CENTRAL ISSUES IN THE STUDY OF INTERNATIONAL CONFLICT
BRUCE BUENO DE MESQUITA
The study of international conflict is undergoing a rapid and dramatic transforma¬tion. Some of the field's most venerable beliefs now confront fundamental challenges to their logic and their empirical reliability. For example, the idea that a balance of power promotes peace and an imbalance war traces back more than two millennia to Thucydides' History of the Peloponnesian War. It is a belief that continues to permeate the thinking of secretaries of state or defense in the United States and of foreign ministers and defense ministers throughout the world. Yet careful research makes it clear that the distribution of power, whether balanced or not, by itself bears neither a logically nor an empirically compelling relationship to the likelihood of international instability or conflict (Niou, Ordeshook, and Rose 1989; Bueno de Mesquita and Lalrnan 1992; Kim and Morrow 1992; Powell 1996).
The study of conflict was long thought of as high politics, with states as the central actors pursuing grand strategies designed to maximize national security. In fact, the view of the state as the key player in the international arena is so strong that the Eng-' lish language does not provide a common word or phrase to describe international relations without invoking the nation as the object of study. Now the subject is being recast as part of the everyday pulls and tugs of domestic affairs, integrating the study of conflict into investigations of all other aspects of politics. Where once national interest was at the forefront of research on international affairs, increasingly the study of the political ambitions of and institutional constraints faced by leaders is supplanting concern about the national interest. Indeed, insights from spatial models and other approaches to domestic politics challenge the fundamental meaning of "the national interest." When it is possible to assemble many majority coalitions around competing interests within the same population of national decision-makers it is difficult to say that the interests backed by one coalition are closer to the national interest than the policies backed by another coalition (McKelvey 1976, 1979; Schofield 1978).
In this chapter I touch upon some debates regarding international affairs and the ways we go about studying conflict and peace. International relations scholarship can loosely be divided into those who focus on such structural aspects of the international system as the global distribution of power, or the alignment of nations into two blocs—the bipolar world of the cold war—or many blocs and those who attend to the ways in which domestic political dynamics shape international relations. For much of the post–Second World War years, structural perspectives including neo¬realism (Waltz 1979; Glaser 1992; Schweller 1994, 1997), liberalism (Keohane and Nye 1977; Keohane 1984), and power transition theory (Organski 1958; Organski and Kugler 1980; Tammen et al. 2000; Lemke 2002) have contended for domination as explanations of variance in cooperative and conflict-prone behavior. By structural perspectives I mean theories, such as those just listed, for which the central concern is how aggregate characteristics of the international system such as the distribution of power or distribution of wealth among states (as rational unitary actors) determines interactions leading to international conflict or cooperation.
Neorealism treats international affairs as anarchic; that is, a self-help system lacking a dominant power that can enforce agreements. It hypothesizes that political stability in the form of the survival of states is ensured by national efforts to maximize security through alignments. Its focus is on explaining conflict and constancy in relations among states under the supposition that war is the natural state of affairs. Liberalism countered by assuming that international relations are hierarchic rather than anarchic and that a hegemonic power, such as the United States, stands atop the international system. Its focus is on the prevalence of cooperation rather than conflict and so naturally draws researchers into investigations of such international political economy arenas as trade policy, the use of economic sanctions, and international banking policy. The power transition shares with liberalism a focus on hierarchical power relations but centers its attention on how domestic economic growth might influence the risk of system-transforming wars; that is, wars that fundamentally alter who is on top and, therefore, what policies or norms of action are imposed on the community of nations.
During the past decade there has been a shift away from structural perspectives toward ones that look at how domestic political institutions constrain and create incentives that shape cooperation and conflict (Putnam 1988; Bueno de Mesquita and Lalman 1992; Fearon 1994; Downs and Rocke 1995; Schultz 1998, mom; Bueno de Mesquita et al. 1999; Werner 1996; Goemans 2000; Bueno de Mesquita et al. 2003; Gartzke and Gleditsch 2004).
The purpose of this chapter is to survey both categories of perspectives and to high¬light how different the view of international politics is when focused on political lead¬ers responding to domestic political incentives rather than unitary states responding to aggregate characteristics of the collectivity of states. To this end, this chapter proceeds as follows. Section z examines structural perspectives. Section 3 discusses the evolving research programs that look inside the state to grasp how conflict or cooperation relates to ordinary politics. Section 4 illustrates the differences across approaches by selecting two empirical issues, briefly evaluating how different ap¬proaches address these issues. The two empirical puzzles are (1) the difference be¬tween the record of nation-building following military intervention and the claimed intentions of interveners, and (2) reconciliation of the existence and use of foreign aid as a policy instrument and the record of its ineffectiveness in alleviating poverty.
2 Structural Theories
In this section I summarize key claims and evidence from the two most prominent structural theories: neorealism and the power transition. Although liberalism is an equally important strand of theorizing, its focus on cooperation places it somewhat outside my purview. Therefore, other than in passing commentary, I overlook the strengths and weakness of this approach. Those interested in this aspect of inter¬national relations research might see Morrow (1994), Levy (1998), and Simmons (1998) for insights into alternative ways to think about problems of coordination and cooperation. Space constraints preclude discussing other structural perspectives.
The structural point of view is most prominently represented in Kenneth Waltz's (1979) neorealist theory. Neorealism begins with a handful of seemingly straight¬forward, innocuous, and parsimonious assumptions. These include the view that international politics is anarchic, states seek to maximize their security against threats to sovereignty; bipolar environments produce substantially less uncertainty about the responses of other states to crises than do multipolar environments; and states seek to increase their power unless doing so is expected to put their security at risk later on.
By definition, an anarchic international system, as intimated earlier, is one in which states cannot make binding commitments enforced by some supernational authority or hegemonic state. Therefore, international politics is best characterized as a non-cooperative game. It is a game in which all states are assumed to share common goals: to make themselves secure and, subject to achieving security, to make themselves more powerful. Neorealists also assume that international interactions arise in a prisoner's dilemma setting (Waltz 1979, 109; Gowa and Mansfield 1993), but, incongruously, also argue that because power is relative, gains for one state must mean losses for others (Waltz 1979, 170-1; Grieco 1988), suggesting a zero sum game. Of course, one cannot logically hold simultaneously that international affairs are characterized as a prisoner's dilemma and as a zero sum game. Some have tried to repair this contradiction by introducing additional parameters into a simple game form that allows the game, under different parameter values, to morph into either the prisoner's dilemma or a zero sum game (Snidal 1991; Powell 1991). In the end there is not much evidence to support the belief that much insight into international affairs is acquired by focusing on the extent to which decision-maker utility functions revolve around relative gains or absolute gains.
Neorealists go on to argue that to improve security or to gain power, states cluster together in blocs that serve as self-help mechanisms. When all states are aligned into only two blocs with each bloc dominated by one especially powerful state, everyone can be certain about how everyone else will react to a crisis. With more than two blocs and more than two large powers, crises engender more uncertainty because a pair of states in a dispute involving two blocs cannot be certain about how those in other blocs will respond. Here, the literature conflates the idea of bipolarity or multipolarity with complete and perfect information or uncertainty.
The core hypotheses of neorealism are generally thought to be that:
1. Bipolar systems, because they involve little or no uncertainty, are more stable than multipolar systems (Waltz 1979);
2. Power tends to become balanced because imbalances place some states' survival at greater risk (Niou, Ordeshook, and Rose 1989); and
3. Alliance partners' responses to threats in a multipolar environment depend on perceived dependence on one's allies or complacency toward the alliance structure (Christensen and Snyder 1990).
Several formal models attempt to capture essential elements of neorealism. One reason why there is more than one such model is that neorealists tend to disagree about the exact meaning of some fundamental concepts, such as security or stability (Elman and Vasquez 2003; Volgy and Bailin 2003). Different formalizations emphasize different meanings attributed to core concepts while placing the informal arguments of neorealists in a political economy framework that helps uncover precise implica-tions and potential sources of inconsistency
Niou, Ordeshook, and Rose (1989), for instance, present a model of an anarchic, self-help system in which states seek survival above all else.. Their model demon¬strates that stability has two distinct characteristics that are conflated in informal examinations of neorealism and the balance of power. One, which they call system stability, refers to environments in which no states—that is, no members of the international system—cease to exist so that the composition of the system remains fixed. The other, which they call resource stability, is achieved when the distri¬bution of power does not change among the states in the international system.
They note that the international system can experience system stability without resource stability.
Niou et al. deduce four theorems that seem consistent with neorealist expectations about the stability of the international system. These theorems, stated informally here, indicate that essential states—states whose resources are necessary to avoid such an imbalance of .power that there is system instability—are never eliminated from the international system; essential states never become inessential; inessential states never become essential; and inessential states are always eliminated from the international system. They also show that an infinite variety of distributions of power are consistent with system stability though not necessarily resource stability in a multipolar setting. These distributions of power all meet the criteria for being balances of power in the logical, if not the intuitive, sense of that term. Thus, the phrase "balance of power" appears to have little discriminatory power in their formal representation of neoreal-ism. What is more, all four of the theorems that follow from their formalization are empirically false.
By way of illustration, we can surely agree that the Soviet Union and Austria-Hungary were essential states that no longer exist. The United States, inessential in the late 1.700s or early i800s, is obviously essential today. Contrary to the theorems, states rise and decline in their essentialness and states come and go in terms of existence. Indeed, so prevalent is state failure that it is a topic of current research (Beck, King, and Zeng z000).
Robert Powell's (1993) formalization of neorealism shows that under conditions of anarchy, pursuit of national security is unlikely to dominate state resource allocations. Contrary to views expressed by Waltz (1979) and earlier realists, peace and stability can be achieved while spending relatively little on defense. Contrary to a basic tenet of realist thought, neither war nor its imminent threat is the normal state of affairs in anarchy. Powell examines the trade-off between consumption spending and defense spending and demonstrates that equilibrium defense spending in a bipolar system is rather modest if the environment is anarchic. He shows that a small amount of military spending is sufficient to deter the threat of an attack by a would-be attacker who is interested in improving her people's future consumption and that, since this holds for both parties to a potential arms race, neither has an incentive to spend above the minimal amount needed to achieve successful deterrence.
Additionally, Powell demonstrates that a long shadow of the future (i.e. patience) can decrease cooperation and increase the risk of conflict. His result depends on the timing of costs and benefits. By giving up some consumption now to spend more on arms—that is, adopting a strategy of defection--La state can create a first-strike ad¬vantage that improves its chances of gaining a future stream of consumption benefits if the adversary does not sacrifice current consumption for arms. Thus, unlike the well-known folk theorem applied to repeated prisoner's dilemmas, Powell presents a game in which defection—arms rather than consumption—today yields greater benefits tomorrow because of a current military advantage instead of yielding a larger benefit today (the PD temptation payoff P) resulting in long-term punishment and, therefore, loss of expected benefits over time. This latter result contradicts a central claim in the research agenda put forward by neoliberals (Axelrod 1984; Axelrod and Keohane 1986).
Buena de Mesquita and Lalman (1992) construct a realist and a domestic variant of their international interaction game from which they deduce several propositions. In keeping with neorealist expectations, they show that an imbalance of power is a logically necessary, but not sufficient, condition for war in the realist variant. Additionally, they show that regardless of information conditions, it is impossible in the realist variant for one state ever to choose to acquiesce to the demands of another state. Finally, they demonstrate that war in the realist variant can only arise as a consequence of uncertainty in keeping with Waltz's hypothesis about the different risks of instability under bipolarity and multipolarity. This is in contrast to their demonstration that commitment problems can lead to war even with complete and perfect information. Fearon (1995), in fact, identifies three rationalist explanations for war: asymmetric information, commitment problems, and a dispute over indivis¬ible goods. Chiozza and Goemans (2004) suggest that when we focus on individual decision-makers rather than states the number of rationalist explanations for war proliferates still farther.
The empirical record contradicts the three neorealist hypotheses Bueno de Mesquita and Lalman (1992) deduce from their realist variant of the international in¬teraction game and supports at least partially the propositions that follow from their domestic variant (Bueno de Mesquita and Lalman 1992.; Signorino 1999; Smith 1999; Bennett and Stam 200o), thereby reinforcing the evidence associated with Niou et al.'s formalization, suggesting that the realist account is poorly supported by evidence.
2.2 Power Transition
Power transition theory was first developed by A. F. K. Organski (1958) and then re¬filled by Organski and Kugler (1980), Tammen et al. (2000), Lemke (2002), and others. It views international politics as hierarchic rather than anarchic. It assumes that all states are interested in imposing international organizations, institutions, rules, and regulations that govern international intercourse on the international system. States are assumed to maximize their control over "the rules of the game" or the status quo norms and policies in the international system, ideas also echoed by Gilpin (1981) and by Krasner (1983) in developing the literature on international regimes.
Power—or relative material resources—is taken as exogenously given and is as¬sumed to be the determinant of who controls international affairs. The most powerful state sits at the apex of a power pyramid in which those below the apex aspire to rise to challenge the dominant state for control of the "rules of the game:' States are assumed to be divided into two broad coalitions. Those in the satisfied coalition find the dominant state's rules and norms acceptable while those in the dissatisfied coalition do not. The opportunity to challenge for control arises when a dissatisfied state's in¬ternal growth rate is fast enough relative to the dominant state's that the "challenger" can be expected to pull equal in power to and then overtake the dominant state. In the temporal window during which this power transition occurs, war is predicted to be likely if the challenger is dissatisfied (as in the Franco-Prussian War of 1870-1). A peaceful transition is anticipated if the challenger is part of the satisfied coalition (as between the UK and the USA in the twentieth century).
Two formalizations of power transition suggest important modifications to this theory. Kim and Morrow (1992) show that the likelihood of war is not materially influenced by differences in the growth rates between challengers and the dominant state. This is so because a more rapid growth rate for the so-called challenger makes waiting to attack more attractive. The longer the challenger waits, the better its chances of prevailing. At the same time, rapid growth by the challenger makes delay risky for the declining dominant state because that state's odds of victory are decreas¬ing with time. Who will move first and whether the contenders will fight or resolve their transition peacefully depends on their respective willingness to take risks as it is the degree of asymmetry in the shape of their utility functions, and not growth rates per se, that determines the trade-off between fighting now, fighting later, or never fighting. Kim and Morrow's formalization also shows that equality of power is not an inherently critical factor determining the risk of war. Their theoretical claims are buttressed by substantial empirical tests that are consistent with their formalization and that depart significantly from claims made in informal statements of the power transition theory.
Powell (1996) replicates the theoretical structure of the power transition in his investigation of the consequences of appeasement and the risks of what he calls salami tactics by which the dominant state slowly gives very small bits of value away until eventually nothing is left. Except under special conditions that include the anticipation of a reversal in the declining fortunes of the dominant state, Powell demonstrates that with incomplete information about the demanding state's satiation point for extracting resources from the rival, there are numerous conditions that lead to or away from war and that with complete information war never arises as an equilibrium outcome in his power transition/appeasement game. Like Kim and Morrow, Powell fails to find support for the central tenets of the power transition theory or for balance of power theory.
Neorealism and power transition focus on states as the central actors in international politics. For neorealists, what happens within states is largely irrelevant for under¬standing the stability of the international system or the security of states. Each state is assumed to have an abiding interest that any leader follows. For power transition theorists, conditions within states matter to the extent that they influence growth rates, but those conditions are assumed to be exogenously fixed and immutable. The two theories differ in their view of international politics, with neorealists assum¬ing that international politics is anarchic and power transition theorists assuming it is hierarchic. They also disagree about what the goals are that are pursued by states: security or control over the "rules of the game." Although the power transition assumes that different states have different policy goals in mind (i.e. different rules), its predictions are not driven by the substantive differences in goals. Thus, domestic political competition over national policy plays no part in either power transition theory or neorealism.
The empirical record does not support the main claims of neorealism and only partially supports the core claims of power transition theory. One apparently well-established empirical regularity in particular—the democratic peace—has raised se¬rious doubts about the centrality of these and other theories concerned with system structure or with theories that treat the state as a unitary actor. Several empirical regularities are associated with what has come to be called the democratic peace. Fore¬most among these is the observation that democracies rarely, if ever, fight wars with each other while democracies and autocracies fight with one another (and autocracies with other autocracies) with regularity. The evidence for these and other democratic peace regularities highlights the realization that domestic characteristics of regimes lead to sharply different patterns of foreign policy behavior, a fact that cannot be true if states are rational unitary actors whose patterns of behavior are determined by factors outside the domestic politics of the state as argued by structuralists.
Pueblo Libre es la red de Seguridad Ciudadana para todos los vecinos de Pueblo Libre iniciando Enero 2,019.