Military Resources and International War

A Statistical Theory of Interconnected Conflict

by Jeffrey Alan Carnegie (Author)
©2021 Monographs XVI, 178 Pages


Do leaders make war decisions individually or do they consider other ongoing conflicts? Most researchers assume dyadic independence out of convenience. In this book, Jeffrey Alan Carnegie presents evidence that this is a faulty assumption. First, he develops a formal theory in which limited military resources act as a constraint on the ability of leaders to engage in international conflicts. Country leaders plan accordingly by considering the entire picture of conflicts. This theory, in contrast to dyadic dependence, implies a different data-generating process for international conflicts. Second, he tests both theories using summary statistics, network analysis, and logistic regression. All three methods show strong support for military resource division theory. Carnegie further shows that the dependent effects change with time, even when controlling for polarity. The effects also differ between regions, which implies cultural factors at work. Third, he suggests for the future that researchers use multiple methods to account for different types of dependencies, because no single method can address them all. He shows how to make the best use of logistic regression and social network analysis for conflict statistics. He offers suggestions to policy makers for how best to incorporate this theory in analysis. Finally, Carnegie concludes by comparing predictions of the two theories regarding conflicts for the United States, especially Iran and North Korea. This book will be of interest to conflict researchers in academia and government who want to better understand the effect of coalitions on modern warfare.

Table Of Contents

  • Cover
  • Title
  • Copyright
  • Dedication
  • About the author
  • About the book
  • This eBook can be cited
  • Table of Contents
  • List of Tables
  • List of Figures
  • Acknowledgments
  • List of Abbreviations
  • Chapter One: Introduction
  • 1.1 A Brief History of Multiple Simultaneous Conflicts
  • 1.2 Competing Theories
  • 1.3 Methodology and Empirical Results
  • 1.4 TL; DR
  • Chapter Two: Theories of Conflict, with and without Dyadic Independence
  • 2.1 Butter, Guns, and Decisions
  • 2.2 The Path to War
  • 2.3 Hypotheses
  • Chapter Three: Empirical Evidence
  • 3.1 Risk Analysis and Statistical Equations
  • 3.2 Cursory Evidence
  • 3.3 Network Analysis
  • 3.4 Logistic Regression
  • 3.5 Summary of Main Evidence
  • Chapter Four: Temporal Differences
  • 4.1 Evidence for Behavioral Changes over Time
  • 4.2 Network Dependency Changes over Time
  • 4.3 Discussion of Temporal Dynamics
  • Chapter Five: Regional Differences
  • 5.1 Regional Differences in Network Dependencies
  • 5.2 Discussion of Regional Differences
  • Chapter Six: Conclusion
  • 6.1 Comparison of Methods
  • 6.2 North Korea and Iran
  • 6.3 Future Research
  • Appendix One: References
  • Appendix Two: Formal Theory
  • Appendix Three: Data and Variables
  • Appendix Four: Full Data Tables
  • Index

←viii | ix→

Table 1.1: Actual number of state-years by capability and degree.

Table 1.2: Actual number of dyad-years by capability of attacker and defender.

Table 1.3: Actual number of state-years by number of initiated and received conflicts.

Table 1.4: Number of conflicts and proportion initiated for all years.

Table 3.1: Density and expected number of 3-cycles and 4-cycles.

Table 3.2: BTERGM results.

Table 3.3: Logit results for peace using capability proportion.

Table 3.4: Logit results for escalation using capability proportion.

Table 3.5: Logit results for threat using capability proportion.

Table 4.1: Time period BTERGM models excluding World Wars.

Table 4.2: Time period BTERGM models for World Wars.

Table 4.3: Time and polarity models of peace.

Table 4.4: Time and polarity models of escalation.

Table 4.5: Time and polarity models of threat.

Table 5.1: Regional models of Australia and Oceania.

Table 5.2: Regional models of Sub-Saharan Africa.

Table 5.3: Regional models of East and South Asia.

Table 5.4: Regional models of Latin America.

←ix | x→

Table 5.5: Regional models of the Middle East and North Africa.

Table 5.6: Regional models of Europe (including the United States and Canada).

Table 6.1: Comparison of logit and BTERGM.

Table A4.1: Abbreviated logit results for peace using capability ratio.

Table A4.2: Logit results using capability proportion, model Peace CP 3 (details).

Table A4.3: Logit results using capability ratio, model Peace CR 4 (details).

Table A4.4: Abbreviated logit results for escalation using capability ratio.

Table A4.5: Logit results using capability proportion, model War CP 2 (details).

Table A4.6: Logit results using capability ratio, model War CR 4 (details).

Table A4.7: Abbreviated logit results for threat using capability ratio.


XVI, 178
ISBN (Hardcover)
Publication date
2021 (March)
New York, Bern, Berlin, Bruxelles, Oxford, Wien, 2021. XVI, 178 pp., 23 b/w ill., 30 tables.

Biographical notes

Jeffrey Alan Carnegie (Author)

Jeffrey Alan Carnegie (PhD in politics, New York University) is a Teaching Assistant Professor at Montana State University in Bozeman, Montana.


Title: Military Resources and International War