DMI, has come under increased scrutiny by behavioral and social scientists.This article subjects one explanation of DMI, the international contagion of similar events, to a critical examination. Contagion regards DMI as the result of macrosystemic effects upon the subsystems. Utilizing data on successful DMI events from 1948-67, tests for the interdependence of DMl events were conducted utilizing the Poisson distribution, contagious Poisson distribution, and contagious binomial and tests for autocorrelation.Some evidence for international contagion was found within various regional subsystems, but the utility of contagion of DMI as a pure macrosystemic phenomena was found to be severely constrained. F+;,HENEVER one seeks to explain the w behavior of the components of any system, two approaches for analysis are available. One is to search for macrosystemic regularities, the other is to emphasize the unique characteristics of each subsystem. The choice for strategy of analysis is in part determined by the type of problem area under investigation. The literature on organizational merger illustrates the problems of the appropriate level of analysis (Pfeffer, 1972). Do industrial mergers result from market or macrosystem effects, industry or subsystem characteristics, or firm specific or component attributes? Should macrosystem effects not be found to be present, investigation at subsystem or component levels is appropriate.Similar arguments about appropriate levels of analysis pervade discussions of political violence. Are violent actions the result of solely individual characteristics or are they the result of the influence of KEY WORDS: domestic military intervention, contagion, cross-national military contagion, Poisson process. the social environment in which the actor exists? Berkowitz (1973) speaks of the imitative effect of societal levels of violence as a determinant of individual violent activity. Bandura (cited in Berkowitz, 1973) relates the incidence of airline hijacking to the effects of the international occurrence of similar events. The same arguments concerning proper level of analysis permeate discussion of a particular form of political violence, domestic military intervention, DMI. DMI is said to have occurred when military forces choose to assume control over their host governing apparatus. Is DMI the result of intracomponent events, characteristics of military organizations, or other unique characteristics? Is DMI the result of cue taking from the macros...

show abstract

Alliance Behavior in Balance of Power Systems: Applying a Poisson Model to Nineteenth-Century Europe

Cited by 49 publications

References 19 publications

Event Count Models for International Relations: Generalizations and Applications

Event Count Models for International Relations: Generalizations and Applications

Variance Specification in Event Count Models: From Restrictive Assumptions to a Generalized Estimator

Epidemiology of domestic military intervention: Evaluation of contagion as an explanatory concept

Contact Info

Product

Resources

About