Cover Page

Contents

Guides to International Studies

The series Guides to International Studies builds on the expertise that exists in the various substantively organized sections of the International Studies Association (ISA), the largest professional association of scholars of international studies. ­The series emerged to solve a need for detailed volumes linked to the thematic sub-­sections of the ISA; each building on the work amassed by the ISA’s International Studies Compendium Project, and yet offering faculty, researchers, ­professionals, and students valuable stand-alone surveys. Each volume is carefully organized and reflects the latest developments and research within its field.

Image

Acknowledgments

We are grateful to scholars from the Scientific Study of International Processes section of the International Studies Association for contributing their time to this important project. We thank Ashley Leeds, Robert Denemark, and Thomas Volgy for encouraging us to put this edited volume together. These chapters will be subject to external review and we express our gratitude to the scholars who served as initial reviewers and assisted in ensuring that the research here was of the highest quality. Finally, we would like to thank Mark Nieman for his research assistance with the project.

About the Contributors

D. Scott Bennett is Distinguished Professor of Political Science at the Pennsylvania State University, and Associate Director of the Correlates of War Project. He is the author or co-author of numerous articles about international conflict, and is the co-author of the book, The Behavioral Origins of War (2004). He is also the author of software used by researchers analyzing international conflict data. His research interests include using statistical and simulation models to study international conflict, war termination, insurgency, and the effects of democracy on war. At Penn State, he teaches courses on general international relations, international conflict, and conflict resolution.

Steve Chan is Professor of Political Science at the University of Colorado at Boulder. He was the recipient of the Karl W. Deutsch award in 1988, Boulder Faculty Assembly award for Excellence in Research in 1994, and CU Parents Association’s Marinus Smith Award in 2004. His research interests cover international relations, political economy, foreign policy, decision making, and East Asia. His work has appeared in journals such as the American Political Science Review, Comparative Political Studies, International Interactions, International Studies Quarterly, Journal of Conflict Resolution, Journal of Peace Research, Security Studies, and World Politics.

Paul F. Diehl is Henning Larsen Professor of Political Science at the University of Illinois at Urbana-Champaign. He also serves as Director of the Correlates of War Project, the largest data collection effort on international conflict in the world and is Founding Director Emeritus of the College of Liberal Arts and Sciences Teaching Academy. He received his PhD in Political Science at the University of Michigan in 1983 and has held faculty positions at the University of Georgia and SUNY-Albany. He has published over 20 books including The Dynamics of International Law (2010) and War and Peace in International Rivalry (2000) and he is the author of over 100 articles on international security matters. He is the recipient of numerous grants and awards including those from the National Science Foundation, United States Institute of Peace, and the Lilly Foundation. He was the 1998 recipient of the Karl Deutsch Award given by the International Studies Association to the leading young scholar on peace and conflict issues. He also received the LAS Dean’s Award for Excellence in Undergraduate Teaching and the University of Illinois Campus Award for Excellence in Undergraduate Teaching, as well as being a four time winner of the Clarence Berdahl Award for Excellence in Undergraduate Instruction. He is past President of the Peace Science Society (International) and Vice-President Elect of the International Studies Association. His areas of expertise include the causes of war, UN peacekeeping, and international law.

Caroline A. Hartzell is a professor of Political Science at Gettysburg College. Her work on peace durability has appeared in American Journal of Political Science, Conflict Management and Peace Science, International Organization, and Journal of Conflict Resolution.

Paul R. Hensel (PhD, University of Illinois, 1996) is Professor of Political Science at the University of North Texas. His research concerns international conflict and conflict management, and has recently emphasized territorial claims and the management of cross-border rivers. He has published articles in the American Journal of Political Science, Conflict Management and Peace Science, GeoJournal, International Negotiation, International Organization, International Studies Quarterly, Journal of Conflict Resolution, Journal of Peace Research, and Political Geography.

Paul K. Huth (PhD, Yale University, 1986) is Professor of Government and Politics and Research Director of CIDCM at the University of Maryland. He is the author of Extended Deterrence and the Prevention of War (1988), Standing Your Ground (1996), and (with Todd Allee) The Democratic Peace and Territorial Conflict in the Twentieth Century (2002). He is currently working on a new book, Casualties of War.

Kelly M. Kadera is Associate Professor at the University of Iowa. She is the Co-Organizer of Journeys in World Politics and former Co-Program Chair of the International Studies Association. Her research uses dynamic models to understand international conflict. She is currently investigating potential biological foundations of interstate conflict (with Sung Woo Kim and Jonathan Ring) and the nonlinearities of the relationship between power distributions and militarized disputes (with Dan Morey).

Brett Ashley Leeds is Associate Professor of Political Science at Rice University. Her research focuses on the design and influence of cooperative agreements and institutions, particularly in the area of international security. She has written extensively on the politics of military alliances and on the influence of domestic politics on international relations.

Zeev Maoz is Distinguished Professor of Political Science at the University of California – Davis, as well as Distinguished Fellow at the Interdisciplinary Center, Herzliya, Israel. He is the former President of the Peace Science Society (International) (2007–8). He is the author and editor of 12 books in international relations and his scholarly articles have appeared in leading journals such as the American Political Science Review, American Journal of Political Science, World Politics, Journal of Conflict Resolution, International Security, International Studies Quarterly and others. His fields of specialization include strategy and international security affairs, international ­relations theory, decision making and negotiations, and political methodology. In the field of Middle East politics, his work concentrates on the international relations of the region. His work in this area includes editing and contributing to two special issues of the Journal of Strategic Studies on regional security in the Middle East. He is also one of the first Israelis to have published a major article in the Egyptian daily ­newspaper, Al Ahram.

Sara McLaughlin Mitchell is Professor of Political Science and Collegiate Scholar at the University of Iowa. She is Co-Director of the Issue Correlates of War Project, an Associate Editor of Foreign Policy Analysis, and co-founder of the Journeys in World Politics ­workshop. She is co-author of Domestic Law Goes Global: Legal Traditions and International Courts, she has edited three special journal issues, and she has published more than two dozen ­journal articles and book chapters. She is the recipient of four major research awards from the National Science Foundation, as well as numerous research grants from the University of Iowa and Florida State University. Her areas of expertise include inter­national conflict, democratic peace, international organizations, diversionary theory, international courts, conflict management, territorial, maritime, and river issues, and time series analysis.

Will H. Moore is a professor of political science at Florida State University. His interests include violent political conflict, human rights, dissent, repression, and foreign policy behavior, and his research can be found in American Journal of Political Science, International Organization, International Studies Quarterly, and Journal of Conflict Resolution, among other outlets.

T. Clifton Morgan is the Albert Thomas Professor of Political Science at Rice University. His current research focuses on the development of a general theory of foreign policy and on the use of economic sanctions as an instrument of foreign policy. He has also published work on bargaining in international crises and on the influence of domestic politics on international conflict.

James D. Morrow is Professor of Political Science at the University of Michigan. Professor Morrow’s research addresses theories of international politics, both the ­logical development and empirical testing of such theories. He is best known for ­pioneering the application of noncooperative game theory, drawn from economics, to international politics. His published work covers crisis bargaining, the causes of war, military alliances, arms races, power transition theory, links between international trade and conflict, the role of international institutions, and domestic politics and ­foreign policy.

Alyssa K. Prorok is a PhD candidate in the Government and Politics Department at the University of Maryland. Her current research projects focus on internal political violence, the negotiating behavior of combatants during civil war, and the link between international disputes and support for violent non-state actors. She has presented research on these topics at the APSA, MPSA, ISA, and Peace Science annual conferences.

Karen Rasler is Professor of Political Science at Indiana University. Her research interests are in general theories of international conflict and cooperation; relative decline of world powers; war and statebuilding processes; societal consequences of war; modeling long cycles of war; distribution of power and technological innovations; political violence and internal wars. She has taught a core seminar in international relations, American foreign policy, political violence and revolutions, introduction to political inquiry (research design). She has published journal articles in American Political Science Review, American Journal of Political Science, American Sociological Review, International Studies Quarterly, Journal of Conflict Resolution, Journal of Peace Research and International Interactions.

Idean Salehyan is an associate professor of political science at the University of North Texas and Co-Director of the Social Conflict in Africa Database Project. His research interests include civil and international conflict, international migration, and politics of natural disasters. He is the author of Rebels Without Borders: Transnational Insurgencies in World Politics (2009).

Branislav L. Slantchev is Associate Professor of Political Science at the University of California – San Diego. He studies crisis bargaining, the conduct and termination of wars, and militarized coercion. His work has appeared in the American Political Science Review, American Journal of Political Science, International Studies Quarterly, Security Studies, and Political Research Quarterly.

Ahmer Tarar is an associate professor of political science at Texas A&M University. His current research focuses on crisis bargaining and the causes of war. His articles appear in American Journal of Political Science, American Political Science Review, International Studies Quarterly, and Journal of Conflict Resolution.

William R. Thompson is Distinguished Professor and Donald A. Rogers Professor of Political Science at Indiana University. His teaching interests focus on international relations theory, conflict processes and international political economy. His research interests are similar with a current emphasis on long-term historical-structural change, the rise and fall of major powers, long economic waves and their consequences, and war impacts. A number of articles, monographs, and book chapters have been published on such topics as regional subsystems, military coups, alliance processes, war rivalries, and long waves of economic growth.

Clayton L. Thyne is an Assistant Professor of Political Science at the University of Kentucky. His research focuses on domestic conflict/instability, regime types and democratization, and international education policy. He is the author of How International Relations Affect Civil Conflict: Cheap Signals, Costly Consequences (2009). He has also published articles in journals such as International Studies Quarterly, Conflict Management and Peace Science, Journal of Peace Research, Comparative Political Studies, and Journal of Conflict Resolution.

Jaroslav Tir is an Associate Professor in the Department of Political Science at the University of Colorado, Boulder. He is the author of Redrawing the Map to Promote Peace: Territorial Dispute Management via Territorial Changes (2006) and of articles published in the American Journal of Political Science, Journal of Politics, International Studies Quarterly, Journal of Peace Research, Journal of Conflict Resolution, Political Geography, International Interactions, and Conflict Management and Peace Science. His current research interests include territorial and ethnic conflict management, diversionary theory of war, and international river-sharing issues.

John A. Vasquez is the Thomas B. Mackie Scholar in International Relations at the University of Illinois at Urbana-Champaign. His books include Territory, War, and Peace (with Marie T. Henehan) (2010), The War Puzzle Revisited (2009), and The Steps to War (with Paul D. Senese) (2008), He has published articles in the American Political Science Review, International Studies Quarterly, World Politics, Journal of Peace Research, British Journal of Political Science, among others. He has been President of the International Studies Association and the Peace Science Society (International).

Amy Yuen is an assistant professor of Political Science at Middlebury College. Her research examines conflict processes and intervention in civil and interstate wars. Her work appears in International Organization, Political Analysis and Journal of Conflict Resolution.

Frank C. Zagare is UB Distinguished Professor of Political Science at the University at Buffalo. A former Vice President of the International Studies Association, he is the author, co-author, or editor of six books including The Games of July: Explaining the Great War (2010), Perfect Deterrence (with D. Marc Kilgour) (2000), The Dynamics of Deterrence (1987), and Game Theory: Concepts and Applications (1984). In 2005 he received the Susan Strange Award from the International Studies Association.

Dina Zinnes is Merriam Professor Emerita at the University of Illinois, Urbana-Champaign. She was a founder and first Program Chair of the Scientific Study of International Processes section and President of the International Studies Association. Her research helped introduce formal models to the field of international relations. She is currently writing a book geared toward teaching the art of modeling.

Editors’ Introduction

Sara McLaughlin Mitchell, Paul F. Diehl, and James D. Morrow

The Scientific Study of International Processes (SSIP) is a robust group of scholars dedicated to its name. One of the oldest sections in the International Studies Association, SSIP focuses on the rigorous analysis of arguments and evidence. The SSIP community broadly falls into two camps: one primarily addresses the collection and statistical analysis of data, while the second uses mathematical models to elaborate theories of international processes. Neither milieu is exclusive; some members of each group do research that crosses into the other, and members of both research groups examine evidence and arguments outside of the other camp. The common project of the SSIP community brings rigor to the logical structure of theories and the assessment of evidence for and against those theories. This rigor aims at making the creation and analysis of data and the elaboration of theory more visible to the ­scientific community.

Most research in SSIP focuses on questions of security and conflict, such as why wars occur, which conflict management strategies are most successful, and the consequences of conflicts for future interstate interactions. Historically, those questions drove early efforts at data collection and analysis and model building, including the development of the Correlates of War (COW) Project. As both the accumulation of data sets and the range and sophistication of models have increased, new research has addressed questions from other areas of international studies, such as political economy, international cooperation, and human rights. Recent work on conflict has also focused more explicitly on explaining intrastate violence such as civil wars and protests as well as violence conducted by non-state actors such as transnational terrorist attacks.

We organize this edited volume using the two major research approaches in the SSIP tradition. The first set of chapters in Section I examines methodological issues and approaches in SSIP research. These chapters help the reader understand what ­methods are used and why they were adopted to answer certain questions. In doing so, the chapters provide a gateway into understanding those methods. Kadera and Zinnes provide a historical overview of the creation of the SSIP section, describing its evolution over the past several decades. Zagare and Slantchev focus on the use of game theory to study conflict processes, noting in particular the use of zero-sum and nonzero-sum models and summarizing the major debates that game theory has helped to settle in conflict studies. Hensel summarizes the major data collection efforts that have been undertaken by the SSIP community including the data sets on interstate and intrastate conflict, crises, rivalry, territorial change, regime type, human rights behavior, and contentious issues. Bennett discusses the benefits and challenges to teaching the SSIP approach to undergraduate students describing the importance of generalizability, the quality and depth of evidence for theoretical claims, and the use of different ­pedagogical tools in the classroom. Finally, Morrow’s chapter identifies some ­challenges that arise when integrating the two broad approaches of the SSIP tradition and discusses recent approaches such as the Empirical Implications of Theoretical Models (EITM) approach that seek to integrate the traditions.

The second set of chapters, in Section II, examines substantive topics of research that have played an important role in the development of the SSIP approach to understanding conflict and cooperation. Rasler and Thompson review several systemic theories of interstate conflict including balance of power theory, power transition theory, long cycle theory, and world systems theory respectively. Tir and Vasquez examine how contiguity and territorial disputes have increased the risks for interstate militarized disputes and wars. They also describe territorial conflicts inside the state, such as partitions and secessionist movements, and how these domestic conflicts relate to interstate disputes. Leeds and Morgan analyze the literature on arms races and alliances as potential forces for conflict or peace through deterrence, starting from the traditional balance of power and power transition approaches. Chan provides an overview of the democratic peace literature from Kant’s work on “Perpetual Peace” to modern ­analyses of the Kantian tripod for peace and potential threats to the democratic peace (e.g., the dangers of democratization).

Moore and Tarar discuss how domestic factors influence interstate conflict processes including the relationship between civil conflict and interstate conflict, the diversionary theory of war, and the way in which domestic institutional characteristics influence interstate bargaining. Salehyan and Thyne review the work on civil war onset, duration, and outcomes, concentrating on such factors as greed, grievance, geography, and international influences. Prorok and Huth focus on the expansion and diffusion of war, the conduct of war such as respect for the laws of warfare or civilian targeting, and factors that influence the duration and outcomes of interstate and civil wars. Hartzell and Yuen appraise recent research on the durability of peace following interstate and civil wars describing how the characteristics of conflicts and belligerents and the nature of third party conflict management efforts influence the prospects for peace.

The final section of the book takes stock of what we have learned with an SSIP approach and identifies avenues for future research. Maoz points to many successes of the SSIP community including the development of sophisticated data sets and numerous formal models to help understand conflict processes, the increasing sophistication of statistical models for analyzing SSIP data sets, and successful attempts to integrate the logic of paradigmatic approaches in the study of international relations, such as realism and liberalism. He also describes several shortcomings of SSIP research such as a disconnect between theory and empirical tests, a heavy emphasis on the dyadic level of analysis, and a failure to explain fully why a small number of countries fight most interstate conflicts in history (“fightaholism”).

The chapters in this edited volume provide readers with a very thorough introduction to the SSIP approach and the numerous contributions it has made to the broader understanding of conflict processes. Readers will gain valuable insight into the data sets, methodological advances, formal models, and theoretical arguments advanced by the SSIP research community. The increasing prominence of this research in major social science journals and book publishers attests to the success of the SSIP approach. This volume will help conflict scholars expose this material to a younger generation of SSIP researchers.

Section I

The SSIP Approach

1

The Origins and Evolution of SSIP

How Methods Met Models, with a Short Interlude

Kelly M. Kadera and Dina Zinnes

Introduction

The Scientific Study of International Politics (SSIP) became a section of ISA in 1993, largely out of necessity. Not so long ago, believe it or not, quantitative research in ­international relations had trouble seeing the light of day. The historically dominated field could not understand the statistical, mathematical, and data-based research, and traditionalists often found the results reported in such studies to be trivial and uninteresting, making both publication and convention panel participation difficult. Initially, SSIP researchers believed that subject-oriented panels were best: mixing traditional and quantitative research on a single panel with a common subject focus, such as ­deterrence, would encourage cross-fertilization and provide a greater understanding of the problem under study. This proved to be a false assumption and a ­misunderstanding of the role that public venues play in the development of research. Conventions, in particular, provide a researcher with the opportunity to get feedback on preliminary results from the intellectual community. If panelists, and consequently the audiences they attract, do not understand the methodology behind a piece of research or are hostile to the approach, panels only become wasted opportunities or fora for exchanging useless barbs. Faced with the challenges of obtaining spots on panels, let alone some that provided a productive exchange, Dina Zinnes and Hayward Alker gathered signatures and drafted the charter that led to the establishment of the SSIP section. Because ISA allocates convention panels to each section based on the size of the ­section and its panel attendance from the previous convention, formal status as a ­section ­guaranteed the SSIP community exposure time in potentially productive public ­environments where panelists and audiences spoke the same language. How did this group of scholars develop a research agenda that was substantial enough to warrant the formation of a new ISA section? Our chapter traces the progress of those who set down the foundations of SSIP.

In sketching the origins and evolution of the SSIP community, it is tempting to ­propose a linear chronology from one set of studies to another. To do so, however, would distort the realities of what happened. While some approaches did indeed lead to others, the story line contains several subplots that begin independently and only much later merge with the rest of the field in intriguing ways. But we are getting ahead of ourselves. Let us begin by telling the more straightforward, time-ordered tale, explain why it had to be so, and then turn to the parallel but largely distinct narratives. In the end, we show how the story lines have merged and where their shared themes are headed.

The Linear Chronology

The linear part of the story can be traced to the two world wars. International politics as a field of inquiry did not exist prior to World War I. And to the extent questions about ­international phenomena were entertained at all, they occurred largely in the context of law: the rules that should govern the interactions of states with one another. What, for example, was a just war, that is, under what conditions was it morally, and therefore legally, acceptable, for one state to attack another? But the horrors of World War I turned attention away from should, the ethics of state interaction, to questions about why things happen as they do. Hence, realism became the new approach as ­students of international politics sought to understand the dynamics of state ­interaction that led to such events as war.

The even greater catastrophe of World War II reinforced this new emphasis on understanding why and how things happen in the international arena. For if you did not understand the hows and whys of two world wars, what would prevent a third from happening? This brought scientists from other disciplines to the study of international conflict and war. Psychologists, sociologists, economists, and even physicists and ­mathematicians sought to use their skills to study and hopefully prevent future major conflagrations. The flurry of research by these scholars reinforced the realist ­perspective on understanding what is and added to it a demand for observation and the use of rules for measurement. These are the events that set the stage for what has become SSIP, known in its early years by names such as Quantitative International Politics or Interpolimetrics to emphasize the observational/measurement component.

As researchers began to tackle the problem of international conflict, it became clear that existing arguments about the whys and wherefores of interactions between states were vague and underdeveloped. These needed to be spelled out explicitly so they could be subjected to empirical appraisals. What affected what? The first steps towards clarification produced “frameworks” and “propositional inventories.”

Frameworks identified and classified the key variables thought to be responsible for inter-nation interactions, providing suggestions as to how they might be linked. One classic in this genre was Kaplan’s System and Process (1957). Kaplan argued that it was possible to define different types of international systems based on the characteristics of nation participants and the rules of the game, so to speak, which they used in interacting with one another. Snyder et al. (1962) proposed a very different type of ­framework. Their goal was to provide a generic outline of the domestic and ­international variables that structure foreign policy outputs. Rosenau’s “pretheory” (1966) was yet another noteworthy framework. While it was also an attempt to understand the foreign policy decision process, the goal was to demonstrate the differences in foreign policy outputs of nations as a function of key domestic variables. Propositional inventories, on the other hand, were far more specific and focused on particular variables and hypotheses. These inventories were extensive lists of hypotheses about what affected what, largely obtained from a careful reading of the descriptive and historical texts on international politics. Snyder and Robinson (1961) offered one of the most extensive of these inventories, covering all facets of decision making in international politics.

As the arguments about how the international world worked were clarified, it became increasingly obvious that there was a second, more serious, problem. If hypotheses ­generated by frameworks or inventories were to be subjected to empirical scrutiny one had to have access to data. But what constituted data in this field and how would one obtain relevant observations? The methods of observation used in other ­disciplines – experimentation, participant-observation, interviews – were either not relevant or not feasible. Two solutions emerged: (1) simulation; and (2) what might be called ­archeology. Let us discuss them in that order, breaking down the archeology ­investigations into subtypes.

Data Generation, Part I: The Simulation

Simulation was an adaptation of the data generation process used in psychology and was spearheaded by Harold Guetzkow, a social psychologist (see Guetzkow 1962 and Guetzkow et al. 1963). Guetzkow argued that international processes could be studied in a mock laboratory in which people played the roles of national decision makers and teams of 3–4 players were stand-ins for nations. The Inter-nation Simulation (INS) project consisted of 5–7 teams with members of each team playing the roles of head of state, domestic affairs advisor and foreign policy advisor. At the onset of a simulation period each team was given a profile describing its resources, decision structure (e.g., parliamentarian), and form of government (i.e., its relationship to its citizens). Teams were also given a scenario of the history of the international system up to the point at which the run was to begin; for example, who was allied with whom, whether there had been wars, whether there were existing international organizations, trading partners, and so on. As the simulation proceeded, teams were permitted to (1) interact through messages or prearranged person-to-person visits (state visits or summit meetings), (2) form alliances, (3) create international organizations, (4) declare war, (5) provide aid, and (6) trade resources.

At the conclusion of a run of the simulation, each team filled out a form indicating how it was allocating its original resources: a certain amount for trade, aid, for research and development, for domestic products, and the like. These forms were collected and, while teams took a coffee break, calculations proceeded to determine what ­happened to each nation as a consequence of its resource allocation decisions. For example, in countries that had begun with minimal resources and a history of ­population unrest, inadequate allotments to domestic development could lead to a revolution. Alternatively, large allocations to research and development in one round could mean that the nation would have a major breakthrough in weapons technology in the subsequent run and consequently the capacity to develop nuclear weapons. If a war had been declared then the calculations would determine winners and losers.

It is important to note that the calculations of the consequences of the actions and allocations of the players were governed by explicit mathematical formulae, not simply by the judgments of the experimenters. Moreover, these formulae, or rules, were ­constructed based on the INS team’s hypotheses about how the international system worked. In particular, they were based on some of the same hypotheses and arguments that underlay the frameworks and propositional inventories. Hence, the INS ­researchers made extensive attempts to model the structure of their simulations on what was known about the real-world operation of international politics. This was done in an effort to answer the critics that argued that INS had a serious validity issue: how could teams of high school or college students represent the decisions of national statesmen?

The validity question stalked INS researchers for many years and the attempt to solve or simply answer the charge led these researchers in several directions. One was to attempt to replicate a real world event such as World War I. Structuring the simulation scenario to parallel the principal participants and their relationships prior to 1914, they ran the simulation to see if war would indeed occur. Although the desired result was obtained, questions were raised as to the extent to which the participants (who were not told that the run was an attempt to approximate a real world event) produced war because of their familiarity with the historical context.

Another, very different, approach was to evaluate the underlying propositions that governed the plays and outcomes of the simulations. Thus data on nations were ­collected and hypotheses about the relationship between national attributes on the one hand and national behavior on the other were tested. Intriguingly, these efforts gave birth to a very different project that soon took on a life of its own: Rudolph Rummel’s Dimensionality of Nations (DON). Rummel was a student at Northwestern when the efforts to validate INS were underway and he was originally in charge of the data collection and hypothesis testing. But Rummel’s interest in INS was quickly ­overtaken by his fascination with data collection and hypothesis testing itself. Rummel’s interest in understanding how the attributes of nations translate into behavior led him, much later, to propose how the relationship between national attributes across nations might determine their interactions: thus we have one of the nonlinear spin-offs: field theory. But again, we are ahead of the story. We will return to Rummel’s spin-off work in our discussion of archeology.

A third, less direct, INS response to the validity critique was to point out that some questions of great concern to the field could only be studied in the laboratory. While the laboratory was not perfect, perhaps, they argued, it could at least provide insights. One of the more important attempts in this direction was Richard Brody’s (1963) study of the Nth country problem: how a widespread change in military technology, the advent of nuclear weapons, could affect the structure of an international system and the interactions among the nations. Brody constructed a bipolar INS system: two major powers (highly resource endowed) with a number of smaller nations in alliance with each major power. The simulation was run for a number of periods under these conditions and the perceptions and actions of the nations recorded. It was found that the smaller nations interacted almost entirely with the superpower in their alliance, having little to do with other smaller nations in their alliance or any of the nations in the opposing alliance. Additionally, all nations within a given alliance saw members of the opposing alliance as threatening. Then nuclear weapons were introduced. Several of the smaller nations discovered at the beginning of a new round that they had struck it rich, so to speak: due to previous investments in research and development, these nations now had nuclear capabilities. The interaction patterns changed dramatically. The bipolar alliance structure crumbled and the perceptions of threat were now ­ubiquitous. Every nation became fearful of all the others. Whether these runs provided an insight into the breakdown of social interactions between groups or said something of consequence about international politics is, of course, open to question. The results, however, are nevertheless intriguing. And because there is only one system to observe at any point in time, empirical testing is difficult, meaning such simulations may remain scholars’ best tool for understanding relatively rare, but significant, ­phenomena at the aggregate level.

Data Generation, Part II: Archeology of Actors

The other answer to the data question was to turn to the traces of things that had ­actually happened – just like conducting an archeological dig. Using the historical record, this approach sought to collect, in a systematic and explicit fashion, ­information about the characteristics of nations, their behavior and their interactions. Those who chose this route believed that these data were far more real than the information ­gathered from simulation runs. With time, however, it became evident that the validity problem did not disappear. It now simply showed itself in another form: how do you define a war, an alliance, or even an interaction between nations?

Thus began the many data generating projects of the 1960s and 1970s. While all of these projects rely on some written record – whether it be the historians’ accounts, yearly statistical compilations by various international agencies, or newspaper reports – they were anything but unified. These enterprises varied in the principal questions that drove them, the sources used to extract the data, the definitions of the variables of interest and the methodologies applied. To take but one example of how these data collections differed even in the definition of a single variable, consider the three ­different data sets on war. For Quincy Wright (1942), a political scientist, a war existed and was recorded in his collection only if there had been a legal declaration of war by one state against another. When J. David Singer, another political scientist, initiated the Correlates of War (COW) project, a war was defined in terms of the number of individuals killed on the battlefield; an event was counted as a war only when 1,000 people had been killed (Singer 1979). On the other hand, Lewis Fry Richardson, a meteorologist and Quaker, sought to understand why any disagreement ended up in the death of even a single individual. Thus his data collection focused on “deadly ­quarrels” and contained murders at one end of the continuum through gangland ­executions to the world wars at the other (Richardson et al. 1960b).

Despite these important differences, the many data collection efforts can be ­classified as being (1) attribute oriented, (2) behavior/interaction oriented or (3) a ­combination of the two. The attribute collectors were principally concerned with recording, over time, the characteristics of nations – population size, GDP, square miles of territory, number and composition of minority groups, regime type, and so on. These projects became the various World Handbooks (Russett and Banks 1968; Russett et al. 1968; Taylor and Jodice 1968; Taylor and Hudson 1975a, 1975b; Taylor et al. 1975) and Ted Gurr’s Polity enterprise (Eckstein and Gurr 1975).

The behavior/interaction projects on the other hand were concerned with tracking the events that transpired between nations. To a large extent these event data efforts shared a common focus on international crises as a potential prelude to war. The ­oldest of these is WEIS, the World Event Interaction Survey, initially begun by Charles McClelland (1971). McClelland argued that international crises could be predicted by classifying events into hostile, neutral, and cooperative types and monitoring the ­co-occurrence of event combinations. Crises, he posited, were the culmination of sets of particular kinds of interchanges between nations. Using principally the New York Times (and later adding the London Times), every action taken by any nation towards any other nation was coded by indicating the day on which it occurred, the nation perpetrating the event, the target of the event, and the event type. Thus WEIS was a massive daily chronology of every action directed by one nation towards another. Using indices from information theory such as Hrel (Miller 1964), McClelland (1972) ­demonstrated how combinations of hostile/cooperative acts could predict the onset of international crises.

Approximately a decade later, and unaware of the ongoing WEIS effort, Edward Azar began work on COPDAB, the Conflict and Peace Data Bank (Azar 1980). Born in the Middle East with strong family ties to the region, Azar looked at the interactions between nations from a different perspective. Azar’s concern, like McClelland’s, was with international crises, but unlike McClelland, Azar focused on recurring crises between the same participants, the ongoing, seemingly endless, Arab – Israeli conflict that would cyclically heat up, cool down, and heat up again. Azar called these crises “protracted conflict” and his data collection efforts were an attempt to chart the course of long, drawn-out interactions. Like McClelland, Azar developed a classification scheme for these events in terms of hostility and cooperation. However, unlike McClelland, Azar saw this classification as a scale that ran from high to low levels of hostility and from low to high levels of cooperation. Sensitive to scaling issues, Azar utilized methodologies from psychology to assign weights to the conflict – cooperation categories so that the intensity of hostility or cooperation could be meaningfully assessed for a designated time period, permitting the researcher to observe the ebb and flow of conflict or cooperation.

In addition to the differences in their classification schemes, WEIS and COPDAB differed in the sources used to extract the events. Azar argued that Middle Eastern events were inadequately covered in the Western press and that reliance on a single source, like the New York Times, would provide a distorted picture of what was ­happening in that region. Consequently he turned to the use of multiple regional news outlets. While both WEIS and COPDAB had to struggle with reliability and validity issues in the definition of the various types of events and the training of coders, the use of multiple sources created an additional problem for COPDAB: knowing when an event recorded in one source was the same or different from an event recorded in another source. Unless one could make this distinction, COPDAB would run the risk of over-recording events and thus falsely magnifying the amount of conflict or cooperation occurring on a given day.

Although COPDAB began with a focus on the Middle East, it soon expanded to world coverage, rivaling WEIS. This led to comparisons and evaluations of the relative merits of the two (Howell 1983; Vincent 1983) and, to some extent, arguments over the usability of one versus the other. However, because the classification schemes ­differed in their definitions of types of interactions and the fact that COPDAB came with scaled weights, the superiority of one over the other was never clear. This, together with the fact that the two projects had very different funding sources, kept both alive. DARPA, the Defense Advanced Research Projects Agency of the Defense Department, had been deeply involved in supporting WEIS and was reluctant to switch gears after putting so much behind the efforts to create a crisis indicator. COPDAB, on the other hand, began with small university backing and then, on and off, received limited ­support from the National Science Foundation (NSF).

Three other event data collections emerged: (1) Frank Sherman’s (1994) Sherfacs, which focused on the phases of conflict escalation; (2) Wilkenfeld, et al.’s ICB (International Crisis Behavior) project (see Brecher at al. 1988; Wilkenfeld et al. 1988), which centered on foreign policy behavior and crises as the units of analysis; and (3) Pearson’s (1974) foreign interventions data set, which allowed analysis of the outcomes associated with various types of interventions.

What came to be called the 1914 Study represents a very different type of event data collection process. Like WEIS and COPDAB the focus was on international crisis. But unlike these efforts, the 1914 Study was interested in only one particular crisis: World War I. Looking at this single cataclysmic event, researchers attempted to capture the play-by-play sequence of events that led to World War I. This detailed account of an international disaster focused on how the principal decision makers in the European capitals reacted to one another – their perceptions and actions – to eventually produce the disaster known as World War I. Using both the classic histories (Fay 1928; Albertini 1957) of this conflict and original documents that were found in the archives of the Hoover Institution, Stanford researchers, under the direction of Robert North, coded the activities and perceptions of the decision makers as the events unfolded from the assassination of the Austrian Archduke to the declarations of war (see, e.g., Zinnes et al. 1961; Zinnes 1962).

The event data projects surveyed thus far can all be characterized by their principal focus on actions. There were, however, two projects which were event based but ­additionally had an important national attribute component: CREON and DON. CREON, the Comparative Research on the Events of Nations (Hermann et al. 1973), grew out of the Rosenau framework mentioned earlier. Like Wilkenfeld and Brecher’s project on comparative foreign policy, CREON was an attempt to understand the ­formulation and execution of foreign policies. But while Wilkenfeld and Brecher were interested in the foreign policy formulation process, the Hermanns, who spearheaded CREON, wanted to evaluate the Rosenau paradigm that linked types of nations to types of foreign policy decisions. Thus the CREON researchers needed to collect data on both the attributes of nations to permit them to properly place a nation in the Rosenau typology and the actions that these nation-types pursued, that is, their foreign policies. They hoped to show that the foreign policy of a small, underdeveloped nation was very different from the foreign policy of a large developed country. This ­cross-national perspective had a dramatic effect on CREON’s event data collection procedures. Because WEIS and COPDAB were interested in how crises evolve, their data sets were collected through time. CREON’s concern with types of nations and types of foreign policies, however, made time irrelevant. Thus CREON events were extracted from news sources by sampling quarters within the years covered.