Interstate rivalry,genocide,and politicide

Abstract

Interstate rivalry not only influences a country’s international behavior, but also its domestic conduct. Here, I focus on the connection between interstate rivalry and domestic government mass killing, specifically genocide and politicide. I argue that interstate rivalry has both direct and indirect influences on a government’s decision to use mass violence against its civilian population. Directly, countries engaged in rivalry experience a heightened state of military tension, which increases the likelihood that the country will resort to political mass killing when handling domestic dissent. Indirectly, rivalry increases the likelihood of both inter- and intrastate conflict, which also increases the likelihood of genocide and politicide. Statistical analysis of all country-years from 1955 to 2011 reveals that interstate rivals are more likely to engage in genocide and politicide than are other states. This research illustrates the way in which interstate rivalry influences a state’s domestic politics and shapes the interactions between government and population. It also highlights the importance of how the international threat environment affects a state’s willingness to engage in domestic political mass murder. These findings indicate that rivals do not only engage in the most violent interstate behavior, but also some of the deadliest domestic politics, as well.

Keywords

civil war genocide mass killing political violence politicide rivalry war

Introduction

Why are some governments more likely to engage in genocide or politicide than others? Several scholars have shown that different forms of violence are interrelated (e.g. Dugan & Chenoweth, 2012; Findley & Young, 2012; Thomas, 2014; Avdan & Uzonyi, 2017). Much of this literature has focused on the connection between different forms of domestic violence. Here, I build from these significant insights to establish the connection between international and domestic political violence. Considering the relationship between international and domestic political violence is important because it allows us to understand how actors propagate, perpetuate, sustain, or overcome problems of commitment, indivisibility, and information at different levels of a state’s political interactions. Additionally, considering these connections provides a more holistic understanding of political violence, as domestic and international violence may share an underlying logic (e.g. Cunningham & Lemke, 2013).

To make my argument, I focus on the connection between interstate rivalry and domestic government mass killing, specifically genocide and politicide.¹ Here, I focus on what Thompson (2001: 560) calls ‘strategic rivals’. These are states that view each other as (1) a competitor, (2) a source of actual or latent military threat, and (3) an enemy. I focus on such rivals because these states are the most likely to use interstate violence and thus are a powerful frame through which to understand international relations (e.g. Goertz & Diehl, 1993, 1995). I focus on state mass murder because genocide and politicide are the most severe forms of domestic violence, making them important to understand so that the international community can better prevent such violence. Additionally, as such atrocities tend to be final policy solutions with high psychological and political hurdles to enacting (see Krain, 1997; Valentino, 2004), genocide and politicide provide a hard test for analyzing how rivalry shapes domestic political behavior.

I argue that rivals are more likely than other states to commit genocide and politicide domestically because they are on heightened military guard given the international threat they face. Heightened military guard results in the military being ready to mobilize, existing in a state of anxiety, and perceiving the need to eradicate domestic threat so as to focus on international threat. Econometric analysis of the occurrence of all genocides and politicides between 1955 and 2011 supports this argument and reveals that interstate rivalry has both direct and indirect effects on the likelihood of political mass murder. These findings are consistent with several episodes of genocide and politicide, such as those in Argentina (1976–80), El Salvador (1980–89) and the Democratic Republic of Congo (1964–79), in which interstate tensions mixed with domestic threat to create conditions ripe for state-sponsored mass killing. These findings indicate that rivals do not only engage in the most violent interstate behavior, but also some of the deadliest domestic politics, as well.

This work thus highlights the domestic consequences of international threat (e.g. Hutchison & Gibler, 2007; Wright, 2014), and makes three contributions to our understanding of both domestic and international political violence. First, it contributes to the interstate rivalry literature by highlighting a significant way in which rivalry influences domestic politics, revealing that interstate tension can influence domestic violence.² This point is important because it highlights a potential feedback loop in which rivalry increases domestic conflict and domestic conflict may help to further fuel the rivalry. Second, it also contributes to the genocide and politicide literature by showing how international conditions can influence government mass killing. While the genocide and politicide literature has focused on the role of interstate war in providing the conditions for genocide and politicide, it has ignored the influence of international tensions more broadly. Third, it is the first work providing a theory that brings both direct and indirect causes of genocide and politicide together. Previous scholars have not explored the connection between these two types of causes of such mass killings (e.g. Krain, 1997; Harff, 2003; Shaw, 2007).

The influence of interstate rivalry on state behavior

A number of scholars have shown that interstate rivals behave differently in the international sphere than their non-rival peers. For instance, rivals are more likely to become locked in an arms race than are other pairs of states (e.g. Rider, Findley & Diehl, 2011) and support insurgents (Salehyan, Gleditsch & Cunningham, 2011) and terrorists (Findley, Piazza & Young, 2012) fighting against their enemy. Rivals are also more willing to use diversionary force against one another to pacify domestic unrest (e.g. Mitchell & Prins, 2004). Additionally, rivals react to exogenous factors, such as capability balances, in different ways than non-rivals (Thompson, 2001; Rasler & Thompson, 2000; Colaresi & Thompson, 2002). Ultimately, these factors result in interstate rivals tending to fight more often than other states (e.g. Goertz & Diehl, 1993, 1995), especially over territorial issues (Lektzian, Prins & Souva, 2010), and becoming locked into a cycle of militarized disputes and war (e.g. Bennett, 1996, 1997; Goertz & Diehl, 1995). However, rivals are also more likely to seek binding international arbitration to help settle their disputes than are other contentious dyads (Allee & Huth, 2006; Gent & Shannon, 2011). Combined, these various findings have produced a robust literature on the international behavior of rival states.

Scholars have also shown that interstate rivalry affects the domestic process of the state along with its international behavior. For example, Vasquez (1993) shows that the presence of an international rival creates conditions within a state that favor hawkish politicians. When rivalry exists, domestic audiences tend to oppose their governments providing unreciprocated cooperation to the state’s rival (Colaresi, 2004). Instead, they become more likely to elect hardline politicians who reinforce the rivalry (Deutsch & Merritt, 1965; Mansbach & Vasquez, 1981; Singer, 1982; Snyder, 1989; Vasquez, 1985, 1993). Domestic demand for the continuance of the rivalry is most heightened when national identity is fused with the issue at stake, such as territorial claims in the Argentine–Chilean rivalry (Thies, 2001). However, as rivalries intensify and the enemies experience a high-concentration of military conflicts, public support for the continuance of the rivalry dissipates, forcing politicians to seek peace with their enemy (Morey, 2011).³ This work has great merit in showing the mechanisms through which rivalry influences domestic politics, which in turn fuel the continuation of the rivalry at the international level.

However, little attention has been paid to the manner in which international rivalry can influence the ways in which the state uses violence against its own population.⁴ In an effort to rectify this important lacuna, I build from those scholars who use rivalry as a powerful frame through which to analyze interstate conflict, and turn my attention to domestic political violence. As interstate rivalry places the government in a state of heightened military aggressiveness (e.g. Snyder, 1989; Vasquez, 1985, 1993) and creates conditions of unrest and conflict within the country, I argue that states engaged in interstate rivalry are more likely to perpetrate policies of genocide and politicide against their domestic populations to control domestic dissent. Eliminating domestic threat quickly allows the state to better focus on overcoming its international threat.

The role of interstate rivalry in genocide and politicide

Governments must possess both motivation and opportunity to engage in genocide and politicide. While specific motivations may vary across cases, scholars tend to agree that these motivations are most likely to instigate mass atrocity when a government is particularly fearful of failure and desperate to maintain power (e.g. Chalk & Jonassohn, 1990; Kuper, 1981; Melson, 1992; Staub, 1989). Valentino (2004) posits that such motivation must be present because of the higher psychological and political barriers to perpetrating such mass violence compared to less severe human rights abuses. Given these higher barriers to action, governments are unlikely to engage in genocide or politicide unless they face extreme threat to their power. Krain (1997) calls such threatening situations ‘big opportunities’ for genocide and politicide. Big opportunities for violence tend to be situations of militarized threat, such as civil war (e.g. Harff, 2003; Valentino, Huth & Balch-Lindsay, 2004) or interstate war (Midlarsky, 2005; Wayman & Tago, 2010). Such big opportunities for atrocity also provide the government further opportunity for genocide and politicide because they provide some cover for the government to engage in mass killing against undesired groups, as the atrocities may be somewhat hidden among the violence of war (e.g. Horowitz, 1976).

The scholarship on ‘big opportunities’ for violence tends to focus on how such conditions have a direct effect on a government’s use of atrocities. Focusing on interstate rivalry, I broaden this perspective to consider how international threat and competition can have both direct and indirect influences on the likelihood of government mass murder. Specifically, I contend that interstate rivalry increases the likelihood that a state will perpetrate genocide and politicide by altering a regime’s motivation and opportunity for political mass murder through both direct and indirect effects. Directly, rivalry influences the likelihood of mass killing by increasing the level of perceived threat within the regime, bringing hawkish leaders to power and making these political elites more willing to use violence to maintain power, even if such threats are imagined or greatly exaggerated (Semelin, 2013). Indirectly, rivalry increases the likelihood of genocide and politicide through creating conditions that give rise to domestic and international conflict that the increasingly fearful elites will react to with severe violence. In these ways, interstate rivalry creates both the motivation (fear) and opportunity (conflict) for governments to commit mass atrocities against their domestic populations. Given these dynamics, many states engaged in interstate rivalry have perpetrated mass killings against their domestic populations. For example, in the 1960s, the Democratic Republic of Congo engaged in mass killings of its population while locked in rivalry with Congo-Brazzaville; in the 1970s, Pakistan engaged in politicide while in rivalry with India and Afghanistan; in the 1980s, El Salvador perpetrated politicide while engaged in a rivalry with Honduras; and in the 1990s, Sudan conducted a campaign of political killing while involved in rivalries with Uganda, Ethiopia, and Eritrea.

Figure 1 summarizes the process through which interstate rivalry influences the likelihood of genocide and politicide within a state through direct and indirect means. It is important to note, as illustrated in Figure 1, that rivalry has a distinct direct effect on the likelihood of genocide and politicide outside of its indirect effect on such atrocities through the causal pathways of civil and

Figure 1.

Causal pathway between rivalry and mass killing

interstate war. It is this unique effect of rivalry that helps set the stage for domestic government mass violence.

The direct effect: Fear as motivation

I now consider these mechanisms in more detail, beginning with the manner in which rivalry provides a direct effect on government mass murder by creating motive for the regime to commit such killing. The high political, psychological, and moral costs of these atrocities demand that political elites must have strong motivation to perpetrate mass murder. Interstate rivalry provides this motivation and can help explain why some states are willing to resort to genocide and politicide, while many others are not. The constant fear of threat and violence that hangs over a state engaged in interstate rivalry creates a situation high in distrust and tension within the state (Thompson, 2001). While much of this distrust is directed at a state’s rival, such fear may seep into the domestic political arena, as the government begins to perceive increasing threat to its stability and power. As international threat mounts, the government begins to fear that some domestic groups may support the enemy. It is important to note that pre-existing active opposition of the regime – and support of the rival – is not required for fearful elites to imagine that such threat exists. Nor are salient pre-existing cleavages necessary. Instead, the elites may simply perceive non-active support for the regime as tacit support of the enemy, or the government may invent or exaggerate imagined divisions among groups in society in an effort to consolidate power (Semelin, 2013). The likelihood that the government sees threat from groups throughout society increases with the level of international threat it faces (Melson, 1992). This can lead to domestic in-group/out-group dynamics, as the government looks for domestic opposition to its stability and moves to eliminate this potentially threatening out-group (Mylonas, 2012; Wright, 2014). Note that while previous scholars highlight the role of real or imagined security concerns in the genocidal process, they usually focus on features of the domestic out-group (Valentino, 2004; Downes, 2008). Here, I emphasize that the regime’s anxieties and fears come from the international environment in which the state operates – interstate rivalry – instead of the features of the out-group and social cleavages.

While rivalry may create a rally ‘round the flag’ (e.g. Colaresi, 2005), such support is short-lived as war costs mount (Morey, 2011). Midlarsky (2005) notes that since governments tend to be risk averse, political elites may often perceive threat to their power to be more severe than it is in reality. It is this perceived threat that drives political mass murder (e.g. Chalk & Jonassohn, 1990; Kelman, 1997; Shaw, 2007). As a government comes to perceive that threat to its power is high, and the costs of perpetrating genocide and politicide come to appear low relative to the benefits of remaining in power, the government’s likelihood of resorting to mass killing increases in an effort to maintain control of the state. Valentino (2004) notes that in this calculation, the government must perceive that repression alone can no longer sustain the regime’s stability. Interstate rivalry helps the government form this perception because rivalry adds a layer of interstate threat on top of domestic instability. According to Bulutgil (2015), such psychological mechanisms are even more important to the dynamics of wartime atrocity than the regime’s larger military aims. Rivalry thus directly helps provide the government incentive to eradicate domestic threat (perceived or real) so as to focus on international threat.

Such perceived threat becomes increasingly likely to lead to genocide and politicide when the military gains a stronger position within the state. As military officials tend to interpret threat in terms of state security and military matters, governments in which the military possesses an important role in helping craft policy are more likely to use violence to squash threat (e.g. Barnes, 2006; Cribb, 2001; Mason & Krane, 1989). Interstate rivalry creates a condition in which hawkish leaders gain more power in the regime and the military becomes more influential in helping craft state policy (e.g. Deutsch & Merritt, 1965; Mansbach & Vasquez, 1981; Singer, 1982; Snyder, 1989; Vasquez, 1985, 1993). As the regime comes to place a higher value on military strength and heightened military aggressiveness, political elites within the government become more willing to use violence to end the threats they face or perceive and to outbid one another in an attempt to promote their own interests (Colaresi, 2005), as was the case in 1970s Argentina (Scheina, 2003) and 1980s El Salvador (e.g. Stanley, 1996; UN, 1993). As interstate rivalry raises the profile and position of the military, while simultaneously creating tension, unrest, and perceived threat throughout the state, the likelihood of genocide and politicide should increase. Thus, since rivalry directly shifts perceptions, fear, and a willingness to use violence among the regime:

Hypothesis 1: Interstate rivalry directly increases the probability that a country engages in genocide and politicide.

The indirect effect: Conflict as opportunity

While rivalry provides motive for mass violence, elites also need opportunity to engage in genocide and politicide. I argue that rivalry also helps provide this opportunity and additionally has an indirect effect on mass killing. Recent research on genocide and politicide indicates that these atrocities are strategic tools used by governments to maintain power when facing threat to their power, especially during civil and interstate war (Krain, 1997; Midlarsky, 2005; Rost, 2013; Wayman & Tago, 2010). Rivalry indirectly influences the likelihood of genocide and politicide by creating conditions that give rise to domestic and international conflict, which increase the probability that a government will engage in such atrocities.

As reviewed above, since rivals react more harshly to external shocks, arm against one another, and promote instability in their enemy, rival states are more likely to engage in interstate war than are other states (e.g. Goertz & Diehl, 1993, 1995). The occurrence of interstate war raises the likelihood of genocide and politicide because conflict increases a regime’s loss aversion (Midlarsky, 2005). Regimes that fear loss from a conflict become more likely to commit domestic killing against groups that they perceive to support the enemy and cause such loss (Straus, 2015). Eliminating these groups allows the government to compensate its supporters with the possessions of the unwanted group and helps the government safeguard itself from any chance that the group may rebel in the future. Additionally, regimes that find it difficult to restore their stability following war may perpetrate mass murder in an effort to maintain power (Krain, 1997). For example, upon losing its 1971 war against its rival India, and the resulting independence of Bangladesh, an unstable Pakistani government became quick to use mass killings as a tactic to deal with further opposition groups agitating for independence, such as the Balochi (Marshall, Gurr & Harff, 2015; Sarkees & Wayman, 2010). By increasing the risk of interstate war, rivalry indirectly increases the likelihood of genocide and politicide.

However, there are further indirect effects, as well. Given this higher rate of interstate conflict and tension, rivalry leads to increased demands on a state’s economy and population. To perpetuate its interstate competition, a state will need to devote an increasing portion of its economy to military projects and arms races (e.g. Powell, 1993; Rider, 2009; Whitten & Williams, 2011). This money is diverted from infrastructure and social spending projects, which can slow the economy and hinder the welfare of the population (Goertz, Jones & Diehl, 2005). Furthermore, conscription may increase to provide the necessary military personnel should war occur. These developments can lead to unrest within the state’s population as public spending provisions falter and individuals are forced into military service (e.g. Rosecrance, 1987; Morey, 2011). For example, El Salvador’s heavy military investments during its rivalry with Honduras in the 1970s exacerbated socio-economic inequalities and slowed economic growth in the country, which led to massive protests and general unrest throughout El Salvador (Haggerty, 1988; Williams & Walter, 1997).

Unrest becomes dangerous to the regime as it transforms into armed rebellion. As a state contends with its rival, the government may not be able to successfully buy off members of the opposition group (e.g. Walter, 2006) or maintain power through repression alone, increasing the likelihood that the unrest develops into civil war. In such a situation where the government faces civil war and remains constricted by its interstate rivalry, it is likely that the civil conflict will become protracted, increasing costs to the government and making the government more desperate to end the fighting quickly. In such a situation, mass killing becomes attractive to the government as a strategy to quickly eradicate the opposition and its supporters to maintain power (Downes, 2008), as the Salvadorian government did by the mid-1980s (Mason & Krane, 1989) and the Pakistani government did throughout the 1970s (Marshall, Gurr & Harff, 2015).

Domestic mass killing may temporarily detract from military efforts to compel or deter the state’s rival by redistributing troops to the domestic front. However, such a strategy may appear to the regime to be effective in the long term if it allows the state to quickly quell the unrest to more fully focus on the interstate threat. As the threat of interstate war mixes with civil war, the state faces ‘big opportunities’ for government-sponsored mass killing and becomes more likely to resort to genocide and politicide (Krain, 1997). Thus, rivalry indirectly increases the likelihood of genocide and politicide by creating conditions of domestic conflict, as well as interstate war, which provides the government opportunity to commit these mass atrocities.

Hypothesis 2: Through an increased likelihood of civil and interstate war, interstate rivalry indirectly increases the probability that a country engages in genocide and politicide.

Research design

Unit of analysis

I argue that the heightened military environment of a country engaged in an interstate rivalry increases the likelihood that the government will perpetrate mass killing campaigns against its domestic population. To test this argument, I construct a dataset for all states from 1955 to 2011 in which the unit of analysis is the country-year. Data availability on genocides and politicides (Marshall, Gurr & Harff, 2015) limits the start of the analysis to 1955, whereas data availability on interstate rivalry (Thompson & Dreyer, 2011) limits conclusion of the analysis to 2011.

Dependent variable and estimator

The dependent variable in this analysis is Genocide/Politicide and is drawn from Marshall, Gurr & Harff (2015) who follow the definitions and guidelines of Harff (2003) to identify and code episodes of genocide and politicide. Harff (2003) codes the occurrence of mass killing based on five criteria: (1) complicity by the state in actions undertaken that endanger human life; (2) intent of authorities to isolate group members for mistreatment; (3) victims are members of an identifiable group; (4) the government’s policies cause prolonged mass suffering to the victims; and (5) the government’s actions pose a threat to the survival of the group. These criteria emphasize intentional killing to distinguish mass killings from accidental deaths. I assign an indicator variable of 1 for each year in which a state engages in genocide and/or politicide; 0 for all other years. Given the dichotomous nature of my dependent variable, I employ a logistic regression to test my hypotheses. The Online appendix lists the cases of mass killings coded by Marshall, Gurr & Harff (2015).

Independent variables

Rivalry

The key independent variable for my analysis captures whether or not a state is engaged in a strategic interstate rivalry each year. Given that my theory focuses on a state’s perceptions and fears about its strategic rivals, I use Thompson & Dreyer’s (2011) data that employ Thompson’s (2001) coding of rivals, which is based on state perceptions and requires each state to view each other as (1) a competitor, (2) a source of actual or latent military threat, and (3) an enemy. States that are engaged in at least 1 interstate rivalry in a given year are assigned an indicator variable of 1; all others are 0. I conduct additional analyses using the count of rival militarized interstate disputes (MIDS), which are threats, shows, displays, or uses of force, and the rivalry lifecycle in place of this indicator variable (see Rider et al., 2011). Each rivalry variable is lagged one year to establish the sequence of causality within the data.

Given my theoretical discussion of rivalry’s indirect effect on genocide and politicide, I also need to include measures of civil and interstate war. As civil and interstate war are theorized to be intervening variables, measures of these conflicts are only included in the mediation analysis of the indirect effects and not in the main analysis, except in a final model to show the robustness of the rivalry variable. Each of these variables is also lagged one year.

Civil war

Governments engaged in civil war are more likely to perpetrate genocide and politicide than are other governments (e.g. Krain, 1997; Melson, 1992). Additionally, civil war may force a state to redirect its resources away from its interstate rivalry to fighting its domestic enemies. This may be especially true if the insurgent movement is supported by the state’s rival (Salehyan, Gleditsch & Cunningham, 2011). To capture the effect of civil war, I include an indicator variable of 1 to each state fighting a civil war in a given year; 0 to all others from the Correlates of War Project (Sarkees & Wayman, 2010).

Interstate war

Rivals are more likely to engage in interstate war with each other than are other states (e.g. Goertz & Diehl, 1993, 1995). Interstate war provides governments both the incentive and the opportunity to engage in genocide and politicide to maintain control domestically (Colaresi & Carey, 2008; Krain, 1997; Kuper, 1981; Midlarsky, 2005; Rost, 2013; Wayman & Tago, 2010). To capture the effect of interstate war, I include an indicator variable of 1 for each state that is engaged in interstate war each year; 0 to all others from the Correlates of War Project (Sarkees & Wayman, 2010).

Since other political and demographic features of a country may correlate with the government’s willingness to engage in mass killing each year, and better explain the relationship between rivalry, genocide, and politicide, I include six additional variables to capture these effects and control for alternative or confounding arguments. Each variable, other than the time polynomial, is lagged one year. The Online appendix provides descriptive statistics for all variables in my models. Analysis reveals that multicollinearity is not a significant issue for these data, as the variance inflation factor for each variable is less than 2.

GDP_pc

Rivalry can divert money and manpower from the economy (Goertz, Jones & Diehl, 2005; Rosecrance, 1987). Poor governments that lack capacity are more constrained in their ability to handle domestic threats and international rivals simultaneously than are wealthier states. As poorer states are more likely to face domestic opposition and fail when attempting to suppress opposition, poorer states are more likely to resort to genocide and politicide to maintain domestic control (e.g. Freeman, 1991). Furthermore, poorer states are unable to pursue as wide a range of possible alternative policies to placate domestic opposition groups as are their wealthier counterparts. They are therefore more likely to resort to violence to handle domestic threat. To capture the effects of state wealth, I include the natural log of each state’s gross domestic product per capita (Gleditsch, 2002).

Democracy

Democratic leaders must be careful to avoid costly interstate conflict (e.g. Bueno de Mesquita & Siverson, 1995) and the use of domestic violence (e.g. Hegre et al., 2001) or they are likely to be removed from power. These incentives, coupled with legal constraints on a democratic executive’s power, decrease the likelihood that a democracy will engage in genocide or politicide (e.g. Harff, 2003; Rummel, 1995). To capture democracy, I use each state’s –10 (high autocratic) to 10 (highly democratic) Polity IV score (Marshall, Gurr & Jaggers, 2013). As a state’s Polity IV score increases, and the state becomes more democratic, the state will be less likely to experience genocide or politicide.

Exclusionary ideology

Ideology shapes a government’s goals. Some governments possess exclusionary ideologies that call for the eradication of specific domestic or international groups (see Fein, 1993). Therefore, I use Harff’s (2003) data on exclusionary ideologies to control for whether each government subscribes to ‘a belief system that identifies some overriding purpose or principle that justifies efforts to restrict, persecute, or eliminate certain categories of people’. Regimes with such exclusionary ideologies are assigned an indicator variable of 1; 0 otherwise. Harff’s original data only extend to 2001. I therefore follow her coding rules to extend these data to 2011.⁵

Population size

Larger populations provide greater threat to the government and thus provide more opportunities for conflict and mass killing (Besancon, 2005). Therefore, I include the natural log of each country’s population size (World Bank, 2015).

Coup

Coups can also threaten a government by targeting the political elites directly with the use of their military’s forces. When coups are successful, the new elites may use mass killings to eradicate supporters of the former regime. When coups are unsuccessful, the incumbent elites may use mass killing to eradicate supporters of the coup leaders (See Wayman & Tago, 2010). For this reason, I use Marshall & Marshall’s (2016) data to include an indicator variable of 1 for each country-year that experienced a coup; 0 otherwise.

Time polynomial

As genocides and politicides are likely to reoccur in states that have experienced such killing (e.g. Harff, 2003), I control for temporal dependency in the data by including the number of years, years squared, and years cubed since the last occurrence of mass killing within a state (Carter & Signorino, 2010). The time polynomial begins at 0 for the first year each country enters the dataset. It counts the number of ‘peace years’ until the country experiences genocide/politicide or the dataset ends.

Raw data

Before reviewing the results of my econometric analysis, I begin by considering the raw data. Table I displays the cross-tabulation between Rivalry and Genocide/Politicide. As Table I displays, states engaged in interstate rivalries

Table I.

Cross-tabulation of rivalry and genocide/politicide

	No genocide/politicide	Genocide/politicide	Total
Non-rivalry	5,145	69	5,214
Rivalry	3,176	205	3,381
Total	8,321	274	8,595

Chi² = 149.307; p = <0.000.

have nearly 200% more state-years experiencing government-sponsored mass killing than do non-rivalry states. Of the 274 observations of genocide and politicide, 205 (75%) occur in states engaged in interstate rivalry. A Chi square test indicates that these differences are statistically significant (Chi² = 149.307; p = 0.000). These patterns in the raw data strongly support my claim that rivalry is correlated with a higher likelihood of genocide and politicide.

Analysis of rivalry’s direct effect

Table II displays the results of my main econometric analysis. I begin by investigating rivalry’s direct effect on the occurrence of genocide and politicide (Hypothesis 1) before comparing this effect to rivalry’s indirect effect on episodes of government mass murder (Hypothesis 2). Model 1 is a parsimonious model that includes only my key variable of interest, Rivalry, and the time polynomial to control for time dependence in the data. The logistic regression provides significant support for my argument. States engaged in interstate rivalry are nearly 170% more likely to perpetrate mass killing than are non-rivalry states. Model 2 introduces the five additional control variables, discussed above. As the results of Model 2 indicate, interstate rivalry increases the likelihood of genocide and politicide by nearly 80% after including these control variables. These models provide strong support for my argument that interstate rivalry increases the likelihood of genocide and politicide, directly (Hypothesis 1).

To gauge the sensitivity of my analysis to various model specifications, I estimate five robustness checks on Model 2. First, as mass killings occur in only 3% of country-years, I re-estimate Model 2 using a rare events logistic regression (King & Zeng, 2001). As Model 3 indicates, my results are robust to this estimator. Second, to determine whether any one country is driving my findings, I re-estimate Model 2 using a jackknife robustness technique in which each country is systematically removed from the estimation. Model 4 indicates that no one country is driving these results. Third, to determine whether unobserved country effects are influencing the results of my analysis, I re-estimate Model 2 using country fixed effects. Model 5 indicates that my results are robust to the inclusion of these fixed effects. Fourth, while I theorize civil and interstate war to be intervening variables on the relationship between rivalry and mass killing, I include measures of each in Model 6 to gauge the robustness of my rivalry variable to their inclusion. As Model 6 indicates, my results remain consistent. Lastly, Downes (2008) argues that as civil wars becomes protracted due to guerrilla warfare – where a rival may support the weak rebels – a government becomes more desperate to end the fighting and thus more willing to kill civilians to help raise costs on the enemy and end the war. In Model 7, I include a measure of guerrilla warfare from Banks (2010) and find that my results remain consistent.⁶

The six models containing the control variables also provide support for existing explanations of the causes of genocide and politicide. I find that larger population size (e.g. Besancon, 2005) is associated with a higher likelihood of genocide and politicide in Model 5. Similarly, in Models 6 and 7, I find that exclusionary ideology (e.g. Harff, 2003), interstate war (Colaresi & Carey, 2008; Krain, 1997; Kuper, 1981; Midlarsky, 2005; Rost, 2013), and civil war (Krain, 1997; Melson, 1992)⁷ each increase the likelihood of genocide and politicide. Conversely, in all models, I find that wealthier states are less likely to engage in political mass murder (e.g. Freeman, 1991). Only regime type (Rummel, 1995; Harff, 2003) and coups (Wayman & Tago, 2010) appear to have no statistically significant effect on the likelihood of genocide or politicide. Probing the regime type result, I find that this pattern holds if I replace the Polity IV Score with an indicator variable of democracy (Polity 2–6) or Vreeland’s (2008) xpolity score. However, Uzonyi et al.’s (2012) ACC measure indicates that democracy reduces the likelihood of mass killing in some models. Additionally, while democratic transitions are understood to be prone to domestic violence (Snyder, 2000), I find that such transitions are statistically insignificant in my models. For space concerns, each of these models is presented in the Online appendix.

Several robustness questions may still exist for this analysis. I attempt to address those here and present the results in Table III. First, military capacity may alter the regime’s ability to handle domestic and international

Table II.

Testing the direct effect of interstate rivalry on genocide and politicide

	Model 1: Base	Model 2: Controls	Model 3: RELOGIT	Model 4: Jackknife	Model 5: Country FE	Model 6: Conflict	Model 7: Guerrilla
	Coefficient (Robust standard error)	Coefficient (Robust standard error)	Coefficient (Robust standard error)	Coefficient (Robust standard error)	Coefficient (Robust standard error)	Coefficient (Robust standard error)	Coefficient (Robust standard error)
Rivalry_t-1	0.991* (0.225)	0.583* (0.196)	0.576* (0.195)	0.583* (0.243)	0.607* (0.293)	0.479* (0.205)	0.434* (0.195)
Civil war_t-1						1.291* (0.296)	1.018* (0.248)
Interstate war_t-1						1.419* (0.466)	1.255* (0.565)
Ln(GDP_pc)_t-1		–0.433* (0.145)	–0.426* (0.145)	–0.433* (0.175)	–0.549* (0.192)	–0.413* (0.142)	0.433* (0.145)
Democracy_t-1		–0.020 (0.024)	–0.019 (0.024)	–0.020 (0.028)	–0.032 (0.025)	–0.025 (0.025)	–0.018 (0.026)
Exclusionary ideology_t-1		0.660 (0.352)	0.655 (0.351)	0.660 (0.433)	0.666 (0.416)	0.857* (0.317)	0.751* (0.319)
Population size_t-1		0.156 (0.093)	0.154 (0.093)	0.156 (0.119)	0.318* (0.135)	0.061 (0.107)	0.080 (0.098)
Coup_t-1		0.571 (0.317)	0.566 (0.316)	0.571 (0.338)	0.455 (0.330)	0.388 (0.348)	0.332 (0.364)
Years without genocide/politicide	–2.401* (0.306)	–2.062* (0.273)	–2.024* (0.272)	–2.062* (0.316)	–1.677* (0.136)	–1.835* (0.239)	–1.994* (0.265)
Years without genocide/politicide²	0.177* (0.030)	0.150* (0.026)	0.146* (0.026)	0.150* (0.032)	0.120* (0.012)	0.131* (0.023)	0.143* (0.025)
Years without genocide/politicide³	–0.004* (0.001)	–0.003* (0.001)	–0.003* (0.001)	–0.003* (0.001)	–0.002* (0.000)	–0.003* (0.001)	–0.003* (0.001)
Constant	0.272 (0.273)	1.859 (1.502)	1.822 (1.500)	1.859 (1.832)	–0.461 (2.017)	1.911 (1.690)	2.222 (1.614)
N	8,387	7,607	7,607	7,607	7,607	7,135	6,829
Log pseudolikelihood	–414.807	–374.710		–374.710	–356.564	–352.287	–335.227

Errors clustered by country; *p < 0.05.

Table III.

Additional analysis

			Model 10: Bivariate Probit
	Model 8: Power	Model 9: New state	Genocide	Politicide
	Coefficient (Robust standard error)	Coefficient (Robust standard error)	Coefficient (Robust standard error)	Coefficient (Robust standard error)
Rivalry_t-1	–0.645 (0.734)	0.559* (0.196)	0.163 (0.151)	0.330* (0.147)
Power_t-1	–0.493* (0.196)
Rivalry_t-1* Power_t-1	–0.209^† (0.107)
New state_t-1		–1.137* (0.360)
Ln(GDP_pc)_t-1	–0.195 (0.178)	–0.443* (0.144)	–0.165 (0.105)	–0.238* (0.097)
Democracy_t-1	–0.019 (0.024)	–0.016 (0.024)	–0.020* (0.008)	–0.014 (0.015)
Exclusionary ideology_t-1	0.859* (0.336)	0.639 (0.348)	0.223 (0.268)	0.425 (0.229)
Population size_t-1	0.765* (0.234)	0.126 (0.087)	0.144* (0.052)	0.123* (0.056)
Coup_t-1	0.536 (0.303)	0.529 (0.314)	0.868* (0.150)	0.175 (0.175)
Years without genocide/politicide	–1.975* (0.266)	–2.029* (0.257)
Years without genocide/politicide²	0.142* (0.025)	0.146* (0.025)
Years without genocide/politicide³	–0.003* (0.001)	–0.003* (0.001)
Years without genocide			–0.420* (0.085)
Years without genocide²			0.017* (0.005)
Years without genocide³			–0.000* (0.000)
Years without politicide				–0.833* (0.117)
Years without politicide²				0.056* (0.011)
Years without politicide³				–0.001* (0.000)
Constant	–8.989* (4.162)	2.354 (1.463)	–1.149 (0.786)	0.437 (0.990)
ρ			0.317* (0.129)
N	7,135	7,607	7,607
Log pseudolikelihood	–365.869	–370.474	–496.274

Errors clustered by country; ^†p = 0.1; *p < 0.05.

threats simultaneously. When I include the interaction between interstate rivalry and the Correlates of War’s measure of state military strength (Singer et al., 1972) in Model 8, I find that the positive relationship between rivalry and mass killing is produced by militarily weaker states that are engaged in interstate rivalry, and is decreasing as power increases. See Figure 2. This finding fits with the idea that when states are weaker, and less able to address international concerns, they are more likely to seek domestic scapegoats to punish (Midlarsky, 2005). Second, states might be most prone to domestic violence shortly after independence (Wimmer, 2002). However, in Model 9, I find that this period is less likely to see mass killings than others. From the rivalry perspective, this may be because new states are still forming identities and have yet to fully establish strategic rivals.

Figure 2.

Power, rivalry, and genocide/politicide

Table IV.

Probing the rivalry mechanism

	Model 11: Rivalry age	Model 12: MID count	Model 13: Enduring rivals
	Coefficient (Robust standard error)	Coefficient (Robust standard error)	Coefficient (Robust standard error)
Enduring rivalry_t-1			0.125 (0.399)
MID count_t-1	0.101* (0.029)	0.008 (0.007)
MID count² _t-1	–0.003* (0.001)
MID count³ _t-1	0.000* (6.72e-06)
Ln(GDP_pc)_t-1	–0.452* (0.164)	–0.353* (0.175)	–0.644* (0.174)
Democracy_t-1	–0.026 (0.026)	–0.003 (0.032)	0.035 (0.027)
Exclusionary ideology_t-1	0.607 (0.343)	0.514 (0.392)	1.104* (0.288)
Population size_t-1	0.153 (0.114)	–0.017 (0.101)	0.088 (0.096)
Coups_t-1	0.565 (0.330)	0.578 (0.367)	0.474 (0.355)
Years without genocide/politicide	–2.077* (0.289)	–2.312* (0.316)	–1.468* (0.161)
Years without genocide/politicide²	0.151* (0.028)	0.182* (0.028)	0.096* (0.016)
Years without genocide/politicide³	–0.003* (0.001)	–0.004* (0.001)	–0.002* (0.000)
Constant	2.245 (1.750)	3.566* (1.730)	4.536* (1.591)
N	6,046	2,683	2,340
Log pseudolikelihood	–358.195	–227.403	–197.206

Errors clustered by country; *p < 0.05.

Third, Straus (2015) cautions against conflating measures of genocide and politicide. Using a bivariate probit model on Harff’s (2003) distinction between genocide and politicide in Model 10, I test whether rivalry influences the occurrence of genocide and politicide in different ways. I find that rivalry is more likely to increase politicide than genocide. This finding suggests that international threat may pose divergent incentives for states to use mass violence when facing communal versus cross-cutting political threats (see Uzonyi, 2015, 2016 for a similar point) and that salient communal divisions are not necessary for international threat to increase the likelihood of domestic violence as Bulutgil (2015) posits. These additional tests help provide nuance to my analysis, providing a more refined understanding of these dynamics, and reveal that the connection between rivalry and domestic mass killing is robust.

In Models 11–13 of Table IV, I investigate the direct rivalry mechanism further. In Model 11, I examine whether the lifecycle of the rivalry influences the occurrence of mass killing. To conduct this analysis, I follow Rider et al. (2011) and use the count, count squared, and count cubed of each state’s rivalry MIDs to measure the lifecycle of a rivalry. As Model 11 indicates, rivalry increases the likelihood of mass killing in the earlier and later periods of interstate competition. However, when conflict is at its highest between rivals, leaders tend to focus more on perpetrating interstate violence than domestic violence. In Model 12, I restrict the sample to only states engaged in rivalry and explore whether an increase in MIDs for these states is associated with an increased likelihood of genocide and politicide. As Model 12 reveals, the actual number of MIDs for each rival is not statistically associated with the likelihood of killing. Lastly, I explore whether enduring rivals, as coded by Klein et al. (2006), have an increased likelihood of genocide and politicide compared to other types of rivals. As Model 13 indicates, enduring rivals are not any more susceptible to government mass atrocities than are other rivals. Overall, these results inform us that rivals are most at risk for genocide and politicide early and late in their lifecycles, but that no one type of rivalry is most likely to commit such violence.

Analysis of rivalry’s indirect effect

Models 1–13 focused on Hypothesis 1, which considered the direct influence of interstate rivalry on the likelihood of political mass killing. However, in addition to having a direct effect on the likelihood of genocide and politicide, rivalry also has an indirect effect on such killing by creating the conditions that lead to opportunities for government mass murder, such as civil and interstate war, as well (Hypothesis 2). To untangle the direct effect of rivalry on state mass murder from its indirect influence on such atrocities, I conduct a mediation analysis (e.g. MacKinnon & Dwyer, 1993). As my dependent variable and each mediating variable are dichotomous,

Table V.

Mediation analysis of rivalry’s indirect effect on genocide and politicide

	Model 14: Indirect effect
	Coefficient (Robust standard error)
Rivalry’s indirect effect: Interstate war_t-1	0.052* (0.009)
Rivalry’s indirect effect: Civil war_t-1	0.070* (0.008)
Rivalry’s total indirect effect	0.122* (0.012)
Rivalry’s direct effect	0.285* (0.032)
Rivalry’s total effect	0.407* (0.031)

Standard errors obtained through 500 bootstrap iterations; *p < 0.05.

the mediation analysis begins with logistic regressions of rivalry’s effect on civil war and interstate war, as well as the effects of rivalry, civil war, and interstate war on mass killing. Then, the coefficients from these models are standardized using the standard deviation of the underlying latent variable for each binary variable. Following this rescaling of the coefficients, the indirect effects of rivalry are computed through a product of coefficients approach (see Imai, Keele & Tingley, 2010).⁸ As is evident from the results of Model 14, which is presented in Table V, there is strong support for the argument that rivalry has both a direct and an indirect effect on government mass killing. As Model 14 indicates, rivalry increases the likelihood of genocide and politicide directly, as well as by influencing the likelihood of these atrocities through the pathways of civil war and interstate war. These findings reinforce the argument that rivalry increases the likelihood of genocide and politicide outside of its influence on the occurrence of domestic and international conflict.

The total indirect effect of rivalry is fairly small (0.122) and the ratio of the indirect effect to rivalry’s direct effect is 0.429. Furthermore, the proportion of rivalry’s total effect that is mediated is 0.300. These results indicate that rivalry has a greater direct effect on genocide and politicide than an indirect effect. However, it is important to note that rivalry’s total effect (0.407) is more than twice that of interstate war’s total effect (0.153) and is also greater than the total effect of civil war (0.381). Thus, while rivalry has a small indirect effect, it has a substantial total effect on the occurrence of government atrocity. Comparatively, these results are important, since civil war and interstate war are seen as primary explanations for genocide and politicide.

Discussion and conclusion

I argued that because interstate rivalry increases the perception of threat within a country and places the military in a state of heightened aggression, states engaged in rivalry are more likely to perpetrate mass killing than are non-interstate rivals. Rivalry also has an indirect effect on genocide and politicide by increasing the probability of civil and interstate war, which each increase the likelihood that a government will commit domestic mass atrocities. Examination of the raw data on all countries between 1955 and 2011, as well as econometric analysis, revealed strong support for this argument.

My theory and findings provide several implications for our understanding of both interstate rivalry and government-sponsored mass killing, as well as avenues for future research. Here, I will focus on two implications and options for future research. First, it is clear that interstate rivalry influences a state’s domestic political situation. Past scholars have noted that rivalry increases arms spending (e.g. Rider, 2009) and strengthens the political position of hawks within the government (e.g. Snyder, 1989; Vasquez, 1985, 1993). These domestic changes work to reinforce the rivalry. Here, I show that these domestic changes also have an influence on the domestic practices of the state. Rather than focusing on how domestic changes feed back into the cycle of rivalry, I show that these changes have serious consequences for the interactions between the government and its own population. While I focus on one grave outcome of these domestic changes, I believe that the heightened tensions and perception of threat caused by interstate rivalry may influence several other domestic processes. For instance, future scholars may wish to examine whether rivalry influences levels of repression within a state or the reoccurrence of civil war. This implication is important because it highlights a potential feedback loop in which domestic conflict helps to perpetuate rivalry (Daxecker, 2011; Mitchell & Prins, 2004) and rivalry increases domestic conflict.

Second, my work here has shown that a state’s international threat environment influences the onset of genocide and politicide. While past scholars have analyzed the role interstate war has on instigating episodes of genocide and politicide, results on this subject have been mixed, with Midlarsky (2005) and Rost (2013) finding interstate war to be a significant predictor of mass killing and Colaresi & Carey (2008) and Wayman & Tago (2010) providing no support for this claim. Here, I expand our understanding of the threat environment. Rather than focus on interstate war, I consider the role of interstate rivalry. Unlike in interstate war in which the military can focus its aggression on a clear enemy from abroad, rivalry leaves a state’s military in a condition of heightened tension without a clear outlet for its aggression. This condition of heightened alertness and aggression mixed with an environment of uncertainty may result in violence directed at domestic civilian targets and the occurrence of genocide or politicide. While I focus on interstate rivalry as one type of threat environment, there are a number of other international factors that may have exacerbating or ameliorating effects on the likelihood of government mass killing. For example, future scholars may wish to consider the role alliances or alignment with the dominant state(s) plays in a government’s decision to engage in genocide or politicide.

Footnotes

Replication data

The dataset and Stata 14 do-files for the empirical analysis in this article, along with the Online appendix, can be found at http://www.prio.org/jpr/datasets and

Acknowledgements

I would like to thank Michael Colaresi and Toby Rider for helpful comments on earlier versions of this draft. I would also like to thank Bill Thompson for sharing his updated data with me.

Notes

References

Allee

Todd L

Huth

Paul K

(2006) Legitimizing dispute settlement: International legal rulings as domestic political cover. American Political Science Review 100(2): 219–234.

Avdan

Nazli

Uzonyi

Gary J

(2017) V for vendetta: Government mass killing and domestic terrorism. Studies in Conflict & Terrorism 40(11): 934–965.

Banks

Arthur S

(2010) Cross-national time-series data archive (http://www.cntsdata.com/).

Barnes

Catherine

(2006) The functional utility of genocide: Towards a framework for understanding the connection between genocide and regime consolidation, expansion and maintenance. Journal of Genocide Research 7(3): 309–330.

Bennett

D Scott

(1996) Security, bargaining, and the end of interstate rivalry. International Studies Quarterly 40(2): 157–183.

Bennett

D Scott

(1997) Measuring rivalry termination. Journal of Conflict Resolution 41(2): 227–254.

Besancon

Marie L

(2005) Relative resources: Inequality in ethnic wars, revolutions, and genocides. Journal of Peace Research 42(4): 393–415.

Bueno de Mesquite

Bruce

Siverson

Randolph M

(1995) War and the survival of political leaders: A comparative study of regime types and political accountability. American Political Science Review 89(4): 841–855.

Bulutgil

H Zeynep

(2015) Social cleavages, wartime experience, and ethnic cleansing in Europe. Journal of Peace Research 52(5): 577–590.

10.

Carter

David B

Signorino

Curt S

(2010) Back to the future: Modeling time dependence in binary data. Political Analysis 18(3): 271–292.

11.

Chalk

Frank R

Jonassohn

Kurt

(1990) The History and Sociology of Genocide. New Haven, CT: Yale University Press.

12.

Colaresi

Michael

(2004) When doves cry: International rivalry, unreciprocated cooperation, and leadership turnover. American Journal of Political Science 48(3): 555–570.

13.

Colaresi

Michael

(2005) Scare Tactics: The Politics of International Rivalry. Syracuse, NY: Syracuse University Press.

14.

Colaresi

Michael

Carey

Sabine C

(2008) To kill or to protect: Security forces, domestic institutions, and genocide. Journal of Conflict Resolution 52(1): 39–67.

15.

Colaresi

Michael

Thompson

William R

(2002) Hot spots or hot hands? Rivalry, recurrent conflict, and escalating risks. Journal of Politics 64(4): 1175–1198.

16.

Cribb

Robert

(2001) Genocide in Indonesia, 1965–1966. Journal of Genocide Research 3(2): 219–239.

17.

Cunningham

David E

Lemke

Douglas

(2013) Combining civil and interstate wars. International Organization 67(3): 609–627.

18.

Daxecker

Ursula E

(2011) Rivalry, instability, and the probability of international conflict. Conflict Management and Peace Science 28(5): 543–565.

19.

Deutsch

Karl W

Merritt

Richard L

(1965) Effects of events on national and international images. In: Kelman

Herbert C

(ed.) International Behaviour: A Social-Psychological Analysis. New York: Holt, Reinhart & Winston, 132–187.

20.

Downes

Alexander B

(2008) Targeting Civilians in War. Ithaca, NY: Cornell University Press.

21.

Dugan

Laura

Chenoweth

Erica

(2012) Moving beyond deterrence: The effectiveness of raising the expected utility of abstaining from terrorism in Israel. American Sociological Review 77(4): 597–624.

22.

Fein

Helen

(1993) Accounting for genocide after 1945: Theories and some findings. International Journal on Group Rights 1(2): 79–106.

23.

Findley

Michael G

Young

Joseph K

(2012) Terrorism and civil war: A spatial and temporal approach to a conceptual problem. Perspectives on Politics 10(2): 285–305.

24.

Findley

Michael G

Piazza

James A

Young

Joseph K

(2012) Games rivals play: Terrorism in international rivalries. Journal of Politics 74(1): 235–248.

25.

Freeman

Michael

(1991) The theory and prevention of genocide. Holocaust and Genocide Studies 6(2): 185–199.

26.

Gent

Stephen E

Shannon

Megan

(2011) Decision control and the pursuit of binding conflict management: Choosing the ties that bind. Journal of Conflict Resolution 55(5): 710–732.

27.

Gleditsch

Kristian Skrede

(2002) Expanded trade and GDP data. Journal of Conflict Resolution 46(5): 712–724.

28.

Goertz

Gary

Diehl

Paul F

(1993) Enduring rivalries: Theoretical constructs and empirical patterns. International Studies Quarterly 37(2): 147–171.

29.

Goertz

Gary

Diehl

Paul F

(1995) The initiation and termination of enduring rivalries: The impact of political shocks. American Journal of Political Science 39(1): 30–52.

30.

Goertz

Gary

Jones

Bradford

Diehl

Paul F

(2005) Maintenance processes in international rivalries. Journal of Conflict Resolution 49(5): 742–769.

31.

Haggerty

Richard

(1988) El Salvador: A country study. Foreign Military Influence and Assistance: Federal Research Division Library of Congress (https://www.loc.gov/item/89048948/).

32.

Harff

Barbara

(2003) No lessons learned from the Holocaust? Assessing risks of genocide and political mass murder since 1955. American Political Science Review 97(1): 57–73.

33.

Hegre

Håvard

Ellingsen

Tanja

Gates

Scott

Gleditsch

Nils Petter

(2001) Toward a democratic civil peace? Democracy, political change, and civil war, 1816–1992. American Political Science Review 95(1): 33–48.

34.

Horowitz

Irving

(1976) Genocide: State Power and Mass Murder. New Brunswick, NJ: Transaction.

35.

Hutchison

Marc L

Gibler

Douglas M

(2007) Political tolerance and territorial threat: A cross-national study. Journal of Politics 69(1): 128–143.

36.

Imai

Kosuke

Keele

Luke

Tingley

Dustin

(2010) A general approach to causal mediation analysis. Psychological Methods 15(4): 309–334.

37.

Kelman

Herbert C

(1997) Social-psychological dimensions of international conflict. In: William Zartman

(ed.) Peacemaking in International Conflict: Methods and Techniques. Washington, DC: US Institute of Peace, 191–237.

38.

King

Gary

Zeng

Langche

(2001) Explaining rare events in international relations. International Organization 55(3): 693–715.

39.

Klein

James P

Goertz

Gary

Diehl

Paul F

(2006) The new rivalry dataset: Procedures and patterns. Journal of Peace Research 43(3): 331–348.

40.

Krain

Matthew

(1997) State-sponsored mass murder: The onset and severity of genocides and politicides. Journal of Conflict Resolution 41(3): 331–360.

41.

Kuper

Leo

(1981) Genocide: It’s Political Use in the Twentieth Century. New Haven, CT: Yale University Press.

42.

Lektzian

David

Prins

Brandon C

Souva

Mark

(2010) Territory, river, and maritime claims in the Western hemisphere: Regime type, rivalry, and MIDs from 1901 to 2000. International Studies Quarterly 54(4): 1073–1098.

43.

MacKinnon

David P

Dwyer

James H

(1993) Estimating mediated effects in prevention studies. Evaluation Review 17(2): 144–158.

44.

Mansbach

Richard W

Vasquez

John A

(1981) In Search of Theory: A New Paradigm for Global Politics. New York: Columbia University Press.

45.

Marshall

Monty G

Marshall

Donna

(2016) Coup D’États Events, 1946–2015. Center for Systemic Peace. INSCR Data Page (http://www.systemicpeace.org/inscrdata.html).

46.

Marshall

Monty G

Gurr

Ted

Harff

Barbara

(2015) PITF-state failure problem set: Internal wars and failure of governance, 1955–2015. Center for Systemic Peace. INSCR Data Page (http://www.systemicpeace.org/inscrdata.html).

47.

Marshall

Monty G

Gurr

Ted

Jaggers

Keith

(2013) Polity IV Project: Political Regime Characteristics and Transitions, 1800–2013. Center for Systemic Peace (http://www.systemicpeace.org/polity/polity4.htm).

48.

Mason

T David

Krane

Dale A

(1989) The political economy of death squads: Toward a theory of the impact of state-sanctioned terror. International Studies Quarterly 33(2): 175–198.

49.

Melson

Robert

(1992) Revolution and Genocide. Chicago, IL: University of Chicago Press.

50.

Midlarsky

Manus I

(2005) The Killing Trap: Genocide in the Twentieth Century. Cambridge: Cambridge University Press.

51.

Mitchell

Sara McLaughlin

Prins

Brandon C

(2004) Rivalry and diversionary uses of force. Journal of Conflict Resolution 48(6): 937–961.

52.

Morey

Daniel S

(2011) When war brings peace: A dynamic model of the rivalry process. American Journal of Political Science 55(2): 263–275.

53.

Mylonas

Harris

(2012) The Politics of Nation Building. Cambridge: Cambridge University Press.

54.

Powell

Robert

(1993) Guns, butter, and anarchy. American Political Science Review 87(1): 115–132.

55.

Rasler

Karen

Thompson

William R

(2000) Explaining rivalry escalation to war: Contiguity, space and position in the major power subsystem. International Studies Quarterly 44(3): 503–531.

56.

Rider

Toby J

(2009) Understanding arms race onset: Rivalry, threat, and territorial competition. Journal of Politics 71(2): 693–703.

57.

Rider

Toby J

Findley

Michael G

Diehl

Paul F

(2011) Just part of the game? Arms races, rivalry, and war. Journal of Peace Research 48(1): 85–100.

58.

Rosecrance

Richard

(1987) Long cycle theory and international relations. International Organization 41(2): 283–301.

59.

Rost

Nicolas

(2013) Will it happen again? On the possibility of forecasting the risk of genocide. Journal of Genocide Research 15(1): 41–67.

60.

Rummel

Rudolph J

(1995) Democracy, power, genocide, and mass murder. Journal of Conflict Resolution 39(1): 3–26.

61.

Salehyan

Idean

Gleditsch

Kristian Skrede

Cunningham

David E

(2011) Explaining external support for insurgent groups. International Organization 65(4): 709–744.

62.

Sarkees

Meredith Reid

Wayman

Frank W

(2010) Resort to War: 1816–2007. Washington, DC: CQ.

63.

Scheina

Robert

(2003) Latin America’s Wars. Vol. 2: The Age of the Professional Soldier, 1900–2001. Washington, DC: Brassey’s.

64.

Semelin

Jacques

(2013) Purify and Destroy: The Political Uses of Massacre and Genocide. New York: Columbia University Press.

65.

Shaw

Martin

(2007) What is Genocide? Cambridge: Polity.

66.

Singer

John David

(1982) Confrontational behavior and escalation to war 1816–1980. Journal of Peace Research 10(1): 1–18.

67.

Singer

John David

; Stuart Bremer & John Stuckey (1972) Capability distribution, uncertainty, and major power war, 1820–1965. In: Russett

Bruce

(ed.) Peace, War, and Numbers. Beverly Hills, CA: Sage, 19–48.

68.

Snyder

Jack

(1989) International leverage on Soviet domestic change. World Politics 30(4): 1–30.

69.

Snyder

Jack

(2000) From Voting to Violence: Democratization and Nationalist Conflict. New York: WW Norton.

70.

Stanley

William

(1996) The Protection Racket State: Elite Politics, Military Extortion, and Civil War in El Salvador. Philadelphia, PA: Temple University Press.

71.

Staub

Ervin

(1989) The Roots of Evil: The Origins of Genocide and Other Group Violence. Cambridge: Cambridge University Press.

72.

Straus

Scott

(2015) Making and Unmaking Nations: War, Leadership and Genocide in Modern Africa. Ithaca, NY: Cornell University Press.

73.

Thies

Cameron G

(2001) Territorial nationalism in spatial rivalries: An institutionalist account of the Argentine–Chilean rivalry. International Interactions 27(4): 399–431.

74.

Thomas

Jakana

(2014) Rewarding bad behavior: How governments respond to terrorism in civil war. American Journal of Political Science 58(4): 804–818.

75.

Thompson

William R

(2001) Identifying rivals and rivalries in world politics. International Studies Quarterly 45(4): 557–586.

76.

Thompson

William R

Dreyer

David

(2011) Handbook of International Rivalries, 1494–2010. Washington, DC: CQ.

77.

United Nations (1993) Report of the UN Truth Commission on El Salvador. United States Institute for Peace (https://www.usip.org/publications/1992/07/truth-commission-el-salvador).

78.

Uzonyi

Gary J

(2015) Civil war victory and the onset of genocide and politicide. International Interactions 41(2): 365–391.

79.

Uzonyi

Gary J

(2016) Domestic unrest, genocide and politicide. Political Studies 64(2): 315–334.

80.

Uzonyi

Gary J

Rider

Toby J

(2017) Determinants of foreign aid, rivalry, and domestic instability. International Interactions 43(2): 272–299.

81.

Uzonyi

Gary J

Souva

Mark

Golder

Sona N

(2012) Domestic institutions and credible signals. International Studies Quarterly 56(4): 765–776.

82.

Valentino

Benjamin A

(2004) Final Solutions: Mass Killing and Genocide in the 20th Century. Ithaca, NY: Cornell University Press.

83.

Valentino

Benjamin A

Huth

Paul

Balch-Lindsay

Dylan

(2004) Draining the sea: Mass killing and guerrilla warfare. International Organization 58(2): 375–407.

84.

Vasquez

John A

(1985) Domestic contention on critical foreign policy issues: The case of the United States. International Organization 39(4): 643–666.

85.

Vasquez

John A

(1993) The War Puzzle. Cambridge: Cambridge University Press.

86.

Vreeland

James

(2008) The effect of political regime on civil war: Unpacking anocracy. Journal of Conflict Resolution 52(3): 401–425.

87.

Walter

Barbara F

(2006) Building reputation: Why governments fight some separatists but not others. American Journal of Political Science 50(2): 313–330.

88.

Wayman

Frank W

Tago

Atsushi

(2010) Explaining the onset of mass killing, 1949–87. Journal of Peace Research 47(1): 3–13.

89.

Whitten

Guy D

Williams

Laron K

(2011) Buttery guns and welfare hawks: The politics of defense spending in advanced industrial democracies. American Journal of Political Science 55(1): 117–134.

90.

Williams

Philip

Walter

Knut

(1997) Militarization and Demilitarization in El Salvador’s Transition to Democracy. Pittsburgh, PA: University of Pittsburgh Press.

91.

Wimmer

Andreas

(2002) Nationalist Exclusion and Ethnic Conflicts. Cambridge: Cambridge University Press.

92.

World Bank (2015) Data Catalogue (https://data.worldbank.org/data-catalog).

93.

Wright

Thorin M

(2014) Territorial revision and state repression. Journal of Peace Research 51(3): 375–387.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

Interstate rivalry,genocide,and politicide

Abstract

Keywords

Introduction

The influence of interstate rivalry on state behavior

The role of interstate rivalry in genocide and politicide

The direct effect: Fear as motivation

The indirect effect: Conflict as opportunity

Research design

Unit of analysis

Dependent variable and estimator

Independent variables

Rivalry

Civil war

Interstate war

GDPpc

Democracy

Exclusionary ideology

Population size

Coup

Time polynomial

Raw data

Analysis of rivalry’s direct effect

Analysis of rivalry’s indirect effect

Discussion and conclusion

Footnotes

Replication data

Acknowledgements

Notes

References

Supplementary Material

GDP_pc