Foreign Policy Orientation, Strategic Interaction, And the Initiation of International Crises

Foreign Foreign Policy Analysis (2005) 2, 165–179

Foreign Policy Orientation, Strategic Interaction, and the Initiation of International Crises JAMES MEERNIK

University of North Texas In recent years, scholars studying U.S. foreign policy and the diversionary use of force have begun to focus more attention on when foreign regimes time crises with the U.S. Many argue that U.S. domestic conditions play a role in this. I argue that these models should take into consideration the foreign policy relationship between foreign governments and the U.S. I develop a theory of crisis initiation that considers the foreign policy orientation of states that may initiate a crisis with the U.S., and second, when the crisis may occur. I argue that by virtue of their foreign policy orientation, orientation, some regime leaders leaders will be more likely likely than others to initiate a crisis with the U.S. Those regimes that are either closely closely aligned with the U.S. or closely closely identified identified with anti-U.S. anti-U.S. interests will be much more likely to initiate crises with the U.S. than those whose foreign policy interests do not lead to such intimate ties of friendship or enmity. I develop and test hypotheses predicting what regime characteristics will be predictive of crisis initiation and when such crises are likely to occur. The results demonstrate that regime type matters more than U.S. domestic conditions in predicting where and when crises involving the U.S. will take place.

Empirical Empirical research on the use of force by American presidents began principally principally by investigating the linkage between domestic, political, and economic conditions in the U.S. and the likelihood presidents would use military force overseas (e.g., Ostrom and Job, 1986; James and Oneal, 1991; Morgan Morgan and Bickers, Bickers, 1992; James and Hristoulas, 1994; Meernik, 1994; DeRouen, 1995, 2000; Meernik and Waterman, 1996; Wang, 1996; Yoon, 1997; Fordham, 1998a, b; James and Rioux, 1998). While initial findings seemed to demonstrate that presidents did engage in diversionary use of force, increasingly research in this area has questioned these findings. More recently, scholars have begun to argue that one principal reason why the evidence for diversion diversionary ary behavio behaviorr is so weak weak is because because of ‘‘strat ‘‘strategi egicc interac interactio tion’’ n’’ (Smith, (Smith, 1996a, 1996b, 1998; Leeds and Davis, 1997; Meernik, 2000, 2001; Clark, 2003). Simply put, other states are dissuaded from initiating a crisis against the U.S. when the president is experiencing difficulties at home, for fear of provoking a diversionary retaliation. This new area of inquiry regarding the causes of international crises expanded our understanding of presidential foreign policy decision making by making clearer how president’s choices are shaped by the actions of external actors. Yet, we must wonder if our focus on the impact of domestic conditions within the U.S. on the occurrence of international crises has led us to ignore the extent to which other factors predict the initiation of international crises involving r 2005 International Studies Association.

Published by Blackwell Publishing, 350 Main Street, Malden, MA 02148, USA, and 9600 Garsington Road, Oxford OX4 2DQ, UK.

166

Foreign Policy Orientation

the U.S. In particular, do factors pertaining to nation-states that might initiate a crisis, such as their foreign policy orientation, influence the probability of such events occurring? In fact, I argue that it is necessary to consider both which types of nation-states are likely to initiate crises, and when these nation-states might provoke a crisis in order to determine both where and when crises begin. Models of strategic interaction may well be underspecified if they focus solely on conditions within the U.S. More generally, it is also time to begin the development of a broader theory of crisis initiation that will allow us to determine if certain types of states are more likely than others to initiate crises. Therefore, the twin purposes of this paper are to better understand the influence of the timing of international crises, especially regarding strategic interaction with the U.S., and to begin to understand the effects of foreign state characteristics on the occurrence of these events. I develop here a theory of crisis initiation that considers principally the foreign policy orientation of nation-states that may initiate a crisis with the U.S. I argue that by virtue of their foreign policy orientation some nation-states will be more likely than others to initiate crises with the U.S. Those nation-states that are either closely aligned with the U.S. or closely identified with anti-U.S. interests will be much more likely to initiate crises with the U.S. than those whose foreign policy interests do not lead to such intimate ties of friendship or enmity. I argue that the type of political and military/security policies pursued by states will be highly predictive of their propensity for conflict involvement. Nonetheless, if some nation-states are more prone to initiate crises than others, their propensity for dispute involvement is not likely to be constant across time. It is also possible that these states time the initiation of crises to take advantage of domestic and political conditions inside the U.S. In addition to considering where international crises might occur, we must also account for when and why nations might strategically time crisis initiation. I develop here a more comprehensive model of international crisis initiation that considers both the extent to which characteristics particular to nation-states and conditions peculiar to the U.S. make crises more likely. The development of this theory of crisis initiation will help us address several interconnected puzzles in the literature on international crises, strategic interaction, military intervention, and U.S. foreign policy. First, do regimes strategically time crisis initiation, and if so, when are they most likely to initiate a crisis? Some argue that nation-states time crisis initiation to avoid periods when U.S. presidents are more likely to retaliate with a diversionary use of force. Others contend that nations will be more likely to precipitate crises when U.S. domestic and political conditions are hurting the president under the assumption that presidents will be preoccupied during such times, and therefore less likely to respond. By developing a more comprehensive theory and empirical test of crisis initiation, we will be able to obtain more informed answers to this puzzle. Second, if regimes do not strategically time crises, what factors determine when crises are initiated? Ultimately, I argue that the probability that a nation-state will initiate a crisis is more likely because of its own foreign policy interests, specifically the nature of its relationship with the U.S. To address these issues, I develop this paper accordingly. First, I describe my theory of foreign policy orientation and crisis initiation. Second, I discuss the literature linking U.S. domestic and political conditions to crisis initiation. Third, I describe other factors that may affect crisis initiation. Fourth, I discuss the measurement of all the variables. I then assess the results, and conclude by suggesting how research in this area should proceed.

Foreign Policy Orientation and the Initiation of Crises Involving the U.S. While it may be impossible to account for all the reasons that might lead a particular regime to initiate a crisis involving the U.S., I contend that it is possible to develop a general theory of crisis initiation that incorporates the most important theoretical

JAMES MEERNIK

167

factors. Specifically, I argue that the foreign policy orientation of a potential crisis nation vis-a´-vis the U.S. will contribute substantially toward the likelihood that the nation will initiate a crisis that involves the U.S. Regimes and their leaders in the conduct of their foreign policies make choices regarding the most effective method to advance a mixture of both external and internal policy preferences. Among the most critical policy choices a regime makes is its foreign policy orientation vis-a´-vis the dominant power in the system. By foreign policy orientation, I mean the degree to which a regime defines itself in opposition to or in common cause with the U.S. politically. I contend that the greater the intensity of the foreign policy orientation of a nation toward the U.S.Fwhether evidencing a greater degree of cooperation or conflictFthe greater the likelihood a regime will initiate a crisis that will involve the U.S. Henceforth, it should be understood that I am analyzing the potential for all nationsFU.S. friends, enemies, and neutralsFto initiate crises involving the U.S. Thus, the initiation of a crisis does not mean that the crisis action taken directly threatens or challenges the U.S. Rather, I am interested in all crises that affect U.S. interests, regardless of who initiates the event. Enemies may threaten the U.S. (e.g., the U.S.S.R. during the Cold War) and friends may involve the U.S. in their crises (e.g., the U.K., France, and Israel during the 1956 Suez crisis). In either case, because the crisis affects U.S. interests, it becomes involved as a crisis participant. In the analysis section, I do distinguish between the crisis initiation proclivity of states with whom the U.S. sides and states against whom the U.S. sides. In one sense, there is nothing inherently novel in this theoretical assertion for we would expect that states that interact a great deal will have more opportunities for involvement in crisis politics. I believe, however, that the substance of a nation’s foreign policy rather than its level of interaction with the U.S. is a more critical factor in explaining crisis initiation. Indeed, the U.S. may have little foreign policy interaction with a nation (perhaps because it is seeking to isolate it [e.g., China early in the Cold War]), but may become involved in crises with it because of that nation’s degree of hostility toward the U.S. Conversely, the U.S. may engage in a wide and deep range of political interactions with another state while the tenor of the relationship may be mostly neutral or ‘‘non-partisan’’ in tone with little reason for crisis politics (e.g., perhaps Nigeria). Therefore, I argue that the quality of a nation’s foreign policy orientation toward the U.S. is more important than the quantity of its interactions.

Foreign Policy Orientation and External Security

Foreign policy orientation with or against the U.S. may be pursued both for external security reasons and for internal political reasons. First, we are likely to find that foreign policy orientation vis-a´-vis the U.S. is determined in part by external security considerations. Researchers generally argue and find that many international crises and wars occur in large part because of the divergence of foreign policy interests among states (Bueno de Mesquita, 1981; Bueno de Mesquita and Lalman, 1992; Brecher and Wilkenfeld, 1997; Reed, 2000). Some regimes may perceive that their security is jeopardized by the hegemon’s security interests if the U.S. poses a threat to a regime’s military, economic, or political influence locally, regionally, or internationally. Those states whose foreign policy orientation indicates considerable divergence between their national interests and the policies of the U.S. should be more likely to initiate crises involving the U.S. in order to lessen the threat posed by the U.S. Indeed, when we conceive of states initiating international crises involving the U.S. (especially in models of strategic interaction) we tend to think in terms of states whose interests are hostile to the U.S.’s. Many international crises undoubtedly involve such adversarial competition, but hostile regimes are not the only ones whose interests can be furthered through crisis politics.

168


Those states whose foreign policy orientation evidences substantial compatibility vis-a´-vis the U.S. should be more likely to initiate crises involving the U.S. in order to increase the amount of security provided by the U.S. Nations that maintain good relations with the U.S., and may depend on the U.S. for security guarantees and regime protection may also have substantial interests that contribute to an increased propensity for crisis initiation. Their close ties with the U.S. signal an a priori American interest and commitment to such regimes, and hence these states may initiate a crisis against a third nation in the expectation that the U.S. will come to their assistance. Regime leaders may believe that given their close ties and cooperation with the U.S., they enjoy a free hand in the international arena and thus act to maximize their interests in the expectation of U.S. backing (e.g., Israel perhaps). Furthermore, the very closeness of this relationship may be threatening to other regimes and create additional opportunities for conflict.

Foreign Policy Orientation and Internal Security

While the importance of international policy differences in crisis proclivity should not be underestimated, we should also highlight the role such differences play in the internal politics of these regimes and how these forces may make crisis initiation more likely. First, it stands to reason that the foreign policy utility some regimes might gain by becoming involved in a crisis with the U.S. would be more than offset by the improbability of winning such a dispute given the tremendous power of the U.S. and its global military reach. Some utility other than potential international foreign policy gains is likely making initiation of such crises an attractive policy option, especially in the case of adversaries. I suggest that the gains made by some regime leaders in maintaining power as a result of an adversarial relationship with the U.S. provide incentives for some states to initiate crises. As we have learned from research on ‘‘in-groups’’ and ‘‘out-groups,’’ there is increased solidarity among people in the face of an external threat (Simmel, 1955; Coser, 1956). To the extent that a regime can justify its own hold on power because of the need for strong and unified leadership given the ‘‘threat’’ posed by the U.S., the greater the enmity between that regime and the U.S., the more solid the regime’s control of the state. Thus, pursuing an adversarial relationship with the U.S. helps maintain the regime in power in order to ‘‘protect’’ the nation. One thinks of Cuba under Castro, Iraq under Saddam Hussein, Yugoslavia under Slobodan Milosevic, and North Korea under its Communist regime as states whose very identities were often defined by their adversarial relationship with the U.S., and whose regime leaders used the ‘‘U.S. threat’’ to bolster their hold on power. Thus, in addition to the foreign policy utility to be derived from challenging the U.S., there is domestic policy utility as well to be enjoyed, often regardless of the crisis outcome. States with substantial foreign policy differences should have a significantly higher degree of interest in initiating international crises that involve the U.S. Foreign policy orientation toward the U.S. will also influence domestic politics within a friendly regime. The American ties will no doubt serve some interests well, especially those whose power and/or wealth may depend on U.S. protection. We would expect that regime leaders who represent such interests will seek to maintain the U.S. security umbrella and rely upon it on occasion when their influence is threatened. But while such dependence upon a patron state may protect a regime from opposing factions within the state to some extent and for some time, the very closeness of the relationship may also provoke hostility among those disenfranchised by the current regime. The regime’s cozy relationship with the U.S. may also inspire complacency among the leaders and contempt for the opposition that encourages further unrest. The greater the perception that the U.S. is complicit in the policies (e.g., those involving human rights, the economy, politics) pursued by the

JAMES MEERNIK

169

regime that harm the interests of the opposition, the greater the likelihood the U.S. will be drawn into power struggles between the regime and the opposition. The U.S. may intervene to protect the regime in power or to insure a transfer of power to some group viewed as more favorable to its interests or less hostile than other groups (e.g., in Latin America throughout the Cold War). Nonetheless, the closeness of such ties may itself be a source of instability and crisis with the nation. Therefore, I expect a curvilinear relationship between foreign policy orientation and the likelihood of crisis initiation. Those states whose foreign policy orientation vis-a´-vis the U.S. is the most (dis)similar should be the most likely to initiate crises involving the U.S. Those states in the middle of this continuum should be the least likely to initiate crises. Hypothesis 1: The greater the difference/similarity in a state’s foreign policy orientation in comparison to the U.S., the greater the likelihood it will initiate a crisis involving the U.S.

When Crises Happen: Strategic Interaction and the Timing of Crisis Initiation The extant research on strategic interaction (Leeds and Davis, 1997; Smith, 1998; Meernik, 2000, 2001; Clark, 2003) is principally concerned with the effect the political and economic situation within the U.S. exercises on the willingness of foreign actors to instigate a crisis. Some argue that the probability of crisis initiation declines as the incentives for a diversionary use of force increase as the crisis initiator postpones action until the president is less likely to use force (Leeds and Davis, 1997; Clark, 2003). The literature on strategic interaction argues that the U.S. should be most likely to intervene in such crises when the president’s political fortunes and the national economy are deteriorating, or when national elections are approaching (Leeds and Davis, 1997; Clark, 2003 [but see Smith (1998), who qualifies such hypotheses as contingent upon the perceived and actual competence of the chief executive]). During such times, presidents would prefer to use force to distract the public from their domestic woes, rally support behind their leadership, and increase their chances of remaining in power. Yet, because foreign actors are aware of the incentives presidents have for using force during such periods, they refrain from acting and thus present few such opportunities for diversionary activity. On the contrary, it is also possible that foreign leaders may believe that when the president is suffering in the polls or the U.S. economy is deteriorating, he will be too consumed by his problems at home to be bothered with troubles abroad. Henry Kissinger commented on this opportunism of foreign leaders when he analyzed why the Soviets threatened to intervene in the 1973 Yom Kippur War during a time of domestic upheaval in the U.S. He writes, ‘‘I could not avoid the conviction that Nixon’s evident weakness over Watergate had not a little to do with the Politburo’s willingness to dare so crass a challenge’’ (Kissinger, 1982, p. 584). Furthermore, if we believe that foreign leaders prefer to initiate crises against incompetent leaders, with whom presumably their chances of success are greater (Smith, 1998), what better indication could they have than the American people’s own assessment of their chief executive? While foreign leaders may perceive that presidents possess certain incentives to use force for diversionary purposes when they are down in the polls or blamed for the ill health of the economy, on balance it may be that such presidents are better targets, especially for regimes whose interests are opposed to the U.S. and wish to avoid U.S. involvement. If the choice is between initiating a crisis when the president is incompetent, distracted, and weakened, and when the president is competent, focused, and strong, the former would seem to be preferable. Therefore, I posit two contrasting hypotheses to represent each of these theoretical arguments.

170


Hypothesis 2a: Strategic interaction: As domestic conditions in the U.S. worsen, the prob ability that a regime will initiate a crisis involving the U.S. will decrease. Hypothesis 2b: Exploiting weakness: As domestic conditions in the U.S. worsen, the probability that a regime will initiate a crisis involving the U.S. will increase.

Control Measures

Together with those determinants of crisis initiation discussed above, I also include two other factors in the model. First, I control for the military capabilities of states. Conventional international relations theories typically maintain that the greater the military advantage possessed by a state, contingent upon the utility to be derived from a successful challenge, the more likely that state is to initiate a crisis. As I am only examining the occurrence of crises relevant to the U.S., I utilize a measure of each foreign nation’s military capabilities. It is also important to account for the inherent proclivity of regimes to become involved in crises. As research on enduring rivalries as well as research on civil wars shows, the same regimes seem to account for a substantial proportion of these events (Goertz and Diehl, 1993, 1995; Bennett, 1998). Whether these crises are related, as some research contends, or whether their frequency can be accounted for simply by the laws of probability (Gartzke and Simon, 1999) is not an issue here. I am suggesting that there is either some inherent, perhaps unmeasured, propensity for regimes to become involved in crises, or a temporal dependence among crises such that once an initial crisis occurs, subsequent crises become more likely. In addition, research on the use of military force by the U.S. tends to show that the U.S. is more apt to take action in crises involving states it has targeted previously (Meernik, 1994, 2001). Once the U.S. begins a pattern of being drawn into crises by a stateFwhether friend or foeFthe tendency may be for such events to develop their own sort of selfperpetuating logic. Whether such crisis interaction continues because of the interests of the initiator state in reducing the threat posed by the U.S. or increasing the security provided by the U.S., or simply because the issues at stake have not been resolved, there is reason to believe that prior involvement in disputes with the U.S. will lead to future dispute involvement. Hypothesis 3: As the number of a regime’s military personnel increases, the probability of crisis initiation increases. Hypothesis 4: The more often presidents have used force in or against a regime in the previous five years, the more likely that nation will initiate a crisis.

Measurement I analyze the initiation of international crises by foreign governments that ultimately involve the U.S. Because I am interested in understanding which nations are more likely to initiate crises and when, the spatial universe of cases consists of all nations, while the temporal universe is potentially infinite. Regimes can choose to initiate a crisis at almost any time. I choose to limit the temporal dimension from which I predict crisis initiation, however, to the nation-year. Therefore, for all nations of the world for which data are available, I predict how many crises a regime initiates in any given year between January 1960 and December 2000. I choose this period because most of the new nations of the world became independent in the early 1960s after decolonization. Before this period there were substantially fewer ´-vis the U.S. I use the Militarized Interstate nations that might initiate crises vis-a

JAMES MEERNIK

171

Dispute (MID) data set1 (Ghosn and Palmer, 2003) to measure crisis initiation (targets of crisis initiators are not included). I considered using the International Crisis Behavior2 (Brecher and Wilkenfeld, 1997) data set in order to maximize the generalizability of this analysis, but because these data do not include many of the crises involving U.S. limited uses of force, I elected to use a more comprehensive data set (see also Fordham and Sarver, 2001). Because I am interested in whether a given nation is a crisis participant, I make use of the MID’s actor-level data. Every nation that is coded as a crisis initiator is included in the measure and all crises initiated by a nation each year are summed to create the dependent variable. I measure the foreign policy orientation of a regime using three measures. First I use data on voting similarity in the United Nations (UN) between a regime and the U.S. from Gartzke, Jo, and Tucker’s UN voting behavior data. 3 This variable ranges between ‘‘–1’’ (perfect disagreement) on General Assembly roll call votes to ‘‘1’’ (perfect agreement). This variable is a tau-b correlation. I expect that those regimes that exhibit the greatest differences in UN voting and those nations that exhibit the greatest similarity should be most likely to initiate crises with the U.S. I square this measure to capture the expected curvilinear relationship. The first coefficient should be negative, while the squared variable coefficient should be positive. Second, to measure the military alignment of a regime, I use Bueno de Mesquita’s Kendall tau-b indicator of the similarity of a regime’s alliance portfolio with the system’s leaderFin this case the U.S. This measure ranges from ‘‘1’’ to ‘‘ 1,’’ indicating more or less similarity in alliance membership. These data are obtained from the EUGENE Program (Bennett and Stam, 2003). I expect that those regimes that exhibit the greatest differences and those regimes that exhibit the greatest similarities with the U.S. will be the most likely to initiate crises. I also square this measure to capture the expected curvilinear relationship. The first coefficient should be negative, while the squared variable coefficient should be positive. Third, I use Polity IV data on democracy ratings to measure the extent to which a nation’s form of government resembles the U.S.’s. I expect that the more democratic the regime, the less likely it will initiate crises involving the U.S. I measure domestic conditions within the U.S. using three variables: (1) the president’s annual average approval rating; (2) the U.S. annual average unemployment rate; and (3) the U.S. annual average inflation rate.4 Presidential approval ratings are measured using data from the Gallup Polls and the Roper Polls for more recent years.5 Unemployment data are available from the Department of Labor’s Bureau of Labor Statistics web page. 6 Inflation data are from the Consumer Price Index data downloaded from the Department of Commerce’s Bureau of Economic Analysis web page.7 I measure a nation’s military capabilities using data from the Correlates of War Project’s National Materials Capabilities Index. Specifically, I measure military capabilities as the number of military personnel in the thousands and lag this variable one year in order to reduce the possibility that crisis initiation and size of the armed forces are reciprocally related. I measure prior U.S. uses of force directed toward À

1

Data are available at http://cow2.la.psu.edu/. Data are available at http://www.icbnet.org/Data/index.html. 3 These data can be found at http://www.vanderbilt.edu rtucker/data/affinity/un/similar/. 4 Presidential election periods perfectly predicted noncrisis months and so I was not able to use such a variable. Stata 8.0 forces out all such variables from these types of MLE models given the lack of variation in an exogenous variable across all levels of the endogenous variable. 5 I use presidential approval ratings obtained from the Gallup organization web site at http://www.gallup.com for all years to create a monthly average. Since accessing this web page originally, however, the Gallup organization has now begun to charge users for access to their data. Other organizations, however, such as Roper (http://roperweb. ropercenter.uconn.edu/), do provide these same data. 6 The website is http://stats.bls.gov/top20.html. 7 The website is http://www.bea.doc.gov/. 2

$

172


each nation by counting the number of crises involving a regime in the previous five years in which the U.S. used military force vis-a´-vis that regime. I expect a positive relationship. In the MID database, all instances where the U.S. scored at least a ‘‘3’’ on the Hostility Level measure (displays of force, uses of force, and war) are counted.

Analysis I utilize a random effects Poisson model using the program Stata 8.0. I use the Poisson model because the data are a count of the number of crises initiated by a nation each year. This model is preferred because of the cross-sectional element of the data and recurrence of individual nations throughout the period under study that may cause heteroskedasticity in the estimates. In general, the results of the models indicate that the type of regime where a potential crisis takes place matters a great deal more than when such a crisis might occur. We see in Table 1 that support for the hypotheses regarding the importance of foreign policy orientation is mixed. As evidenced in the negative coefficient for the military alignment variable and the positive coefficient for its squared measure, we find that those regimes whose military alignments are most dissimilar and most similar to the U.S. are most likely to initiate crises. The test of joint statistical significance does not, however, indicate that the coefficients are jointly distinguishable from zero. The squared measure of political orientation is not statistically significant, although the two coefficients are jointly statistically significant. Those states whose U.N. voting behavior is most divergent from the U.S. and those states whose interests are more similar are most likely to initiate crises involving the U.S. Either their divergent foreign policies lead to foreign policy differences and hostility, or their strategic compatibility leads to closer ties, which come with their own risks of crisis. Yet both types of regimes may come to depend upon the intensity of their relationship with the U.S. to maintain power, and thus may become even more likely to initiate crises to secure their dominance. Hostile regimes may depend upon an adversarial relationship with an external enemy to keep their domestic opponents united behind their leadership, while friendly states use their cooperative relationship with the U.S. to fend off challenges from domestic opponents and advance foreign policy interests against neighboring states. The more democratic a regime is, the less likely it is to initiate crises. In keeping with previous research on the international behavior of democratic and autocratic states, we find that the less democratic a nation, the more conflict prone it is (Benoit, 1996; Prins and Sprecher, 1999). Democracies tend to be peaceful with each other (Bueno de Mesquita and Lalman, 1992; Maoz and Russett, 1993). This finding also supports Huth’s (1998) research that finds major powers are more likely to intervene in conflicts when they share the same type of polity with a state that is threatened by a third, politically dissimilar nation. Although foreign policy orientation matters when predicting crisis initiation, the timing of international crises appears to bear little relationship to domestic conditions in the U.S. The lack of statistical significance among these coefficients provides little justification to believe that U.S. domestic conditions exercise a statistically discernible effect on crisis initiation. Rather, foreign regimes may time crisis involvement to exploit local opportunities for advantage that may have little or no relationship to the vicissitudes of American public opinion or U.S. macroeconomics. The potential to catch an opponent off guard or the desire to seize a military capability advantage may both dictate the specific timing of foreign crises. The number of military personnel in a nation is strongly related to crisis involvement. Nations that maintain sizable numbers of men and women in the armed forces are much more likely to become involved in crises. Either because they are seeking a military advantage, or because their neighbors fear their intentions, a sizable defense establishment leads to more crisis activity. Given the U.S.’s stated

173

JAMES MEERNIK TABLE 1.

Predicting Regime Crisis Initiation, 1960–2000 Random Effects Poisson Model

Variable

Coefficient

Political orientation Political orientation2 Military alignment Military alignment2 Regime polity U.S. prior uses of force Regime military personnel U.S. president’s popularity U.S. unemployment U.S. inflation Constant

À

À

À

À

À

À

N 4529. Wald w2 72.7, p o.0001. Jointly statistically significant (w2 6.67, p # Not jointly statistically significant.

0.4183 0.4582 0.4877 1.4348 0.0243 0.1587 0.0008 0.0054 0.0374 0.0208 1.9953

Standard Error

0.1745 0.3298 0.4737 0.8168 0.0085 0.0403 0.0001 0.0039 0.0254 0.0144 0.3233

t Statistic

p Value

2.40 1.39 1.03 1.76 2.86 3.93 4.48 1.37 1.47 1.44 6.17

.017 .165 .303 .079# .004 .000 .000 .172 .142 .149 .000

À

À

À

À

À

À

n

¼

¼

n

¼

¼

¼

.01).

o

desire to prevent the emergence of any power that might rival its own, its interest in such nations is to be expected.8 Prior uses of force by the U.S. involving a particular nation also increase the likelihood that the same nation will again initiate a crisis. Once the U.S. begins using force against a regime, that regime becomes more likely to become involved in additional crises and thus begins something akin to a conflict spiral between the two nations. The continual militarized disputes between the U.S. and Cuba in the 1960s, Libya in the 1980s, and Iraq and Yugoslavia in the 1990s provide ample evidence of this. Next, in order to provide a more rigorous test of the model, I reran the initial estimates using probit instead of Poisson analysis. The reader will recall that I used Poisson analysis because the endogenous variable is a count measure of the number of times states initiate crises each year. It would be useful, however, to determine if states might be divided simply into those that initiate crises in a given year, and those that do not. It may be that once a state crosses the threshold from being a quiescent member of the international community to a crisis initiator, the major and critical division between states has occurred. To investigate this possibility, I reran the model in Table 1 using random effects probit analysis. The estimates are provided in Table 2. Generally the results are similar to what we just saw although the joint tests of statistical significance for the military alignment measure of foreign policy orientation are statistically significant rather than the test for the U.N. voting measure. The indicators of democracy, military personnel, and prior U.S. crisis behavior vis-a´-vis the states remain statistically significant and in the predicted direction. The measures of the U.S. economyFinflation and unemploymentF remain statistically insignificant. The presidential popularity measure is, however, negative and statistically significant. States are more likely to initiate a crisis the lower the president’s popularity. The reader will recall that this analysis combines crises initiated by states whom the U.S. ultimately sides with and whom the U.S. sides against. Thus, it would be useful to distinguish the impacts of the exogenous factors on the propensity of these two types of states to initiate crises. Before proceeding, however, a word of caution is in order. We should not necessarily assume that states the U.S. sides with are friends and those it sides against are adversaries in the more general sense of those 8

The ‘‘United States National Security Strategy’’ report states that, ‘‘our forces will be strong enough to dissuade potential adversaries from pursuing a military build-up in hopes of surpassing or equaling the power of the United States.’’ http://www.state.gov/r/pa/ei/wh/15430.htm

174

Foreign Policy Orientation TABLE 2.

Predicting Regime Crisis Initiation, 1960–2000 Random Effects Probit Model

Variable

Coefficient


À

À

À

À

À

À

0.2403 0.3320 0.2356 1.0234 0.0165 0.1715 0.0006 0.0060 0.0211 0.0130 1.2841

Standard Error

0.1296 0.2508 0.3612 0.5983 0.0059 0.0420 0.0001 0.0029 0.0187 0.0102 0.2320

t Statistic

p Value

1.85 1.32 0.65 1.71 2.77 4.08 4.30 2.05 1.13 1.26 5.53

.064 .186# .514 .087 .006 .000 .000 .040 .258 .206 .000

À

À

À

À

À

À

n

N 4529. Wald w2 67.3, po.0001. Jointly statistically significant (w2 6.74, po.01). # Not jointly statistically significant. ¼

¼

n

¼

words. In any particular crisis, the U.S. may side temporarily with an adversary or some state with which it does not enjoy good relations for short-term, tactical reasons (as the U.S. occasionally sided with the U.S.S.R. or the PRC in opposition to the other during the Cold War). Similarly, the U.S. may side against a state with which it does enjoy good relations, perhaps because it opposes particular regime leaders (e.g., the Dominican Republic in 1965 and Panama in 1989 were both U.S. allies whom the U.S. invaded). In order to execute this analysis, I created two new endogenous variables. The first measures the number of crises initiated each year by states whom the U.S. ultimately sides with in these situations. The second measures the number of crises initiated each year by states that the U.S. ultimately sides against in these events. I term the former, ‘‘crisis allies,’’ and the latter, ‘‘crisis adversaries,’’ in order to avoid continual usage of the wordier description of such states. The unit of analysis for each set of estimates is the same as in Tables 1 and 2Fthe nation-state-year for all nations for which there are data in the period of 1960–2000. The estimates are provided in Tables 3 and 4. We see straightaway that distinguishing between crisis allies and crisis adversaries reveals important differences in the crisis propensity of both types of states. First, TABLE 3.

Predicting Regime Crisis Initiation by States Whom the U.S. Sides Against, 1960–2000 Random Effects Poisson Model

Variable

Coefficient


À

À

À

À

N 4529. Wald w2 34.9, po.0001. Jointly statistically significant (w2 Jointly statistically significant (w2

2.4285 3.1354 0.5202 3.1046 0.0748 0.2000 0.0007 0.0225 0.0248 0.1401 8.1036

¼

¼

n

nn

¼

16.9, po.01). 10.4, po.01).

¼

Standard Error

0.7106 1.4515 1.6135 2.6944 0.0377 0.1291 0.0007 0.0168 0.1255 0.0564 1.4224

t Statistic À

À

À

À

3.42 2.16 0.32 1.15 1.98 1.55 1.03 1.33 0.20 2.48 5.70

p Value

.001 .031 .747 .249 .048 .121 .304 .182 .843 .013 .000

n

nn

175

JAMES MEERNIK TABLE 4.

Predicting Regime Crisis Initiation by States Whom the U.S. Sides With, 1960–2000 Random Effects Poisson Model

Variable

Coefficient


À

À

À

À

À

3.4133 0.9990 0.1123 0.2532 0.0088 0.6892 0.0005 0.0552 0.3655 0.0914 4.7256

Standard Error

1.0208 1.3940 1.1322 1.6190 0.0259 0.0831 0.0002 0.0193 0.0931 0.0589 1.4208

t Statistic

p Value

3.34 0.72 0.10 0.16 0.34 8.29 2.37 2.85 3.92 1.55 3.33

.001 .474 .921 .876# .733 .000 .018 .004 .000 .121 .001

À

À

À

À

À

n

N 4529. Wald w2 165.7, po.0001. Jointly statistically significant (w2 17.1, po.01). # Not jointly statistically significant. ¼

¼

n

¼

however, we find that those states with whom the U.S. sides in crises are more likely to initiate such events the closer their foreign policy orientation to the U.S., as measured by UN voting similarity, like we found previously. Analogously, those states with which the U.S. sides against in crises are more likely to initiate such events the more different their UN voting patterns. Both sets of coefficients were jointly statistically significant in Tables 3 and 4. We also see that foreign policy orientation as measured by alliance similarities is jointly statistically significant in the case of states against whom the U.S. sides (Table 4), although given that neither coefficient alone is statistically significant, we should not put too much emphasis on these results. Interestingly, the Polity 4 ratings of democracy do not predict the propensity of crisis allies to initiate events, but do help explain the likelihood that states whom the U.S. sides against will initiate crises (Table 3). The less democratic the state, the more likely it will initiate a crisis that will involve the U.S. against it. Generally, the theory of foreign policy orientation is supported in these sets of analyses. The more political component of foreign policy orientation FUN voting similarityFprovides the better explanation of crisis initiation than the more military componentFalliance similarities. It may be that the UN voting data provide a more sensitive measure of foreign policy orientation that can better reflect evolving and short-term fluctuations in U.S. relationships with other states that help predict crisis onset. Interestingly, while the coefficients for both the lagged defense personnel measure and the previous U.S. crisis behavior indicator are positive in both sets of estimates, they are statistically significant only in the case of states that initiate crises whom the U.S. ultimately sides with (Table 4). States with large militaries and states with whom the U.S. has used military force previously are more likely to initiate crises in the present that ultimately involve the U.S. on their side. The latter finding is especially interesting for it seems to suggest that as the U.S. uses force more often in conjunction with some state, the more likely such states are to involve the U.S. in their future problems. One thinks of U.S. allies that come to depend upon U.S. support, such as Zaire, many Central American nations, and even West Germany during the Cold War. Israel, both during and after the Cold War, represents another possible illustration of this phenomenon. Thus, while states whom the U.S. has sided against in the past certainly provoke their share of international crises that involves the U.S., these enduring ‘‘rivals’’ may, in some respects, not be quite as numerous as the enduring ‘‘allies.’’

176


The impacts of U.S. domestic, political, and economic indicators on the crisis behavior of states also reveal some interesting tendencies. Crisis allies especially seem more likely to time the initiation of international incidents to coincide with difficult times for U.S. presidents. As U.S. unemployment increases and as presidential popularity decreases, states with which the U.S. ultimately sides are more likely to initiate crises (Table 4). Perhaps because such states would tend to seek U.S. involvement in a crisis in order to protect their interests, they initiate such events when they believe presidents will be more likely to respond with a diversionary use of force. This finding then provides some support for the strategic interaction hypothesis, although in a rather unexpected fashion. As most studies of strategic interaction and the diversionary use of force assume that the states that initiate crises are doing so against the U.S. or some other nation, the general expectation is that these crisis initiators wish to avoid a U.S. response. Few studies in this field of research,9 to my knowledge, consider the possibility that some states that initiate crises do so to encourage a U.S. response. This finding points to the vital importance of distinguishing between crisis allies and crisis adversaries and their differing preferences regarding the desirability of U.S. involvement. Yet, although crisis allies tend to seek U.S. involvement during times when the U.S. domestic political environment is conducive to a diversionary use of force, crisis adversaries do not evidence any clear and consistent preferences for particular U.S. political conditions in initiating crises. However, the coefficient for the measure of U.S. inflation is statistically significant and positive in Table 3 among states the U.S. sides against, which would indicate that such states are more likely to initiate events when the cost of living is increasing in the U.S. I would like to offer one possible explanation for the more general difference in results between the crisis allies and crisis adversaries. Perhaps the reason why crisis allies are more apt to strategically time crises is because they are better aware of U.S. domestic, political, and economic conditions. If many or most such states do tend to enjoy better relations with the U.S. in general, they may have more access to information about the U.S. political scene and have more experience in interpreting such information. On the other hand, if crisis adversaries tend not to enjoy good relations in general, the quality and quantity of their information about the U.S. political environment, and their sophistication in deciphering it, may be relatively deficient. Thus, its utility in decision making regarding the timing of crisis initiation may be limited. These findings about the differing crisis initiation behaviors between ‘‘crisis allies’’ and ‘‘crisis adversaries’’ call for increased and more focused research.

Conclusion Our understanding of U.S. foreign policy and the use of the military has improved substantially in recent years as more research has been conducted on the interrelationship between presidential decision making on the use of force, and the opportunities provided by the international environment. The present effort builds upon that research agenda and expands our understanding in several ways. First, we see that the spatial aspect of crisis onset matters more than the temporal element. Foreign policy orientation plays a substantial role in determining where conflicts arise, while domestic conditions in the U.S. matter less. Previous research on strategic interaction had highlighted the importance of these conditions, but had not tested to determine the influence exerted by the characteristics of foreign states in the initiation of crises. More specifically, we discover that those nations most similar and dissimilar to the U.S. are the most likely to become involved in crises. We must also be mindful of the role likely played by internal forces in both friendly 9

Certainly other studies, particularly those that focus on extended deterrence, study the propensity of states to bring their allied protectors into crises (see Huth and Russett, 1984, 1993; Huth, 1988b, 1998).

JAMES MEERNIK

177

and adversarial regimes. The intensity of the relationship with the U.S. may spark opposition in the former if regime opponents view the regime leaders as mere clients of the hegemon (one thinks of many of the insurgencies that developed in Latin America during the Cold War [especially Cuba in the late 1950s or Nicaragua in the 1970s], or present day Saudi Arabia). Regimes hostile to the U.S. may come to depend upon this adversarial relationship as a means to unite an otherwise hostile population. The relationship between domestic conditions within the U.S. and the initiation of international crises is more complex. When we do not distinguish between states the U.S. sides with and those it sides against in crises, U.S. domestic conditions appear to exercise little influence in producing more international incidents. Yet, when we do distinguish between these two types of states, we find that among states the U.S. ultimately sides with in crises, there is a marked propensity to initiate such crises when the prospects for what we might consider to be a diversionary use of force are greater. States against whom the U.S. sides do not exhibit such clear preferences. Ultimately, these findings demonstrate once again that future research on crisis initiation needs to focus much more on the types of states that initiate crises and their foreign policy interests. I see two concerns that ought to generate future research in this area. First, as just mentioned, we should investigate in more detail the political and military interests of foreign regimes that affect their timing of crisis initiation. We need a broader understanding of not just the types of foreign states that initiate crises (and how they are influenced by U.S. domestic conditions), but also why they elect to initiate crises at particular points in time for their own national interests. Our models that focus principally upon the impact of U.S. conditions on crisis timing are markedly limited in this regard. For example, we would suspect that many regimes’ leaders would wait until the military balance favors their challenge, or until they have sufficient support from their military (Dassel and Reinhardt, 1999). These possibilities deserve greater exploration. Second, we need to better understand why some states become involved in enduring crisis relationships with the U.S. whether as friends or enemies. Earlier I suggested that some states use their adversarial relationship with the U.S. to unite their government and citizens, while others use their friendly relationships to protect themselves from challengers. More research on the internal dynamics of crisis initiators and the utility they derive from confronting the U.S. needs to be undertaken in order to better understand their crisis initiation interests.

References BENNETT, D. SCOTT (1998) Integrating and Testing Models of Rivalry Duration. American Journal of Political Science 42:1200–1232. BENNETT, D. SCOTT, AND A LAN STAM (2003) Expected Utility Generation and Data Management Program. BENOIT, KENNETH (1996) Democracies Really Are More Pacific (in General): Reexamining Regime Type and War Involvement. Journal of Conflict Resolution 40:636–657. BRECHER, MICHAEL, AND JONATHAN WILKENFELD (1997) A Study of Crisis . Ann Arbor: University of Michigan Press. BUENO DE MESQUITA, BRUCE (1981) The War Trap . New Haven, CT: Yale University Press. BUENO DE MESQUITA, BRUCE, AND D AVID L ALMAN (1992) War and Reason. New Haven, CT: Yale University Press. CLARK, D AVID H. (2003) Can Strategic Interaction Divert Diversionary Behavior? A Model of U.S. Conflict Propensity. Journal of Politics 65:1013–1039. COSER, LEWIS A. (1956) The Functions of Social Conflict . New York: Free Press. D ASSEL, KURT, AND ERIC R EINHARDT (1999) Domestic Strife and the Initiation of Violence at Home and Abroad. American Journal of Political Science 43:56–85.

178


DER OUEN, K ARL (1995) The Indirect Link: Politics, the Economy and the Use of Force. Journal of Conflict Resolution 39:671–695. DER OUEN, K ARL (2000) Presidents and the Diversionary Use of Force: A Research Note. International Studies Quarterly 44:317–328. FORDHAM, BENJAMIN (1998a) The Politics of Threat Perception and the Use of Force: A Political Economy Model of U.S. Uses of Force, 1949–1994. International Studies Quarterly 42:567–590. FORDHAM, BENJAMIN (1998b) Partisanship, Macroeconomic Policy, and the U.S. Uses of Force, 1949– 1994. Journal of Conflict Resolution 42:418–439. FORDHAM, BENJAMIN, AND CHRISTOPHER C. S ARVER (2001) Militarized Interstate Disputes and United States Uses of Force. International Studies Quarterly 45:455–466. G ARTZKE, ERIC, AND MICHAEL W. SIMON (1999) ‘‘Hot Hand’’: A Critical Analysis of Enduring Rivalries. Journal of Politics 61:777–798. GHOSN, F ATEN, AND GLENN P ALMER (2003) Codebook for the Militarized Interstate Dispute Data , Version 3.0. Online: http://cow2.la.psu.edu GOERTZ, G ARY, AND P AUL F. DIEHL (1993) Enduring Rivalries: Theoretical Constructs and Empirical Patterns. International Studies Quarterly 37:147–171. GOERTZ, G ARY, AND P AUL F. DIEHL (1995) The Initiation and Termination of Enduring Rivalries: The Impact of Political Shocks. American Journal of Political Science 39:30–52. HUTH, P AUL K. (1988a) Extended Deterrence and the Prevention of War. New Haven, CT: Yale University Press. HUTH, P AUL K. (1988b) Extended Deterrence and the Outbreak of War. American Political Science Review 82:423–444. UTH, H P AUL K. (1998) Major Power Intervention in International Crises. Journal of Conflict Resolution 42:744–770. HUTH, P AUL, AND BRUCE R USSETT (1984) What Makes Deterrence Work? Cases from 1900 to 1980. World Politics 36:496–526. HUTH, P AUL, AND BRUCE R USSETT (1993) General Deterrence Between Enduring Rivals: Testing Three Competing Models. American Political Science Review 87:61–73. J AMES, P ATRICK, AND A THANASIOS HRISTOULAS (1994) Domestic Politics and Foreign Policy: Evaluating a Model of Crisis Activity for the United States. The Journal of Politics 56:327–348. J AMES, P ATRICK, AND JOHN R. ONEAL (1991) Influences on the President’s Use of Force. Journal of Conflict Resolution 35:307–332. J AMES, P ATRICK, AND SEBASTIAN R IOUX (1998) International Crises and Linkage Politics: The Experiences of the United States. Political Research Quarterly 51:781–812. KISSINGER, HENRY (1982) Years of Upheaval . Boston: Little, Brown and Company. LEEDS, BRETT A SHLEY, AND D AVID D AVIS (1997) Domestic Political Vulnerability and International Disputes. Journal of Conflict Resolution 41:814–834. M AOZ, ZEEV, AND BRUCE R USSETT (1993) Normative and Structural Causes of Democratic Peace 1946– 1986. American Political Science Review 87:624–638. MEERNIK, J AMES D. (1994) Presidential Decision Making and the Political Use of Military Force. International Studies Quarterly 38:121–138. MEERNIK, J AMES D. (2000) Modeling International Crises and the Political Use of Military Force by the United States. Journal of Peace Research 37:547–562. MEERNIK, J AMES D. (2001) Domestic Politics and the Political Use of Military Force by the United States. Political Research Quarterly 54:889–904. EERNIK, M J AMES D., AND PETER W ATERMAN (1996) The Myth of the Diversionary Use of Force by American Presidents. Political Research Quarterly 49:573–590. MORGAN, T. CLIFTON, AND KENNETH N. BICKERS (1992) Domestic Discontent and the External Use of Force. Journal of Conflict Resolution 36:25–52. OSTROM, CHARLES W. JR ., AND BRIAN L. JOB (1986) The President and the Political Use of Force. American Political Science Review 80:541–566. PRINS, B. C., AND CHRISTOPHER SPRECHER (1999) Institutional Constraints, Political Opposition, and Interstate Dispute Escalation: Evidence from Parliamentary Systems, 1946–89. Journal of Peace Research 36:271–287. R EED, WILLIAM (2000) A Unified Model of Conflict Onset and Escalation. American Journal of Political Science 44:84–93. SIMMEL, GEORG (1955) Conflict and the Web of Group Affiliations . New York: Free Press. SMITH, A LASTAIR (1996a) Diversionary Foreign Policy in Democratic Systems. International Studies Quarterly 40:133–153. SMITH, A LASTAIR (1996b) To Intervene or Not to Intervene. Journal of Conflict Resolution 40:16–40.

JAMES MEERNIK

179

SMITH, A LASTAIR (1998) International Crises and Domestic Politics. American Political Science Review 92:623–638. W ANG, KEVIN H. (1996) Presidential Responses to Foreign Policy Crises. Journal of Conflict Resolution 40:68–97. YOON, MI YUNG (1997) Explaining U.S. Intervention in Third World Internal Wars, 1945–1989. Journal of Conflict Resolution 41:580–602.

180

Foreign Policy Orientation, Strategic Interaction, And the Initiation of International Crises

Recommend Documents