Peacekeeper Fatalities and Force Commitments to UN Operations

Abstract

To what extent do peacekeeper fatalities affect states’ contributions to UN operations? While the deaths of peacekeepers are thought to be a factor in states’ decisions to reduce the magnitude of their participation in a mission, not all states respond similarly to peacekeeper fatalities. I hypothesize that democracies and wealthy countries are likely to be more sensitive to peacekeeper deaths than their non-democratic and poorer counterparts. Analyses of UN peacekeeping operations between 1990 and 2011 confirm that peacekeeper fatalities generally have a negative effect upon the size of countries’ contributions to peacekeeping operations, and that wealthy countries are likely to make larger decreases to their contributions than poorer countries. There is less evidence, however, that democracies are more sensitive to peacekeeper fatalities than non-democracies.

Keywords

Political violence United Nations UN Fatalities UN Peacekeeping

Introduction

Peacekeeper deaths are an unfortunate reality for states that contribute personnel to United Nations peace operations. Since 1948, over 3300 peacekeepers—including troops, police, military observers, and UN staff—have died in operations as a result of malicious acts, illnesses, accidents, and other incidents (dos Santos Cruz, 2017; Henke, 2017). However, little is known about the consequences of these fatalities and, in particular, how they affect states’ continued participation in the UN missions where their personnel die. In some cases, states—such as Belgium, after the death of 10 of its soldiers in Rwanda in 1994—respond by reducing their commitment to a given operation (e.g. Power, 2001). In other cases, states—such as Pakistan, following the death of 25 of its personnel in Somalia in June of 1993—maintain or increase their level of participation.¹ Understanding how much of an impact peacekeeper fatalities have on peacekeeping commitments—and under what circumstances states are likely to withdraw personnel after suffering fatalities—has potential implications for UN mission planning, operational effectiveness, and the ability to prevent future losses.

This article seeks to make two main contributions to the growing body of literature on peacekeeper fatalities. First, although the deaths of peacekeepers are often assumed to lead states to decrease their personnel contributions to UN operations, there has been no effort to theorize this phenomenon, nor to test it empirically across a large number of cases. Second, there has been minimal—if any—recognition that states’ responses to peacekeeper fatalities often differ. In some cases, states reduce the size of their contributions following peacekeeper deaths while in others they maintain or even escalate their degree of commitment. Yet there is little understanding of why or how frequently this variation occurs. These theoretical and empirical gaps will be subsequently addressed.

Specifically, I argue that peacekeeper fatalities compel personnel-contributing countries to update their calculations on the costs and benefits of participation in a UN operation. Peacekeeper deaths—whether they occur as a result of malicious acts, accidents, or illness—can indicate to a state that the mission is costlier than previously anticipated, and suggest that the political, material, and human costs of participation may no longer outweigh the benefits. Under these circumstances, states are increasingly likely to decide that maintaining the same level of participation is no longer in their interest, leading them to reduce the magnitude of their contribution. However, not all states are likely to reach the same conclusions. Developing states receive greater benefit from peacekeeping than their wealthier counterparts owing to the relative value they assign to UN reimbursements for deploying personnel. As a result, I hypothesize that they will be willing to incur greater costs—including the deaths of peacekeepers—before altering their behavior in regards to contributions. In terms of domestic political costs, meanwhile, I posit that democratic countries will be more sensitive to peacekeeping fatalities than non-democracies owing to the different domestic electoral pressures they face. Wealthier and democratic countries are thus likely to make greater reductions in their contributions to UN missions following peacekeeper deaths.

To test these hypotheses, I analyze monthly data on countries’ peacekeeping deployment and peacekeeper fatalities in 22 UN operations between 1990 and 2011. The findings indicate that states that incur peacekeeping fatalities are likely to decrease the size of their contributions to those operations, with higher numbers of fatalities associated with larger personnel reductions. When wealthier states lose peacekeepers, however, they are much more likely to withdraw personnel than less wealthy states. There is only weaker evidence, meanwhile, to support the hypothesis that democracies are more likely than non-democracies to reduce their contributions after incurring peacekeeping fatalities. I conclude with a discussion of implications for policy and future research.

Peacekeeper fatalities

Research on peacekeeping has grown substantially over the last two decades. Scholars and policymakers have focused particularly on a variety of issues such as explaining peacekeeping effectiveness (e.g. Doyle and Sambanis, 2006; Fortna, 2008; Howard, 2008), the location of peacekeeping operations (e.g. Fortna, 2004; Gilligan and Stedman, 2003; Townsen and Reeder, 2014), and the supply of peacekeepers (e.g. Bove and Elia, 2011; Bellamy and Williams, 2013; Gaibulloev et al., 2009, 2015; Khanna et al., 1999). Yet much of this research has paid little attention to the causes or consequences of peacekeeper fatalities.

Despite this relative lack of attention, peacekeeper fatalities are not a new phenomenon (Figure 1). For instance, during the United Nations Operation in the Congo (ONUC, 1960–1964), 249 peacekeepers died from accidents, illnesses, and malicious acts committed against the UN. In the 1990s, meanwhile, 161 peacekeepers deployed in the United Nations Operation in Somalia (UNOSOM I and UNOSOM II, 1992–1995) incurred fatalities. Even UNFICYP, a traditional peacekeeping operation in Cyprus, has experienced 183 fatalities (primarily owing to accidents and illnesses) across its six decades of existence.²

Figure 1.

Peacekeeper fatalities per year, 1948–2015.

Recently, scholars have begun to focus more attention on the phenomenon of peacekeeper fatalities. Three main areas of research have emerged. The first looks specifically at trends in peacekeeping fatalities across time (Boshoff, 2008; Dorn, 2008; Rogers and Kennedy, 2014; Seet and Burnham, 2000; Van der Lijn and Smit, 2015). These studies are largely descriptive in nature, examining the numbers of peacekeeper deaths, the types of deaths that occur (i.e. illness, accidents, malicious attacks), and the countries that incur them, but with little effort to build theories or test hypotheses about the observed data.

The second area involves centralizing data on peacekeeper fatalities. One of the main barriers to analyzing peacekeeper deaths has been the dearth of easily accessible data that captures information on the number of fatalities incurred during UN operations. While this information has been available on the UN’s website for a number of years, the data is presented across multiple documents, making it difficult for researchers to easily collate and employ this information in quantitative studies.³ Henke’s (2017) effort to gather this information into a single, analyzable dataset spanning from 1948 to 2015 allows researchers to overcome this obstacle, distinguishing different types of peacekeeper fatalities, including accidents, illnesses, and malicious incidents, while also identifying the nationality of peacekeepers that died. The UCDP Peacemakers at Risk (PAR) dataset, meanwhile, includes data that goes beyond what is provided by the UN. Consisting of georeferenced events data from Africa, PAR includes not only data on peacekeeper fatalities, but also information on injuries and kidnapping (Lindberg Bromley, 2018). Such efforts have mitigated key empirical barriers to conducting research on peacekeeping deaths.

Third, researchers have considered the causes of deliberate violence against peacekeepers. Scholars have posited various explanations for why rebel groups in particular might engage in this type of behavior. For instance, evidence suggests that rebels are more likely to attack peacekeepers when the balance of power favors government forces. Targeting peacekeepers can potentially reduce the UN’s influence by derailing a peace process while, perhaps concurrently, preventing internal fragmentation within a rebel group (Fjelde et al., 2016). However, it is also possible that stronger rebels are more likely to attack peacekeepers. Salverda (2013) argues that rebels in a relatively strong position may be more likely to target UN personnel because peacekeepers present an obstacle to achieving military objectives. If rebels can convince the UN to withdraw, their path to victory should be more attainable. In Rwanda, for instance, it was reported that Hutu militias deliberately targeted Belgian peacekeepers to “thus guarantee Belgian withdrawal from Rwanda” (Dallaire, 1994, para. 2). Such efforts to explain the logic of violence against peacekeepers have important implications, both for theory and practice.

Despite these developments in the study of peacekeeper fatalities, there has thus far been little effort to explore the consequences of peacekeeper fatalities. In particular, we still know little about whether or to what extent peacekeeper deaths impact states’ decisions about continuing to participate in peacekeeping operations. In some studies, it is assumed that peacekeeper deaths lead states to reduce the magnitude of their contributions (e.g. Lindberg Bromley, 2018; Salverda, 2013). However, there has been no effort to explicitly theorize why peacekeeper fatalities would cause countries to withdraw personnel, nor has there been any systematic analysis of whether such behavior is actually observed in practice. Moreover, there has been no acknowledgement of or explanation for why some states reduce their contributions after suffering fatalities, while others maintain—or even increase—their level of participation.

There are numerous reasons why these phenomena are worth addressing. In particular, understanding the extent to which states are likely to withdraw peacekeepers after incurring fatalities could have implications for UN mission planning and outcomes. Personnel turnover in UN missions may be associated with lower levels of institutional knowledge and can make it difficult to fulfill operational mandates. For example, with regards to civilian police deployed to the UN’s operation in Bosnia in the 1990s, Call and Barnett (1999: 57) note that rapid turnover of police contingents “generally meant that new CIVPOL personnel [spent] up to half their stay just getting familiar with the country and environment.” During the UNAMIR operation in Rwanda, meanwhile, Force Commander Romeo Dallaire stated that the departure of Belgium (following the death of 10 of its soldiers in April 1994) would leave him with “no real substantive capability to sustain myself” (Frontline, 2004, para. 68) and that Belgian troops were necessary to carry out key aspects of the operation (Pugliese, 2014). Knowing which countries are more or less likely to reduce their commitments after incurring fatalities could thus be useful information for UN staff seeking to design missions and deploy forces in ways that encourage personnel stability. Examining states’ responses to peacekeeper deaths is also a way of evaluating the various strategic logics that have been hypothesized to explain violence against UN personnel (Fjelde et al., 2016; Salverda, 2013). If attacking peacekeepers is logical for belligerents under certain circumstances, but not others, this could help mission planners avoid deploying peacekeepers into situations where there are more likely to be killed.

How peacekeeper fatalities affect personnel commitments

Once states have begun contributing personnel to a UN operation, why might peacekeeper deaths cause them to reduce their level of commitment? I begin with the premise that contributing personnel to UN operations is a logical decision that involves states weighing the costs and benefits of such an action. To date, scholars have identified numerous benefits that states receive from participating in peacekeeping missions. These include security benefits such as whether a conflict is in the realm of a potential contributors’ national interest (e.g. Findlay, 1996; Neack, 1995); political benefits such as enhancing one’s international prestige or increasing the chances of gaining a seat on the Security Council (e.g. Krishnasamy, 2001; Thakur, 1980); economic benefits, in the form of reimbursements paid to contributors on a per-peacekeeper basis (e.g. Axe, 2010; Thakur, 1984); and domestic institutional benefits, including militaries’ desire to provide their troops with on-the-ground experience in a foreign country (Findlay, 1996). Participation could also be a way for states to demonstrate compliance with international norms of cooperation (Brysk, 2009) or to engage in coup-proofing efforts by removing military personnel from the domestic arena (Kathman and Melin, 2016). Others have argued that democracies are more likely to contribute personnel than non-democracies (e.g. Andersson, 2000; Perkins and Neumayer, 2008), viewing peacekeeping as a multilateral channel to export liberal principles and pursue humanitarian initiatives across the globe (Lebovic, 2004). Moreover, when compared with other forms of military intervention, UN peacekeeping presents a more economical way to potentially improve human security (Hultman et al., 2013) and prevent civil wars from spilling over into other states (Beardsley, 2011).⁴

Each state assigns different weights to such benefits, and the aggregate utility that a state receives from participation varies depending on how much value is placed upon these particular incentives. However, a state’s calculation is not simply based upon the benefits it receives from participation. These benefits are weighed against the cost of contributing personnel to a given UN operation. When a state decides to make—and maintain—a contribution to a mission, it determines that the benefits outweigh the costs. When a state decides to decrease its contribution, one plausible explanation is that its previous calculation has shifted as the costs of participation have increased.

Peacekeeper deaths—whether they are a result of malicious acts, illness, or accident— are arguably the most significant cost that a state will incur after making the initial decision to deploy personnel to a UN operation. These costs can be conceived of in a number of different ways. There are, of course, human costs to consider. The loss of peacekeepers means that families lose children, spouses, friends, and neighbors. These losses are often unanticipated by people living in countries whose governments participate in peacekeeping operations. When a Chinese peacekeeper was killed in South Sudan in 2016, for instance, many in the country were shocked—the risks of peacekeeping were not widely comprehended by the majority of the Chinese population (Page and Stevis, 2016).

There are also material costs to consider. States invest significant resources in training, equipping, and deploying their personnel. For example, in 2005 the baseline cost for a US military personnel-year was estimated at approximately $112,000 (Government Accountability Office, 2005). If a five-year veteran of the US military were killed during a UN operation, the US could thus incur a loss of nearly half-a-million dollars that it had invested in that soldier, even after accounting for the monthly reimbursements that the UN provides personnel contributors. Material costs may also include loss of equipment, particularly if an attack on peacekeepers or an accident results in the destruction of weaponry or a vehicle.

Peacekeeping deaths that occur as a result of malicious acts also potentially create reputational costs for a contributing state. If a peacekeeper is killed in the context of a UN operation, this might imply that the contributing country is not as powerful as it wants others to believe. This could have multiple consequences. First, it could embolden belligerents within the same conflict to launch subsequent attacks against peacekeepers from that particular country. If killing peacekeepers is a strategic maneuver (Fjelde et al., 2016; Salverda, 2013), belligerents may increase violence against personnel from countries that have fallen victim to such attacks in the past. Second, there could be ramifications outside of the mission in question. States may be concerned that peacekeeper deaths are indicative of a sign of weakness, especially if their personnel are killed by less-powerful non-state actors. For example, in response to the Chinese peacekeeper death in South Sudan mentioned previously, a retired Colonel stated, “We can’t passively take a beating” and called for China to respond militarily (Page and Stevis, 2016, para. 14).

Finally, the death of peacekeepers could result in electoral costs. Peacekeeping is often perceived by the public as a low-risk endeavor. Once peacekeepers die, however, leaders must justify to their people why their fellow countrymen and women are returning home in caskets. This can be particularly difficult in contemporary times, given that most peacekeeping operations in the post-Cold War era have occurred in the context of civil wars that pose little or no immediate national security threat to most states (Gelb, 1994; Smith, 2005). As a result, policymakers may find it difficult to justify the death of personnel in countries where interests at stake are minimal or non-existent, which in turn threatens their political positions. For instance, in 2016, the Dutch defense minister and the top Dutch military commander both resigned following the accidental deaths of two Dutch soldiers in Mali during training exercises. “I am politically responsible and am taking that responsibility,” noted the defense minister as she stepped down (BBC News, 2017, para. 5). If the public perceives that the government has engaged in ill-advised foreign endeavors, they could be more likely to oppose the state’s policies (e.g. Gartner and Segura, 1998; Mueller, 1973), which could lead them to seek the removal of leaders as punishment for their actions.

Personnel have died during UN operations for more than half a century; as such, the death of peacekeepers is not wholly unexpected, and may not immediately lead a country to draw down its forces.⁵ However, it is plausible that such occurrences compel states to recalculate the costs associated with participation. Whereas previously a state may have anticipated the likelihood of peacekeeper deaths to be low, the actual loss of a peacekeeper forces a state to reckon with the reality of the situation into which it has deployed personnel. When faced with elevated human, material, reputational, and electoral costs of peacekeeping, states may become likely to reduce their commitment to a given operation. This suggests the following hypothesis:

Hypothesis 1: The greater the number of peacekeeper fatalities that a state incurs in a UN operation, the greater the absolute magnitude of its subsequent personnel reduction to that operation.

If the deaths of personnel force states to grapple with the costs of participating in peacekeeping, why do some states decide that the benefits remain sufficiently high to maintain a constant level of support, while others arrive at the opposite conclusion? I posit that this can be explained by variation in states’ valuation of the benefits of peacekeeping and the costs of peacekeeper deaths. In particular, a country’s wealth and regime type are likely to affect this calculation.

While no country wants to incur fatalities during UN missions, the extent to which the costs of peacekeeper deaths dissuade states from contributing to these operations is dependent, in part, upon the benefits they accrue from participation. If a state perceives relatively lower utility from contributing personnel to the UN, it is more likely that the costs associated with peacekeeping fatalities will outweigh the benefits, thus discouraging participation. For those countries that perceive relatively greater benefit from contributing peacekeepers, on the other hand, peacekeeper deaths are likely to have less impact on their calculations regarding the utility of supplying personnel. If the benefits of participation are high enough, states may be able to justify the elevated costs associated with a particular operation.

Developing and developed states are particularly likely to perceive the benefits of peacekeeping differently, owing to the financial incentives associated with participation. Many scholars have hypothesized that poor countries contribute to UN operations because of the direct financial benefits they receive for doing so (e.g. Axe, 2010; Bove and Elia, 2011; Burrows, 2014; Morrison, 1996; Thakur, 1984). While contributions to UN operations are voluntary, states are compensated for their participation. For each peacekeeper that a country contributes, it receives a reimbursement of $1410 per month (paid directly to the government).⁶ Additionally, states receive compensation for equipment and clothing, personal weaponry, additional pay for specialists, and disability costs (Bove and Elia, 2011).

For developed countries, the UN rate of reimbursement does not come close to covering the costs they expend to support their personnel. By one estimate, the annual costs per military personnel for states such as Canada, France, Italy, and the United States were well over $100,000 per year (Gaibulloev et al., 2015), far more than the reimbursement that the UN currently provides. Personnel contributions are thus not a financially beneficial proposition for wealthy states and, as a result, the prospect of UN reimbursements is unlikely to factor heavily into their decisions about contributing to a UN operation. This stands in contrast to developing countries that may view peacekeeping as an important source of income. Major contributors such as Bangladesh, Ghana, India, Nepal, Pakistan, and Nigeria all spend far less annually on a per-peacekeeper basis compared with the reimbursements provided by the UN (Gaibulloev et al., 2015). And, because the UN directly reimburses governments, states can use these flows of currency as they see fit. These reimbursements can add up quickly: between 2001 and 2010, the UN paid $1.28 billion to the Bangladeshi government in exchange for its peacekeeping provision, allowing it “to purchase and maintain military equipment that it would not be able to obtain under normal circumstances” (Uz Zaman and Biswas, 2013: 3). Such financial incentives may compel poorer states to maintain their involvement in UN operations even as conditions worsen. For example, Tanzania—a country that has stated its resolve to remain a contributor to peacekeeping in the DRC, despite attacks on its personnel—continues to receive “an opportunity to benefit economically through UN compensation payments. This is important at the level of soldiers’ welfare, which is improved and their families enjoy the benefits of their participation in UN peace operations” (Minde, 2017, para. 6).

This divergence has implications for states’ decisions about whether or not to continue contributing to UN operations after incurring peacekeeper fatalities. When peacekeeping is perceived as a low-cost endeavor, developed and developing countries both have incentives to continue contributing personnel. As peacekeeping becomes increasingly costly, states must consider if the benefits continue to outweigh these costs. While wealthy countries may find sufficient benefit in contributing to low-risk missions, these benefits are likely to be overshadowed as the costs of participation increase. For poorer countries, on the other hand, peacekeeping provides benefits that otherwise can be difficult to replace. Drawing down personnel from an operation could have consequences for developing countries’ ability to provide national security, support non-military budget items, bolster soldiers’ income, or expand their economies. Because developing states perceive greater financial benefits of peacekeeping than developed states, they are less sensitive to rising costs associated with fatalities. Satish Mishra, a researcher in Mumbai, stresses this point, noting, “There is a lesser emphasis in developing countries on the human cost of sending peacekeeping forces.”⁷ The cost–benefit calculation in regards to personnel contributions following peacekeeper fatalities is thus different for poorer and wealthier countries, suggesting the following:

Hypothesis 2: As a state experiences increasing numbers of peacekeeper fatalities in a UN operation, the absolute magnitude of its subsequent personnel reductions to that operation will be greater for wealthier states than for poorer states.

Democracies and non-democracies are also likely to react differently to peacekeeper fatalities. In democracies, citizens hold leaders accountable via public opinion and elections. If leaders make decisions with which the public disagrees, they risk being punished at the ballot box (e.g. Bueno de Mesquita et al., 2003; Manin et al., 1999). This has ramifications for how democratic leaders make choices, in terms of both domestic and foreign policy. While early research questioned the link between the public’s opinion on foreign affairs and their electoral preferences (e.g. Almond, 1950; Lippman, 1955), Gelpi et al. (2009: 16) note, “Public opinion may be ‘latent’ on many issues, but when activated by news events, and especially by prominent debates within the elite over foreign policy options, it becomes a factor that policymakers must address.”⁸ Assuming that democratic leaders want to keep their jobs, we should expect them to act prudently in foreign affairs, seeking to avoid policies that contradict public sentiment and jeopardize their future employment (Bennett and Stam, 1998; Reiter and Stam, 2002).

Autocratic leaders, on the other hand, are not as reliant upon the broad support of the public to maintain power. While autocrats must satisfy the domestic elites that support their regime (Weeks 2014), the source of authority in an autocratic regime does not come from the acquiescence of the public through the same electoral processes found in democracies. Autocratic leaders thus have more latitude than democratic leaders to institute policies that their people do not broadly support, in terms of both domestic and foreign affairs. As Kant (1795: 100) noted in Perpetual Peace, autocrats can “decide on war, without any significant reason, as a kind of amusement, and unconcernedly leave it to the diplomatic corps … to justify the war for the sake of propriety”.

A number of existing studies have focused specifically on how electoral accountability shapes democratic leaders’ policies regarding foreign military commitments (Gelpi et al., 2009; Karol and Miguel, 2007; Koch, 2011). Sending personnel abroad is one of the most important policy decisions a leader can make. If the result is sub-optimal (e.g. losing a war) or costlier than anticipated (particularly in terms of human lives), democratic citizens can express their displeasure at the polls. These concerns help explain findings showing that democracies tend to avoid wars that are likely to be bloody or losing efforts (Reiter and Stam, 2002; Siverson, 1995), and that public opposition to war is associated with mounting casualties (Gartner and Segura, 1998; Luttwak, 1996; Mueller, 1973).

In certain cases, however, democratic publics may have more tolerance for costly military actions. If the population views a military endeavor as critical to a state’s national interests, the public might be more willing to accept casualties. In contrast, “the further a particular war or military operation is removed from core national interests, the more the populace will be averse to casualties, and the more decision-makers will seek to avoid them” (Smith, 2005: 492). This suggests that democratic policymakers are not always reluctant to become involved in military endeavors. Rather, their hesitation increases in situations where the likelihood of fatalities is high and there is no clear justification for the losses. When making personnel commitments, leaders are thus cognizant not only of the potential human costs associated with their decision, but also whether the public will view those deaths as a worthwhile sacrifice. This calculus should be relevant for democratic leaders even when making decisions about deploying personnel to less-traditional military operations, such as humanitarian interventions or peacekeeping missions.

Given that most countries lack immediate national interests at stake in the countries where the UN operates, democracies should be more likely to withdraw personnel from operations after peacekeepers die. If voters perceive that personnel from their country have died in a situation where there was no clear national security interest, they may conclude that deployment was an ill-advised decision, and they could choose to punish leaders at the ballot box. Fearing this outcome, democratic leaders are likely to be encouraged to withdraw peacekeepers following fatalities in an effort to preempt voter backlash. Similar reductions should be less frequent among autocracies, where leaders’ job security is not as directly associated with instituting policies that align with the public will.

Hypothesis 3: As a state experiences increasing numbers of peacekeeper fatalities in a UN operation, the absolute magnitude of its subsequent personnel reductions to that operation will be greater for democracies than for non-democracies.

Data

In order to test the three hypotheses, I compiled and analyzed a dataset containing each country’s monthly contributions to 22 UN peacekeeping operations in Africa between 1990 and 2011.⁹ The dependent variable (Personnel Change) is a count variable, indicating the magnitude of a state’s change in the number of personnel that it deployed to a particular operation from the previous month. Information on personnel contributions—including troops, police and observers—comes from Kathman’s (2013) dataset on UN peacekeeping commitments.¹⁰ The main analysis examines these three types of personnel in the aggregate; subsequent robustness tests disaggregate personnel type into troops, police, and observers.¹¹

The key independent variable, Fatalities, is based on UN data collected by Henke (2017). The variable indicates, for each month, the total number of peacekeeper deaths that occurred in the previous three months; robustness tests extend this time period to include peacekeeper deaths that occurred in the previous six months. The main analysis aggregates all types of peacekeeper fatalities (accidents, illness, malicious acts, and “other”), with additional analyses considering these categories in the disaggregate.¹² Lagging this variable by more than one month is important as countries may take time to deliberate over adjusting their contribution levels in the wake of peacekeeper fatalities, and to recognize that such a policy shift may be a result of cumulative fatalities over an extended period. Too long a time period, however, is likely to add significant noise to the statistical model.

In order to gauge the effect of peacekeeper fatalities at different levels of wealth and democracy, I created two interaction terms. The first, Fatalities*GDP is an interaction between the Fatalities term and the state’s log-transformed annual real GDP (Gleditsch, 2002). As Ward and Dorussen (2016: 399) note, this measure of a country’s wealth assumes a positive correlation between income level and the price of a soldier.¹³ The second term, Fatalities*Democracy is an interaction between the Fatalities term and a binary variable (Democracy) denoting if a country scores 6 or higher on the Polity IV scale (Marshall et al., 2013).¹⁴ Following the Polity guidelines, countries with a score of 6 or higher are categorized as democracies; those with a score of 5 or below are labeled as non-democracies.¹⁵ While Polity is not the only available measure of regime type, it is suitable in this instance given that its conception of democracy is based largely on the extent to which elites are constrained by their political societies (Gunitsky, 2015). This aligns with the proposed causal process through which democracy is hypothesized to shape states’ responses to peacekeeper fatalities.

The statistical models also control for numerous factors that could lead a country to alter its contributions to a UN operation in a particular year.¹⁶ First, a key challenge in measuring a state’s decision to withdraw personnel is the UN Security Council’s propensity to adjust the size of the force it deploys to each mission. In order to allow for the possibility that personnel changes are a result of mandated reductions, I include a variable indicating the total change in the authorized number of personnel (Authorized Change) from the previous month (Passmore et al., 2017). The monthly data on mandated personnel is only available for UN operations in Africa, thus restricting the analysis to those operations. A full list of operations included in the analysis appears in the Online Appendix.

Another possible factor influencing a country’s decision about personnel contributions is its geographic proximity to the country hosting the peacekeeping operation. Countries might have greater motivation to maintain contribution levels in missions operating in their own neighborhoods. In these situations, states may have specific interests in ensuring that regional stability is maintained, that conflict does not spill across borders (Passmore et al., 2017), and that refugee flows do not become overwhelming (Uzonyi, 2015). Proximity is measured in kilometers (log) between the capital cities of personnel contributors and states hosting peacekeeping operations (Gleditsch and Ward, 2001).

A previous colonial relationship between a personnel contributor and host state could provide additional motivation for staying in a UN operation (Beardsley and Schmidt, 2012; Uzonyi, 2015). It may be the case that countries perceive a unique responsibility for overseeing their former colonies’ stability, or they may have greater political and economic interests left over from the colonial era that they seek to protect (Perkins and Neumeyer, 2008). This relationship could make a former colonizer less likely to reduce its contributions at any given time. I include a binary variable (Colonial) indicating whether or not a contributing country has colonial ties to a country hosting a UN operation (Hensel, 2014).

Permanent membership on the UN Security Council is another factor that has been hypothesized to impact peacekeeping contributions (Diehl, 2008; Uzyoni, 2015), although there is disagreement about the direction of this effect. Some have suggested that permanent members may be particularly committed to UN operations as a way of legitimizing their status on the Council (Bove and Elia, 2011). Other research, however, indicates that permanent members are less likely to participate in UN operations relative to other states (Stojek and Tir, 2015), and thus may be less likely to sustain their commitments. To control for either outcome, the binary variable P5 is included as an indicator of a country’s status on the Security Council.

Other variables relate to states’ ongoing capability to contribute to UN operations. First, if a country is involved in an armed conflict of its own, it may need to transfer personnel from the UN operation to its own military effort. A binary variable (Other Conflict) indicates whether or not a country is engaged in a conflict in the same year as a peacekeeping operation (Marshall, 2016).¹⁷ Second, some countries have a greater supply of personnel which they could draw upon to provide peacekeepers. States with greater numbers of troops, in particular, may be more capable of maintaining contributions to UN operations given their larger available pool of personnel.¹⁸ Annual measures of military personnel (log) come from Singer et al. (1972).

Countries may also decide to redeploy personnel from a UN operation to a non-UN operation. In order to account for this occurrence, I include a variable indicating the number of personnel (log) a country committed to non-UN operations in the same year (Gaibulloev et al., 2015). There are also situations where the UN and other organizations have operated peace operations in the same country concurrently, such as KFOR (NATO) and UNMIK (UN) in Kosovo. In this situation, a country may transfer personnel from the UN operation to the non-UN operation. A variable indicates the number of personnel (log) a country committed annually to a non-UN operation in the same country (Gaibulloev et al., 2015).

States may be less likely to reduce their commitment if they have participated in an operation for a longer period. These contributors may feel more invested in the outcome of an operation, or experience a sense of sunk costs.¹⁹ Countries that are newer to an operation may feel less invested, and thus may have fewer qualms about withdrawing personnel. A count variable (Contribute Duration) indicates the number of months that a country has participated in the operation.²⁰

In order to gauge how meaningful a country’s personnel reduction is, I include a measure of the total number of peacekeepers (including troops, police, and observers) that a country contributed to the operation in the previous month (Personnel_t₋₁). I also consider the possibility that a decline in contributions to a particular operation reflects a country’s general disengagement from UN peacekeeping operations at a given time. A count of a country’s change in total UN contributions from the previous month (Total Personnel Change) controls for this scenario.

Finally, peacekeeper fatalities may shape countries’ decisions about contributions even if personnel from other countries are the ones dying. Such occurrences may indicate that a mission has entered a phase of heightened risk, leading other personnel-contributing countries to reconsider their participation. To account for this, the variable Mission Fatalities 3 Months indicates the number of total peacekeeper fatalities that occurred in a particular operation over the previous three months.

Data analysis

The dataset contains 40,855 observations (Online Appendix, Table S1). Of these, there are 1616 observations (4.0%) where a contributing state experienced peacekeeping fatalities—of any type and involving any type of personnel—in the previous three months, with a maximum of 28 deaths. In 678 (41.9%) of the instances where a fatality occurred, the contributing state decreased its contribution by an average of 47.7 personnel. In 572 instances (35.4%), meanwhile, states increased their contribution by an average of 42.1 personnel. In the 366 remaining instances (22.6%), states did not alter the magnitude of their contributions after experiencing fatalities. In comparison, when countries experienced no fatalities in the previous three months, they decreased (17.7%) and increased (16.1%) their contributions at nearly equal rates, while making no changes to their personnel contributions 66.1% of the time (Table 1).

Table 1.

Contribution changes and fatalities 1990–2011.

	Decrease contribution	Increase contribution	No change to contribution	N
Fatalities > 0 in previous three months	678 (41.9%)	572 (35.4%)	366 (22.6%)	1,616
No fatalities in previous three months	6,953 (17.7%)	6,329 (16.1%)	25,957 (66.1%)	39,239

To test the hypotheses, I run three regressions (Table 2). Model 1 analyzes the relationship between peacekeeper fatalities and personnel changes generally (H1); Model 2 examines how a country’s wealth mediates its response to peacekeeper fatalities (H2); and Model 3 analyzes the responses of democracies and non-democracies following the deaths of peacekeepers (H3). Because the dependent variable is continuous, I analyze the data using an ordinary least squares regression model with mission fixed effects. The use of a fixed-effects approach accounts for mission-specific factors that do not vary across temporal periods.²¹ I also include month and year dummies to account for the possibility that “UN member-states’ willingness to support peacekeeping may vary within years for different reasons” (Passmore et al. 2017: 371), including weather and the timing of key UN meetings and reports.

Table 2.

Peacekeeper fatalities and personnel changes, 1990–2011.

	(1)	(2)	(3)
	Personnel change	Personnel change	Personnel change
Fatalities	−5.397***	35.366***	−4.028***
	(0.676)	(3.704)	(1.131)
GDP	−0.564	−0.410	−0.571
	(0.366)	(0.366)	(0.366)
Democracy	1.081	1.186	1.233
	(0.783)	(0.782)	(0.789)
GDP × Fatalities		−3.264***
		(0.292)
Democracy × Fatalities			−1.998
			(1.324)
Distance	−1.224**	−0.959*	−1.224**
	(0.522)	(0.522)	(0.522)
Colonial	1.339	1.318	1.311
	(2.529)	(2.525)	(2.529)
Non-UN (in-country)	0.889	0.861	0.887
	(0.632)	(0.631)	(0.632)
Non-UN (total)	−0.235**	−0.232**	−0.240**
	(0.117)	(0.116)	(0.117)
Military Personnel	1.632***	1.601***	1.647***
	(0.455)	(0.454)	(0.455)
Authorized Personnel Change	0.001*	0.001*	0.001*
	(0.000)	(0.000)	(0.000)
Personnel_{t − 1}	−0.003***	−0.003***	−0.003***
	(0.001)	(0.001)	(0.001)
Total Personnel Change	0.100***	0.100***	0.100***
	(0.002)	(0.002)	(0.002)
P5	−4.546***	−4.351***	−4.487***
	(1.571)	(1.569)	(1.572)
Other Conflict	2.280**	2.378***	2.246**
	(0.891)	(0.889)	(0.891)
Contribute Duration	0.005	0.004	0.005
	(0.015)	(0.015)	(0.015)
Mission Fatalities 3 Months	−0.008	−0.005	−0.009
	(0.100)	(0.099)	(0.100)
Constant	21.025	16.874	20.967
	(21.004)	(20.975)	(21.004)

Observations	40,855	40,855	40,855
R ²	0.106	0.109	0.106
Number of Missions	22	22	22

Standard errors in parentheses.

***

p < 0.01; **p < 0.05; *p < 0.1.

Examining the results of Model 1, the coefficient for Fatalities is negative and statistically significant at the 0.001 level. This indicates that, as hypothesized, as a contributing state incurs greater numbers of peacekeeping fatalities, the size of its subsequent contributions decreases. This relationship holds, even after other variables that could plausibly cause a state to reduce its commitment—such as geographical proximity or participation in non-UN operations—are held constant. The results are substantively moderate (Figure 2).²² A country experiencing five peacekeeper deaths in the previous three months is predicted to decrease the size of its contribution by about 20 personnel—approximately half the size of a platoon.

Figure 2.

Predicted change in personnel contributions.

The results of Model 2, meanwhile, support the hypothesized interactive relationship between wealth, peacekeeper fatalities, and changes in commitment. The coefficient for the interaction term, Fatalties × GDP is negative and statsitically significant. This indicates that as a country incurs more fatalities in a given operation, its level of wealth is negatively correlated with the magnitude of its change in peacekeeping contributions to that mission. In simpler terms, if a wealthy country and a poorer country both experience the same number of peacekeeping fatalities, the wealthier country is likely to reduce its contributions at a greater rate than the poorer country.

To gauge how substantive this difference is, I calculated the predicted marginal effects of peacekeeper fatalities for countries at one standard deviation above and below the mean GDP (Fig. 3).²³ The differences are large. After incurring one fatality, countries with a GDP that is one standard deviation below the mean are predicted to increase their contribution size by about seven personnel, while countries with a GDP one standard deviation above the mean are predicted to decrease their commitment by nearly eight personnel. These differences become more stark as fatalities increase. After incurring four fatalities, a poorer country is predicted to increase its contribution by 20 personnel, compared with a decrease of around 34 personnel from a wealthier country (roughly the size of a small platoon).

Figure 3.

Predicted change in personnel contributions by GDP.

The finding that poorer countries increase their contributions following fatalities (rather than decreasing at a lower rate than wealthier countries, or remaining stable) merits additional discussion. Perhaps the most likely explanation is that poorer states still find the benefits of participation in a particular operation to outweigh the costs, but also aim to minimize future fatalities. To address this, states may seek to bolster the safety of their personnel by deploying additional forces. This could be in response not only to a malicious attacks (e.g. to increase defensive posture), but also in response to accidents or illnesses (e.g. to increase the presence of logistics and medical personnel to treat maladies with greater efficacy). For poorer states, the benefits are twofold: increased safety for their personnel and increased reimbursement as a result of larger personnel contributions.

In the third model, I test the hypothesis that, after incurring fatalities, democracies will make greater reductions than non-democracies in regards to their personnel commitments to a given operation. In this case, the interaction term (Fatalities × Democracy) is negative but not statistically significant. While the lack of empirical support for this hypothesis is unexpected, there is some precedent for questioning the “casualties hypothesis” in regards to democracies and humanitarian efforts, including peace operations. For instance, some have indicated that the United States’s decision to withdraw personnel from Somalia in the 1990s after incurring high-profile casualties was clear evidence of a democratic country being unable to tolerate losses when operating in a peacekeeping or humanitarian capacity (e.g. Luttwak, 1995; Mueller, 1996).²⁴ Burk (1999), however, finds little evidence to support this notion. Rather, public opinion for US involvement in Somalia had eroded prior to the loss of American lives in Mogadishu in 1993, making it unlikely that a decline in public support was what caused leaders to enact a change in policy. Moreover, some polls showed that a majority of Americans—up to 60%—actually wanted to deploy more soldiers to Somalia after the killings occurred in order to enact vengeance on the perpetrators (Smith, 2005). The relationship between regime type, casualties, and peacekeeping efforts is likely to be less clear-cut than hypothesized; I examine this relationship in greater detail in subsequent tests.

Robustness tests

Temporal unit of analysis

To test the robustness of the results, I first consider that the three-month period of fatalities may fail to capture domestic policymaking processes that take longer to unfold, or the cumulative effect of multiple fatalities incurred by a contributor over a longer period of time. I thus increase the duration of the period to six months for the number of fatalities that a particular country incurs, as well as the variable indicating the number of fatalities that occur mission-wide during this extended period (Table 3). In regards to Hypothesis 1, the results still show a negative correlation between fatalities and personnel reductions, with even greater substantive effect. After experiencing five fatalities in the previous six months, for instance, a country is predicted to reduce its contribution to an operation by almost 44 peacekeepers (as opposed to 20 when examining a three-month period). There is also evidence to support Hypothesis 2, with wealthier countries reducing their contributions in the six months following fatalities, and poorer countries increasing their contributions during that same time span. In terms of regime type, there is some support for Hypothesis 3: while both democracies and non-democracies decrease their contributions after incurring fatalities in the previous six months, democracies do so at a greater rate. The substantive differences, however, are relatively small. After four fatalities, for instance, non-democracies are predicted to reduce their contributions by 26 peacekeepers compared with 39 for democracies. This does suggest, though, that the statistically insignificant findings about regime type in the main analysis may be impacted in part by the selected temporal unit of analysis.²⁵

Table 3.

Peacekeeper fatalities and personnel changes, 1990–2011 – 6 month fatalities.

	(4)	(5)	(6)
	Personnel change	Personnel change	Personnel change
Fatalities	−9.270***	40.616***	−6.736***
	(0.468)	(2.524)	(0.742)
GDP	−0.643*	−0.305	−0.675*
	(0.365)	(0.363)	(0.365)
Democracy	1.187	1.457*	1.747**
	(0.779)	(0.776)	(0.790)
GDP × Fatalities		−4.058***
		(0.202)
Democracy × Fatalities			−3.800***
			(0.863)
Distance	−1.394***	−0.822	−1.395***
	(0.519)	(0.518)	(0.519)
Colonial	1.062	1.100	0.964
	(2.517)	(2.504)	(2.516)
Non-UN (in-country)	0.913	0.861	0.914
	(0.629)	(0.626)	(0.629)
Non-UN (total)	−0.221*	−0.207*	−0.236**
	(0.116)	(0.115)	(0.116)
Military Personnel	1.580***	1.533***	1.638***
	(0.453)	(0.451)	(0.453)
Authorized Personnel Change	0.001*	0.001*	0.001*
	(0.000)	(0.000)	(0.000)
Personnel_{t − 1}	0.001*	0.001*	0.001
	(0.001)	(0.001)	(0.001)
Total Personnel Change	0.100***	0.099***	0.100***
	(0.001)	(0.001)	(0.001)
P5	−3.827**	−3.585**	−3.636**
	(1.564)	(1.556)	(1.564)
Other Conflict	2.285***	2.633***	2.183**
	(0.886)	(0.882)	(0.887)
Contribute Duration	0.003	−0.000	0.003
	(0.015)	(0.015)	(0.015)
Mission Fatalities 6 Months	−0.302***	−0.283***	−0.299***
	(0.064)	(0.064)	(0.064)
Constant	23.390	14.346	23.227
	(20.905)	(20.808)	(20.901)

Observations	40,855	40,855	40,855
R ²	0.115	0.123	0.115
Number of Missions	22	22	22

Standard errors in parentheses.

***

p < 0.01; **p < 0.05; *p < 0.1.

Disaggregating fatality types

Another possibility is that certain types of fatalities—malicious acts, illnesses, accidents, or “other”—influence the results more than others. If so, the key independent variable, Fatalities, may be masking or over-emphasizing the effects of a particular source of peacekeeper deaths. It is particularly plausible that malicious acts may provoke countries to decrease personnel more than other types of fatalities.²⁶ I thus disaggregated the different types of fatalities and re-ran the regressions. In regards to Hypothesis 1 (Online Appendix, Table S5), all four types of fatalities have coefficients that are negative and statistically significant, indicating that each is correlated with a decline in the number of personnel a state contributes to a particular operation. In regards to Model 2 (Online Appendix, Table S6), when interacting the four types of fatalities with contributor states’ respective GDP, all four coefficients on the interaction terms are negative. Two—malicious acts and accidents—are statistically significant at p < 0.001; illnesses and “other” are statistically insignificant. I also re-tested Hypothesis 3 to consider if democracies and non-democracies respond differently to various types of peacekeeper fatalities. The results (Online Appendix, Table S7) show that democracies are more likely than non-democracies to reduce their contributions after experiencing malicious acts, with statistically significant marginal effects beginning at a threshold of two fatalities. When examining the interaction between democracy and other types of deaths, meanwhile, none of the marginal effects achieve statistical significance.

There are a number of takeaways from these disaggregated findings. First, the effect of wealth on states’ responses to peacekeeper fatalities is generally consistent across different types of fatalities, although this relationship is more robust in regards to malicious acts and accidents than the other types. Second, democracies and non-democracies react differently to peacekeeper fatalities following malicious fatalities, but not after incurring other types of deaths. This suggests that the statistically insignificant finding in the main analysis may be masking a subtler effect of regime type on peacekeeping reductions that is revealed through disaggregated analysis.

Disaggregating personnel type

It is also possible that deaths of different types of peacekeeping personnel—troops, police, and observers—provoke distinct responses that shape the observed results. To test this, I run a disaggregated analysis of contribution changes based on the type of personnel that suffer fatalities. The analyses show that contributions of troops, police, and observers are all negatively correlated with deaths of these same types of personnel, with statistically significantly coefficients for troops and police (Table S8, Online Appendix). The predicted marginal effects of fatalities on police reductions, however, are much smaller than the predicted effect on troop reductions. The effect of wealth on personnel changes following fatalities, meanwhile, is only statistically significant in regards to troops, with substantive results similar to the main analysis (Table S9, Online Appendix). This suggests that troops are the primary driver of the relationships between peacekeeper deaths and personnel reduction that were observed in the main analysis.

Conflict dynamics

Finally, a greater level of insecurity in the host country could convince contributors that their peacekeepers are at greater risk, even if they have not suffered fatalities. Under such conditions, UN personnel may be increasingly likely to be drawn into conflict zones, where they are more likely to become targets of belligerents or succumb to accidents, such as friendly fire. To this end, I add a variable to Models 1 and 2, Conflict Fatalities, that indicates the number of conflict-related deaths that occurred in a host-country in the previous three months.²⁷ Including this measure results in only minimal changes in the coefficient size of the Fatalities variable in Model 1 and the interaction term in Model 2 (Online Appendix, Table S11).

Summary of results

When personnel-contributing states incur fatalities, they are likely to reduce the numbers of peacekeepers they commit to the operations where deaths occurred. Greater numbers of fatalities, meanwhile, are associated with larger personnel withdrawals. This relationship holds regardless of fatality type, and is statistically significant for both troops and police (but not observers). Overall, the evidence largely supports Hypothesis 1.

The findings also show that not all states are impacted similarly by peacekeeper fatalities. In particular, there is a distinction between how wealthier and poorer states respond to personnel losses. Whereas wealthier states are likely to decrease the magnitude of their contributions after incurring peacekeeper fatalities, poorer states are likely to increase their contributions after such incidents. The effect of wealth is substantively large, although strongest following malicious acts and accidents. Disaggregation of peacekeeper type, meanwhile, shows that the results are statistically significant in regards to troops, but not police or observers. These findings suggest general support for Hypothesis 2, while clarifying some of the specific conditions under which wealth is most likely to shape states’ responses to peacekeeper fatalities.

The effect of regime type on personnel contributions following peacekeeper deaths, meanwhile, is less robust. Democracies are likely to make greater reductions in contributions than non-democracies after their personnel die, but this outcome is apparent only in the tests that examine fatalities accumulated over a six-month period. Democracies also are more likely to reduce contributions to missions after their personnel die via malicious acts, but no such association exists following other types of fatalities. Meanwhile, there is no statistically significant correlation between regime type and the type of personnel that incur fatalities. When taken together, the evidence suggests only weak support for Hypothesis 3.

Conclusions

Peacekeeping is risky business. Contemporary UN operations often require states to deploy personnel into situations where peace is fragile or where conflict is still ongoing, frequently placing peacekeepers in danger. Even where immediate threats to their security are relatively low, UN personnel can still fall prey to fatal illnesses or become victims of tragic accidents. While states are aware of these possibilities prior to deployment, their occurrence may still be sufficient cause to update their prior calculations of the costs and benefits from participating in a particular operation. The findings in this article suggest support for this proposition, while also identifying factors that make changes in personnel commitment more or less likely. In particular, wealthier countries are more likely than their poorer counterparts to reduce their contributions to a given operation after incurring fatalities. There is less support, meanwhile, for the hypothesis that democracies and non-democracies respond differently to peacekeeper deaths.

These results are relevant to a number of ongoing debates and discussions on peacekeeping. In regards to improving peacekeeping effectiveness, it is important for the UN to understand the conditions under which countries are most and least likely to depart an operation. High rates of personnel turnover can have negative implications for mission efficacy in terms of higher training costs, loss of institutional knowledge, and barriers to building the trust of local communities. While personnel rotations are a given in any military operation, large-scale removal of personnel in response to peacekeeper deaths is not likely to impart confidence in the UN from the “peacekept,” who often live in situations of great insecurity themselves. Understanding the extent to which peacekeeper fatalities lead states to withdraw personnel could help the UN anticipate, address, and ameliorate the negative consequences of mission turnover after peacekeeper losses.

The findings also speak to the broader literature on peacekeeping contributions, and particularly in regards to the relationship between financial incentives and the domination of peacekeeping operations by developing countries. While past studies have shown that poorer countries are more likely to contribute to UN operations, this study indicates that they are also more likely to stay involved in these missions—even when the going gets tough. The policy implications of this are less straightforward. Perhaps the most superficial takeaway is that the UN could make a concerted effort to deploy personnel from developing countries into higher-risk locales, with the knowledge that they are likely to be more resolved to maintain their level of participation regardless of the human costs. However, such an immoral policy would only exacerbate growing concerns about the ethics of wealthy countries paying for peacekeeping, while allowing poorer countries to suffer the risks of deployment. Instead, the focus should be on developing new ways to keep developed countries involved in peacekeeping once they have begun to contribute, even as the costs of participation rise. Given that financial incentives are unlikely to be sufficient to retain wealthy states, innovative non-monetary inducements will be critical to such endeavors.

In terms of scholarship, shifting the focus from the causes to the consequences of peacekeeping fatalities suggests a fertile avenue of future research. One immediate extension to this study, pending data availability on authorized personnel numbers, is to expand the scope of the study outside of Africa to see if regional variation in outcomes emerge. Thinking more broadly, it is likely that peacekeeper fatalities have implications beyond states’ decisions about whether or not to maintain their current level of participation in a particular operation. For instance, to what extent do peacekeeper fatalities impact political decisions about peacekeeping mandates at the level of the Security Council? Do peacekeeping fatalities have implications for the sub-national geography of peacekeeping deployment, leading UN force commanders to shift personnel away from (or towards) certain regions of a host-country? What kind of impact do peacekeeping fatalities have on the overall effectiveness of UN operations? While explaining the causes of peacekeeping fatalities is critical to helping reduce their occurrence, understanding their consequences could lead to important new insights into how the UN manages violent conflict under changing—and often deteriorating—circumstances.

Supplemental Material

replication_files_PK_fatalities_commitments – Supplemental material for Peacekeeper Fatalities and Force Commitments to UN Operations

Supplemental material, replication_files_PK_fatalities_commitments for Peacekeeper Fatalities and Force Commitments to UN Operations by Andrew Levin in Conflict Management and Peace Science

Footnotes

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Notes

Supplementary material

All data, replication materials, and instructions regarding analytical materials upon which published claims rely are available online through the SAGE CMPS website:

References

Almond

(1950) The American People and Foreign Policy. New York: Harcourt, Brace.

Andersson

(2000) Democracies and UN peacekeeping operations, 1990–1996. International Peacekeeping 7(2): 1–22.

Axe

(2010) Why South Asia loves peacekeeping. The Diplomat, 20 December. Available at: http://thediplomat.com/2010/12/why-south-asia-loves-peacekeeping/

BBC News (2017) Dutch Minister resigns over deaths of Mali peacekeepers, 4 October. Available from: https://www.bbc.com/news/world-europe-41498523

Beardsley

(2011) Peacekeeping and the contagion of armed conflict. The Journal of Politics 73(4): 1051–1064.

Beardsley

Schmidt

(2012) Following the flag or following the charter? Examining the determinants of UN involvement in international crises, 1945–2002. International Studies Quarterly 56(1): 33–49.

Bellamy

Williams

(2013) Providing Peacekeepers: The Politics, Challenges, and Future of United Nations Peacekeeping Contributions. Oxford: Oxford University Press.

Bennett

Stam

(1998) The declining advantages of democracy: A combined model of war outcomes and duration. Journal of Conflict Resolution 42(3): 344–366.

Boshoff

(2008) Personnel fatalities in United Nations missions, by year, up to 30 June 2008. African Security Studies 17(3): 71–74.

10.

Bove

Elia

(2011) Supplying peace: Participation in and troop contribution to peacekeeping missions. Journal of Peace Research 48(6): 699–714.

11.

Brambor

Clark

Golder

(2006) Understanding interaction models: Improving empirical analyses. Political Analysis 14(1): 63–82.

12.

Brysk

(2009) Global Good Samaritans: Human Rights as Foreign Policy. Oxford: Oxford University Press.

13.

Bueno de Mesquita

Smith

Siverson

Morrow

(2003) The Logic of Political Survival. Cambridge, MA: The MIT Press.

14.

Burk

(1999) Public support for peacekeeping in Lebanon and Somalia: Assessing the casualties hypothesis. Political Science Quarterly 114(1): 53–78.

15.

Burrows

(2014) Fijian UN peacekeepers: Cash, experience a lure although fears remain following abductions in Syria. ABC News (Australia), 2 September.

16.

Call

Barnett

(1999) Looking for a few good cops: Peacekeeping, peacebuilding and CIVPOL. International Peacekeeping 6(4): 43–68.

17.

Dallaire

(1994) Request for protection for informant. Facsimile to Maj. Gen. Maurice Baril, United Nations Department of Peacekeeping Operations, 11 January.

18.

Diehl

(2008) Peace Operations. Polity: Cambridge.

19.

dos Santos Cruz

(2017) Improving security of United Nations peacekeepers: We need to change the way we are doing business. Independent Report, 19 December.

20.

Dorn

(2008) Who is dying for peace? In: Center on International Cooperation (ed.), Annual Review of Global Peace Operations 2008. Boulder, CO: Lynne Rienner.

21.

Doyle

Sambanis

(2006) Making War and Building Peace: United Nations Peace Operations. Princeton, NJ: Princeton University Press.

22.

Fjelde

Hultman

Lindberg Bromley

(2016) Offsetting losses: Bargaining power and rebel attacks on peacekeepers. International Studies Quarterly 60(4): 611–623.

23.

Findlay

(1996) Challenges for the New Peacekeepers (no. 12). SIPRI Research Reports.

24.

Fortna

(2004) Does peacekeeping keep peace? International intervention and the duration of peace after civil war. International Studies Quarterly 48(2): 269–292.

25.

Fortna

(2008) Does Peacekeeping Work? Shaping Belligerents’ Choices After Civil War. Princeton, NJ: Princeton University Press.

26.

Frontline (2004) Interview with Romeo Dallaire, 1 April. Available at: https://www.pbs.org/wgbh/pages/frontline/shows/ghosts/interviews/dallaire.html

27.

Gaibulloev

Sandler

Shimizu

(2009) Demands for UN and non-UN peacekeeping: Nonvoluntary versus voluntary contributions to a public good. Journal of Conflict Resolution 53(6): 827–852.

28.

Gaibulloev

George

Sandler

Shimizu

(2015) Personnel contributions to UN and non-UN peacekeeping missions: A public goods approach. Journal of Peace Research 52(6): 727–742.

29.

Gartner

Segura

(1998) War, casualties, and public opinion. Journal of Conflict Resolution 42(3): 278–300.

30.

Gelb

(1994) Quelling the teacup wars: The new world’s constant challenge. Foreign Affairs 73(6): 2–6.

31.

Gelpi

Feaver

Reifler

(2009) Paying the Human Costs of War: American Public Opinion and Casualties in Military Conflicts. Princeton, NJ: Princeton University Press.

32.

Gilligan

Stedman

(2003) Where do the peacekeepers go? International Studies Review 5(4): 37–54.

33.

Gleditsch

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

34.

Gleditsch

Ward

(2001) Measuring space: A minimum-distance database and applications to international studies. Journal of Peace Research 38(6): 739–758.

35.

Government Accountability Office (2005) DOD Needs to Improve the Transparency and Reassess the Reasonableness, Appropriateness, Affordability, and Sustainability of Its Military Compensation System (GAO-05-798). Washington, DC: US Government Printing Office.

36.

Gunitsky

(2015) Lost in the gray zone: Competing measures of democracy in the former soviet republics. In: Cooley

Snyder

(eds), Ranking the World: Grading States as a Tool of Global Governance. New York: Cambridge University Press, pp. 112–150.

37.

Henke

(2017) UN fatalities 1948–2015: A new dataset. Conflict Management and Peace Science. https://doi.org/10.1177/0738894216686789

38.

Hensel

(2014) ICOW colonial history data set, version 1.0. Available at: http://www.paulhensel.org/icowcol.html

39.

Holsti

(1996) Public Opinion and American Foreign Policy. Ann Arbor, MI: University of Michigan Press.

40.

Howard

(2008) UN Peacekeeping in Civil Wars. Cambridge: Cambridge University Press.

41.

Hultman

Kathman

Shannon

(2013) United Nations peacekeeping and civilian protection in civil war. American Journal of Political Science 57(4): 875–891.

42.

International Peace Institute (n.d.) IPI Peacekeeping Database. Available at: http://www.providingforpeacekeeping.org/contributions/

43.

Kant

(1795) Perpetual Peace. In: Reiss

(ed.) (1991) Kant: Political Writings. Cambridge: Cambridge University Press.

44.

Karol

Miguel

(2007) The electoral cost of war: Iraq war casualties and the 2004 U.S. presidential election. The Journal of Politics 69(3): 633–648.

45.

Kathman

(2013) United Nations peacekeeping personnel commitments, 1990–2011. Conflict Management and Peace Science 30(5): 532–549.

46.

Kathman

Melin

(2016) Who keeps the peace? Understanding state contributions to UN peacekeeping operations. International Studies Quarterly 61(1): 150–162.

47.

Khanna

Sandler

Shimizu

(1999) The demand for UN peacekeeping, 1975–1996. Kyklos 52(3): 345–368.

48.

Koch

(2011) Casualties and incumbents: Do the casualties from interstate conflicts affect incumbent party vote share? British Journal of Political Science 41(4): 795–817.

49.

Krishnasamy

(2001) “Recognition” for third world peacekeepers: India and Pakistan. International Peacekeeping 8(4): 56–76.

50.

Lebovic

(2004) Uniting for peace: Democracies and United Nations peace operations after the Cold War. Journal of Conflict Resolution 48(6): 910–936.

51.

Lindberg Bromley

(2018) Introducing the UCDP Peacemakers at Risk dataset, sub-Saharan Africa 1989–2009. Journal of Peace Research 55(1): 122–131.

52.

Lippmann

(1955) Essays in the Public Philosophy. Piscataway, NJ: Transaction.

53.

Luttwak

(1996) A post-heroic military policy. Foreign Affairs 75(4): 33–44.

54.

Manin

Przeworski

Stokes

(1999) Elections and representation. In: Democracy, Accountability, and Representation. Cambridge: Cambridge University Press.

55.

Maoz

(1990) Paradoxes of War: On the Art of National Self-entrapment, Vol. 3. Abingdon: Routledge.

56.

Marshall

(2016) Major episodes of political violence (MEPV) and conflict regions, 1946–2015. Center for Systemic Peace.

57.

Marshall

Gurr

Jaggers

(2013) Polity IV Project: Political regime characteristics and transitions 1800–2012. Center for Systemic Peace.

58.

Minde

(2017) Peacekeeping contributor profile: Tanzania. Providing for Peacekeeping. Available at: http://www.providingforpeacekeeping.org/2016/04/05/peacekeeping-contributor-profile-tanzania/

59.

Morrison

(1996) UN peacekeeping reform: Something permanent and stronger. Brown Journal of World Affairs 3(1): 95–110.

60.

Mueller

(1973) War, Presidents, and Public Opinion. Chichester: John Wiley & Sons.

61.

Mueller

(1996) Policy principles for unthreatened wealth-seekers. Foreign Policy (102): 22–33.

62.

Neack

(1995) UN peace-keeping: In the interest of community or self? Journal of Peace Research 32(2): 181–196.

63.

Page

Stevis

(2016) China discovers the price of global power: Soldiers returning in caskets; President Xi Jinping’s quest to make his nation a military player on the world stage provokes soul-searching as the country absorbs its first combat casualties in decades. Wall Street Journal (Online), 15 November.

64.

Passmore

Shannon

Hart

(2017) Rallying the troops: Collective action and self-interest in UN peacekeeping contributions Journal of Peace Research 55(3): 366–379.

65.

Perkins

Neumayer

(2008) Extra-territorial interventions in conflict spaces: Explaining the geographies of post-cold war peacekeeping. Political Geography 27(8): 895–914.

66.

Power

(2001) Bystanders to genocide. Atlantic Monthly 288(2): 84–108.

67.

Przeworski

Alvarez

Cheibub

Limongi

(2000) Democracy and Development: Political Institutions and Well-being in the World, 1950–1990 (Vol. 3). Cambridge: Cambridge University Press.

68.

Pugliese

(2014) Profile: Romeo Dallaire’s peacekeeping nightmare. Ottawa Citizen, 28 May.

69.

Reiter

Stam

(2002) Democracies at War. Princeton, NJ: Princeton University Press.

70.

Rogers

Kennedy

(2014) Dying for peace? Fatality trends for United Nations peacekeeping personnel. International Peacekeeping 21(5): 658–672.

71.

Salverda

(2013) Blue helmets as targets: A quantitative analysis of rebel violence against peacekeepers, 1989–2003. Journal of Peace Research 50(6): 707–720.

72.

Seet

Burnham

(2000) Fatality trends in United Nations peacekeeping operations, 1948–1998. JAMA: Journal of the American Medical Association 284(5): 598–603.

73.

Singer

Bremer

Stuckey

(1972) Capability distribution, uncertainty, and major power war, 1820–1965. In: Bruce

Russett

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

74.

Siverson

(1995) Democracies and war participation: In defense of the institutional constraints argument. European Journal of International Relations 1(4): 481–489.

75.

Smith

(2005) What costs will democracies bear? A review of popular theories of casualty aversion. Armed Forces & Society 31(4): 487–512.

76.

Stojek

Tir

(2015) The supply side of United Nations peacekeeping operations: Trade ties and United Nations-led deployments to civil war states. European Journal of International Relations 21(2): 352–376.

77.

Sundberg

Melander

(2013) Introducing the UCDP georeferenced event dataset. Journal of Peace Research 50(4): 523–532.

78.

Thakur

(1980) Peacekeeping and foreign policy: Canada, India and the International Commission in Vietnam, 1954–1965. British Journal of International Studies 6(2): 125–153.

79.

Thakur

(1984) Peacekeeping in Vietnam: Canada, India, Poland, and the International Commission. University of Alberta.

80.

Timmons

Bagri

(2013) Poor nations fight, rich nations pay. The New York Times, 10 April.

81.

Townsen

Reeder

(2014) Where do peacekeepers go when they go? Journal of International Peacekeeping 18(1–2): 69–91.

82.

Uzonyi

(2015) Refugee flows and state contributions to post-Cold War UN peacekeeping missions. Journal of Peace Research 52(6): 743–757.

83.

Uz Zaman

Biswas

(2013) Contributor profile: Bangladesh. Providing for Peacekeeping. Available from:http://www.providingforpeacekeeping.org/2014/04/03/contributor-profile-bangladesh/

84.

Van der Lijn

Smit

(2015) Peacekeepers Under Threat? Fatality Trends in UN Peace Operations. SIPRI.

85.

Verba

Brody

Parker

Nie

Polsby

Ekman

Black

(1967) Public opinion and the war in Vietnam. American Political Science Review 61(2): 317–333.

86.

Ward

Dorussen

(2016) Standing alongside your friends: Network centrality and providing troops to UN peacekeeping operations. Journal of Peace Research 53(3): 392–408.

87.

Weeks

(2014) Dictators at War and Peace. Ithaca, NY: Cornell University Press.

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.

0.68 MB

0.00 MB