Paper

The Impact of Human Rights INGO Shaming on Humanitarian Interventions

Amanda Murdie University of Missouri Dursun Peksen University of Memphis

Do transnational human rights organizations (HROs) influence foreign military intervention onset? We argue that the greater international exposure of human suffering through HRO ‘‘naming and shaming’’ activities starts a process of mobilization and opinion change in the international community that ultimately increases the likelihood of humanitarian military intervention. This is a special corollary to the supposed ‘‘CNN Effect’’ in foreign policy; we argue that information from HROs can influence foreign policy decisions. We test the empirical implication of the argument on a sample of all non-Western countries from 1990 to 2005. The results suggest that HRO shaming makes humanitarian intervention more likely even after controlling for several other covariates of intervention decisions. HRO activities appear to have a significant impact on the likelihood of military missions by IGOs as well as interventions led by third-party states.

H umanitarian military interventions are potentially costly endeavors for intervener countries due to the low toleration of the

public for casualties and the high economic and political risks involved in military operations (Gartner and Segura 1998; Gartner, Segura, and Barratt 2004; James and O’Neal 1991; Mueller 1973).1 Despite the significant risks that interventions involve for the in- terveners, the external use of force for humanitarian reasons remains a popular foreign policy tool in the post-Cold War era, justified often by ideas about the world community’s ‘‘responsibility to protect’’ civilians in harm’s way. From 1990 to 2005, on average more than 10 military operations occurred a year for largely humanitarian goals (Pickering and Kisangani 2009). These goals range from ending civil wars to preventing genocides, to rebuilding failed states, or to promoting human rights and democracy.

Given both the possible high risks for interveners and the potential benefits for the targeted countries, what determines when humanitarian interventions take place and when the world community chooses,

for better or worse, to sit on the sidelines and not intervene?2 In other words, what leads to ‘‘boots on the ground’’ for humanitarian reasons? Most existing explanations of the determinants of intervention decisions tend to be state-centric and focus on geo- political and economic factors, often ignoring the potential of nonstate actors to influence foreign policy decision making regarding intervention (Beardsley and Schmidt 2012; Choi 2012; Fordham 2008). Other literature highlights the possibility of a ‘‘CNN effect’’ on intervention decision making, where emotional coverage of both humanitarian disasters and military casualties in global media outlets has been argued to ‘‘drive . . . policy formulation’’ (Robinson 2011, 2). To date, however, empirical support of a general media effect on intervention decision making has been scant, and recent literature on media effects has stressed the need for examining alternative ‘‘global networks’’ through which information is produced and disseminated in ways that could influence foreign policymaking (Gilboa 2005, 337).

The Journal of Politics, Vol. 76, No. 1, January 2014, Pp. 215–228 doi:10.1017/S0022381613001242

� Southern Political Science Association, 2013 ISSN 0022-3816

1Both authors contributed equally to the article. An online appendix with supplementary materials is available at http://dx.doi.org/10.1017/ S0022381613001242. Data and supporting materials to reproduce the numerical results are available at http://www.amandamurdie.org/ research.html.

2 In this study, we define humanitarian armed intervention as ‘‘the threat or use of force across state borders by a state (or group of states) aimed at preventing or ending widespread and grave violations of the fundamental human rights of individuals other than its own citizens, without the permission of the state within whose territory force is applied’’ (Holzgrefe 2003, 18).

215

In this article, we focus on one particular alternative ‘‘global network’’ of information: the work of human rights nongovernmental organizations (human rights INGOs or, as used hereafter, HROs).3 We argue that the information produced by HROs is a special corollary of the supposed ‘‘CNN effect’’ on foreign policymaking and can drive the decision of humanitarian interven- tion. Human rights organizations, like the well-known Amnesty International or Human Rights Watch, work to improve human rights mainly through the strategic reporting of stories on grave human rights situations and changing human rights conditions. Many HROs collect information on abuses and work to ‘‘name and shame’’ or ‘‘shame and blame’’ a targeted regime in the popular press, mobilizing others to take actions to protect a repressed population from ongoing abuse and precarious human rights situations (Keck and Sikkink 1998; Murdie and Davis 2012). Human Rights Watch, for instance, was a frequent source of information on the human rights atrocities in Libya under the Gaddafi regime. After the fighting first intensified in February of 2011, Washington director of Human Rights Watch, Tom Malinowski, called on the United States and other developed countries to act to stop the ‘‘ongoing mas- sacre in Libya.’’ When the UN Security Council passed a resolution authorizing the use of military forces to protect civilians in Libya in March 2011, Human Rights Watch executive director Kenneth Roth remarked that the Security Council had ‘‘at last lived up to its duty’’ (2011). Other organizations have released reports of human rights atrocities and called for foreign policy actions in Sri Lanka, Sudan, and Nepal, among other locations, in the last 20 years.

We therefore argue that there is an ‘‘Amnesty International effect’’: human rights organizations influ- ence the likelihood of humanitarian armed interven- tions. Unlike the general media, HROs are often highly credible actors and work to directly mobilize action related to the ‘‘responsibility to protect.’’ As such, even after taking the actual human rights situation within the country and its general media coverage into account, when HROs increase attention of human rights prob- lems and vulnerabilities within a regime through a ‘‘naming and shaming’’ strategy, it is more likely that the world community will take notice and intervene, either unilaterally or as part of a collective international effort. Our argument does not rest on whether HROs

are actively calling for military intervention or not; simply by bringing attention to a population’s vulner- abilities to current and potential future atrocities, HROs are providing information and mobilizing in ways that can sway foreign policy decision making in favor of a humanitarian intervention.

We test the empirical implication of our argument on a sample of all non-Western countries from 1990 to 2005. In line with our argument, we find that HRO shaming makes humanitarian intervention more likely, even after accounting for several other covariates of intervention decisions.

Determinants of Humanitarian Intervention

Our theoretical argument of an ‘‘Amnesty International effect’’ draws on literatures concerning intervention onset, the media and foreign policy, and human rights transnational advocacy. To begin, a variety of political and economic factors have been argued to affect the likelihood of armed interventions. Much of the litera- ture is divided on whether the severity of the crisis at hand or the strategic and economic importance of the location where the crisis is occurring matter for intervention, with many studies finding some com- bination of these determinants to be at play (e.g., Beardsley and Schmidt 2012; Choi 2012; Fordham 2008; Mullenbach 2005; Mullenbach and Matthews 2008; Rost and Grieg 2011).

Beyond these political and economic determinants of intervention, some work has examined whether emotive press coverage of humanitarian crises, often termed as the ‘‘CNN effect,’’ is ‘‘capable of indepen- dently driving and influencing policy formulation, at times against the interests and wishes of foreign-policy establishments’’ (Robinson 2011, 23; 2000; Gilboa 2005; Jakobsen 2000). Theoretically, as media coverage generates more public interest on an issue, voters, interest groups, and other influential domestic actors will appeal to leaders to take action. This media-driven public pressure will create a policy imperative for pol- iticians to ‘‘do something’’ to satisfy citizens’ expect- ations and avoid media criticism. If this press coverage concerns human rights conditions and the potential for future atrocities, it follows that this information will help embolden public demands for government action to help a repressed population (Jakobsen 2000; Robinson 2000).

The suggested effect of the media was largely touted in the 1990s, as many popular writings and foreign policy experts mentioned the role of the

3We define human rights INGOs as any nonprofit organization that has a mission statement consistent with the Universal Declaration of Human Rights (UDHR), is not controlled by a government, and has an international presence (Keck and Sikkink 1998).

216 amanda murdie and dursun peksen

media in driving foreign policy. For example, then UN secretary-general Boutros Boutros-Ghali commented in 1995 that CNN was effectively an additional member of the UN Security Council. Prior to the 1993 ‘‘Blackhawk Down’’ coverage of American casualties in Somalia and the international community’s inaction in the Rwanda genocide in 1994, much of the discussion of the CNN effect was that it would make humanitarian actions more likely.

Despite the strong theoretical appeal of the media effect argument, most studies conclude that the effect of news coverage of human suffering on armed interventions is very limited (Gilboa 2005; Jakobsen 2000; Robinson 2000). Robinson, in a recent review of the literature, concludes that the ‘‘majority of researchers would agree that news media do not have the transformative influence attributed to them at the height of the 1990s CNN effect debate’’ (2011, 5). Other scholars raise the possibility that the news media is partisan, elite driven, and motivated by profits (e.g., Baum and Groeling 2009; Entman 2004; Groeling and Baum 2009). As such, the media may not be a unified mover of public opinion and may have interests that already match elite opinions on an issue or produce divergent public pressure, perhaps leading to overall inaction. Because of this, the effect of the general media as a driver of foreign policy may be overstated, including its role in driving humanitarian interven- tions (Robinson 2011).

Our study complements and adds to the relevant literatures on armed interventions and news media effects by focusing on the possible impact of a specific set of nonstate actors (HROs) on the likelihood of humanitarian interventions. More specifically, we argue that the work of human rights organization often provides the agenda for mass media coverage and the information these organizations produce can serve as a motivator for foreign policy action, including humanitarian interventions into conditions ripe for mass atrocities. Further, HROs can be direct mobilizers for public action and can work directly with foreign policy decision makers. In the following section, we develop a theoretical framework that explains the causal relationship between HRO shaming and the initiation of humanitarian missions.

The Role of Human Rights Organizations in Humanitarian

Intervention Decisions

The extant literature on human rights organizations is clear that HROs work to provide information about

human rights conditions to the global community, especially during times of severe conflict and human rights disasters when international journalists may have already left a war-torn area for fear of the situ- ation deteriorating further (Keck and Sikkink 1998; Murdie and Davis 2012; Ron, Ramos, and Rodgers 2005). Theoretically, as part of a larger transnational network, the organizations respond to calls by re- pressed populations and then disseminate information to increase both IGO and third-party state attention to the situation through a ‘‘shaming and blaming’’ or ‘‘naming and shaming’’ strategy (Keck and Sikkink 1998). By ‘‘naming’’ the perpetrators of abuses and ‘‘shaming’’ them in the popular press, organizations hope to expand a transnational network of actors that will ultimately influence the decision calculus of the repressive regime, leading them to stop a repressive campaign or rethink their response to political oppo- sition. Bob (2005) has even argued that organizations only respond to the most news worthy of populations. Although the power of HROs may be ‘‘soft,’’ through influencing these third-party actors, ‘‘hard’’ political actions could be undertaken to help the repressed population. Past work has found that shaming by HROs can influence the decision of states to place economic sanctions on a repressive regime (Murdie and Peksen 2013), and shaming by UN actors has been linked to changes in foreign aid (Lebovic and Voeten 2009). They also influence foreign direct investment (Barry, Clay, and Flynn 2013) and can change a local population’s opinion about their own human rights conditions (Davis, Murdie, and Steinmetz 2012). In short, HROs are considered influential actors for transnational advocacy and serve as agenda setters when it comes to human rights action. The informa- tion they produce is frequently compiled and released on a global scale (Ron, Ramos, and Rodgers 2005).

Now, to be clear, HROs vary in their actual calls for intervention. Some have been outright advocates of military interventions in the past; many groups actively supported the 1991 intervention in Iraq, Somalia in 1993, and recent action in Sudan (Kaldor 2001; Sudan Tribune 2012). Others have been out- spoken opponents against any sort of military action, including the hundreds of organizations that came out against U.S. intervention in Iraq in 2003 and the group African Rights, which actually broke away from Human Rights Watch over opposition to military intervention in Somalia in 1993, contended that the Somalia intervention ‘‘was not designed to address Somalia’s real problems and was in fact destined to make things worse’’ (AR 2012). The then-secretary general of Amnesty International, arguably the largest

human rights shaming and interventions 217

and most influential human rights organization (Wong 2012), even said in 2000 that the organization does ‘‘not call for military action, nor . . . oppose it, but [does] campaign on how such interventions should be conducted’’ (Sané 2000, 89).

Our argument, however, does not rest on whether the organization is actively in support of intervention or not: the information that HROs produce is espe- cially designed to motivate public opinion and sway others to join the cause to ‘‘do something’’ in the name of helping the vulnerable population. Regardless of whether other actions of the HRO are to influence the nature of this action, the information alone against a repressive regime should embolden international community responses. This is clear in the canonical literature on how HROs are supposed to work to improve human rights in a repressive state (Keck and Sikkink 1998; Risse, Ropp, and Sikkink 1999).

Although this argument may seem straightfor- ward, it flies in stark contrast to many of the dominant theoretical arguments about foreign policy decision making and the general influence of nonstate actors on state behavior (Snyder et al 2003; Waltz 1979). Many HROs have negligible budgets, and most are not ‘‘at the table’’ when it comes to even minor foreign policy decisions, let alone the decision of a country to put its troops in harm’s way for humanitarian purposes (Fitzduff and Church 2004). Realists, for instance, often see HROs as simply instruments of the state; as such, the influence of the information pro- duced by HROs should not be influential in foreign policy decision making once the underlying political and economic factors are considered (Geeraerts 1995).

With all this going against a likely effect, then why should the information produced by HROs matter for the decision to intervene within a country? First, we argue that HROs are often seen as highly credible actors in a sea of misinformation. Because of this, unlike general news reports, reports that cite a HRO will be more likely to be believed, leading to a more likely change in public and elite opinion and/or pressureforactiononbehalfoftherepressedpopulation. This is not to say that exaggerations or misinformation by HROs do not occur but that organizations have very real constraints on their ability to exaggerate and need credibility in order to gain supporters (Gourevetich and Lake 2012). Reports of incredible stories or partisan biases make front-page news and can constrict support (Shavin 2012).

In general, the public has more confidence in HROs and related organizations than in television or government. The World Values Survey asks individu- als about their confidence in various aspects of society.

According to their survey, between 1995 and 2008 globally, 62.49% of individuals say their confidence in ‘‘charitable or humanitarian organizations,’’ the clos- est categorization in the World Values Survey to HROs, is ‘‘quite a lot’’ or ‘‘a great deal.’’ Only 47.18% say something similar about television, and 47.21% say the same thing about government. The disparity between confidence in organizations and confidence in either the government or television is even greater in Western countries, which could be argued to be the likely sources of any global push for humanitarian intervention. We think these survey results help illustrate the potential power of HRO information.4

Even though HROs are being argued to work through the same general causal pathway that is supposed to link CNN and other news outlets to foreign policy action, on the whole, HROs are more credible actors. As such, stories and reports that cite HROs should be especially important movers of action for humanitar- ian purposes, including foreign military interventions.

Because the motivations of HROs are often dif- ferent than mainstream media, they can also provide information on cases that have been missing in existing general media reports or can bring attention to issues which may have dropped off of the general public’s radar. Many times, HROs heighten their reporting about a country because they are concerned about the escalation of ongoing human rights abuses that could turn into mass killings and other human rights disas- ters. For example, in the summer of 2012, Amnesty International released a report on the how ‘‘the fre- quency and brutality of [Syrian] government reprisals against towns and villages supportive of the opposition has escalated’’ (2012, 8). The Amnesty International official responsible for the report remarked that in- ternational actions must be taken to stop the escalation; people in Syria ‘‘were asking why the international com- munity has virtually deserted them, and that is what people just don’t understand’’ (PRI 2012). Similarly, Amnesty International has recently called on action in Sudan, saying ‘‘the UN Security Council and African Union (AU) must take immediate action to halt indiscriminate attacks in Southern Kordofan’’ or the situation ‘‘will only get worse’’ (AI 2013). These types of shaming activities call for international community actions in ongoing situations in order to avoid escala- tion of atrocities.

Further, HROs also have made their organiza- tional survival rest on providing poignant, vivid accounts of existing human rights atrocities and the

4Please see the online appendix for a table of these summary results.

218 amanda murdie and dursun peksen

possibility of further abuses, often directly from those affected by the situation. Many HRO reports are the most troubling about an atrocity, providing a real name and face to a foreign and distant situation. Because of this, these accounts could be among the most memorable ones for the general public, pro- viding a focal point for action (Risse, Ropp, and Sikkink 1999).

There are many accounts of shaming information by HROs making its way into public statements, bills, and policies in support of armed intervention. Take, for example, some anecdotes from the United States, a frequent state involved in armed intervention. Amnesty International’s report on political killings and torture made it into a U.S. Congressional Record on January 11, 1991, entitled ‘‘Congressional approval of military action to enforce U.N. Security Council resolution against Iraq pending failure of economic sanctions’’ (CR-1991-0111). In the discussions before the U.S. House Resolution entitled ‘‘Peacekeeping Operations in Kosovo Resolution’’ (145 Cong Rec H 1179), congressmen from Michigan and Florida re- ferred to information from Human Rights Watch concerning ongoing killings by Serbians of ethnic Albanians ‘‘over the age of 15.’’ Right after referring to the evidence from Human Rights Watch, former representative David E. Bonior (Michigan) then continued on the U.S. House floor by saying that something must be done: ‘‘Why would Milosevic do anything but stall, not agree to a peace agreement, if the United States Congress says in a vote later today . . . that we, in fact, will not deploy troops?’’ (145 Cong Rec H 1179).

Additionally, unlike the examples of both the benefits and costs of intervention in the general media (Jakobsen 2000), HRO information is more likely to be in one direction: for action—any action—in support of the civilian population where the atrocities are occurring. HRO reports are unlikely to include accounts of the potential military costs of intervention. In fact, we read all accounts of shaming by HROs in our dataset, discussed below, for the years 2004 and 2005 and did not find any evidence that HROs were reporting on the potential costs to the intervening state of a possible intervention.

Finally, HROs can couple their shaming efforts with other forms of mobilization in support of action on behalf of the targeted state, including directly working with policy makers or urging their members to take actions on their behalf (Fitzduff and Church 2004). As such, unlike news media outlets that are not directly pushing foreign policy leaders to take certain policy positions, some HROs will couple their informa-

tion with direct interaction with policy leaders and mobilization campaigns (Keck and Sikkink 1998). For example, in the 1980s, as Forsythe discusses, many HROs, including Amnesty International and the organ- izations now referred to as Human Rights Watch and Human Rights First, met with the U.S. House Foreign Affairs Subcommittee on Human Rights and Interna- tional Organizations (1988, 125–26). HROs frequently testify in hearings in the U.S. Congress; since 2008, when Congress created the Tom Lantos Human Rights Commission, HROs have been included in hearings and briefings almost bimonthly.5 Similarly, HROs have frequently testified in the Canadian House of Commons and the U.K. Foreign Affairs Committee, including Amnesty International U.K. recently providing evidence of human rights abuses in Libya. In a recent written report to the U.K. Foreign Affairs committee, for example, Amnesty International recommended that ‘‘The U.K. Government must put human rights at the top of its agenda likewise in its relations’’ with the Libyan government (‘‘Further Written Evidence From Amnesty International UK’’ 2011).

In short, this line of reasoning implies a testable hypothesis: even after accounting for political/economic factors and any media effect, information by HROs should still increase the likelihood of humanitarian military intervention. The rest of this article tests this proposition using a large sample of non-Western states in the post-Cold War time period.

Research Design

To statistically examine the impact of HRO shaming on armed humanitarian interventions, we gathered time-series, cross-section data for the 1990–2005 period, inclusive. The country-year is the unit of anal- ysis. That is, each datum represents a country i in a given year t. In order to avoid any bias resulting from case selection, the data analysis excludes all Western liberal democracies from the analysis. Liberal de- mocracies are unlikely targets of interventions since the likelihood of complex humanitarian emergencies in democratic countries are considerably lower than countries with low respect for political rights and civil liberties (e.g., Fearon and Laitin 2003; Pickering 2002). Overall, the data analysis includes 129 coun- tries for which the data are fully available.

5See http://tlhrc.house.gov/

human rights shaming and interventions 219

Outcome Variable: Armed Humanitarian Interventions

We use data on armed humanitarian interventions from the International Military Intervention (IMI) dataset (Pearson and Baumann 1993; Pickering and Kisangani 2009). Within the dataset, military in- tervention is defined as ‘‘the movement of regular troops or forces (airborne, seaborne, shelling, etc.) of one country into the territory or territorial waters of another country, or forceful military action by troops already stationed by one country inside another, in the context of some political issue or dispute’’ (Pearson and Baumann 1993, 4).

As noted above, in this study we focus on nonconsensual armed interferences with the goal of mitigating and/or ending ongoing humanitarian crises. To correspond with this definition, we cre- ated a dichotomous Humanitarian Intervention variable based on the information from the ‘‘social-protective intervention’’ and ‘‘humanitarian intervention’’ variables in the IMI dataset. The Humanitarian Intervention variable takes the value of one for the year a human- itarian intervention is initiated in a country and zero otherwise. 6 According to the IMI, the social-protective interventions involve all armed operations with the stated goal of protecting minority groups facing severe socioeconomic, military, or political threat to their physical security. The humanitarian missions, on the other hand, involve military deployments with such goals as saving lives, relieving suffering, and providing assistance during complex humanitarian emergencies such as civil wars and genocides.7

Covariates of Humanitarian Interventions

To assess the impact of HRO shaming on the likelihood of humanitarian interventions, we use two different

measures, HRO Shaming (count) and HRO Shaming (intensity) (Murdie and Davis 2012). These measures account for all negative HRO events directed at a country in a given year that are captured in Reuters Global News Service. The data are originally collected by the Integrated Data for Event Analysis (IDEA) project and were provided by Virtual Research Associates (Bond et al. 2003).

The HRO Shaming (count) variable captures the number of shaming events in Reuters Global News Service directed at a government or government agent. The event is determined to be shaming if it can be coded as ‘‘conflictual’’ using the Goldstein (1992) scale of the level of conflict and cooperation standard in events data. The data captures these events for over 1,100 HROs, as identified by their mission statements in the 2008 Yearbook of International Organizations, the standard reference for international nongovernmental organization (INGO) data (Boli and Thomas 1999). We use the shaming count to explore whether the number of naming and shaming events increases the probability of an external armed inter- vention with humanitarian motives. Our second mea- sure, HRO Shaming (intensity), captures the intensity of total shaming events. Intensity here comes from the Goldstein (1992) scale; this scale has been reversed so that a more intensively negative shaming event would be given a higher weight on the scale. We use the intensity measure to investigate whether the extent of critical statements and actions by HROs increase the probability of a humanitarian intervention in a poten- tial target country.8

To avoid the omitted variable bias, we also add several other major covariates of humanitarian interventions on the right-hand side of the equation. To account for the impact of the overall level of human rights violations, we include the Human Rights Abuses variable, which is the Political Terror Scale (PTS) (Gibney, Cornett, and Wood 2010). We expect that countries with significant human rights abuses are more likely to be targeted with external humanitarian interventions. The PTS data include two index variables—the State Department Index and the Amnesty International Index—which are 5-point ordered indices that provide information regarding the magnitude and severity of abuses of physical integrity rights. The indices range from 1 to 5

6Summary statistics for all variables can be found in our online appendix. We also ran models where we separated out IGO-led and non-IGO led humanitarian interventions. IGO-led interven- tions were those conducted under the auspices of regional or global intergovernmental organizations (e.g., the United Nations, European Union, and North Atlantic Treaty Organization); non- IGO interventions were military missions led by individual countries without any IGO involvement. We find that HRO shaming has similar positive effects across both intervention types (see the online appendix).

7The IMI also considers the deployment of military forces to provide assistance during natural disasters and to evacuate foreign nationals from a country as humanitarian interventions. We exclude those cases from the analysis since we are particularly interested in the question of whether external actors are inclined to use force to protect the citizens of another country facing threats to their basic human rights and physical security.

8 One important issue that requires attention here is the possible reciprocal causation (i.e., simultaneity bias) between HRO activities and the initiation of humanitarian interventions. We address this issue in detail using a Granger causality test below and provide additional tests based on an instrumental variable approach in our online appendix.

220 amanda murdie and dursun peksen

where higher scores indicate a higher level of human rights abuses. In this article, the State Department index was used, which covers more countries and time points to consider in the analysis.

To explore the impact of news media coverage on the probability of armed missions with humanitarian motives, we use two different measures. The first measure, Media Exposure, controls for the total num- ber of reports about a country appearing in Reuters Global News Service in a given year (Bell, Clay, and Murdie 2012; Murdie and Davis 2012). We use the natural log of this variable to correct for the skewness of the data. The second media variable, Negative News Media Coverage, focuses only on news media coverage concerning human rights violations. This variable controls for the total number of media reports ap- pearing in the Economist and Newsweek with the keywords ‘‘human rights.’’ The data come from Ramos, Ron, and Thoms (2007). Because the data are available until 2000, the time frame for the models controlling for Negative News Media Coverage is from 1990 to 2000.9

Countries with strong economic and military capabilities are less likely targets of military intervention (Pickering 2002). This is due to their ability to deter adverse military interventions from neighboring states or major powers in the international system. Moreover, states with strong socioeconomic and military capa- bilities are more effective in avoiding domestic insta- bilities (Fearon and Laitin 2003). To account for State Capacity, we use the Correlates of War’s (COW) the Composite Indicator of National Capability index, which includes total population, urban population, iron and steel production, energy consumption, military personnel, and military expenditure (Singer 1988). The State Capacity variable is logged to correct for the skewness of the data.

Earlier research finds that democratic states are less frequent targets of military intervention (Pickering

2002). The Democracy variable is the polity score derived from the Polity IV dataset (Marshall and Jaggers 2000), and each country’s democracy score ranges from 210 to 10, where 10 represents the highest level of democracy. The Oil Producer variable is a dichotomous measure that takes the value of 1 if a country receives more than one-third of its export revenues from oil exports and 0 otherwise. This variable controls for the possibility that oil-rich countries un- dergoing humanitarian emergencies are more likely to face an armed intervention (Choi 2012; Fordham 2008).

The Ethnic Fractionalization variable accounts for the suggested positive relationship between ethno-linguistic diversity and violence. It varies from 0 (total homogeneity) to 1 (total heterogeneity). Ethnically divided societies are more prone to in- ternal conflicts (Horowitz 1985). The model also includes the squared term of ethnic fractionalization, Ethnic Fractionalization Squared, to control for the possible curvilinear association between ethnic di- versity and the likelihood of foreign interventions. Collier and Hoeffler (1998) claim that if a social group or a few groups hold the majority of the total population, the likelihood of intercommunal conflict and discrimination raises to considerably high levels, which might subsequently trigger external interven- tion. The data for oil and ethnic fractionalization are from Fearon and Laitin (2003).

A Civil War dummy variable accounts for the presence of violent insurgencies in a country. Civil War is coded as 1 if a country experiences a civil war with at least 25 battle-related deaths per year and 0 otherwise. Third-party military interventions are more likely to occur in war-torn societies to reduce the human cost of the violence, cease the conflict between the state and rebel forces, or to undermine any possible negative spillover effects of civil wars threatening the regional stability such as massive population displace- ment (Regan 2000). We use the list of intrastate armed conflicts collected by the UCDP/PRIO Armed Conflict Dataset (Gleditsch et al. 2002). We also control for the presence of politicide/genocide (Genocide) in a country in a given year to examine whether countries under- going mass killings and ethnic cleansing are more likely to face armed interventions. This data is from the Political Instability Task Force (Marshall, Gurr, and Harff 2012).

The model also includes the Economic Sanctions variable, coded 1 for the years that a country is under economic coercion, 0 otherwise. We control for this variable to explore whether military intervention becomes more likely especially in countries where

9It is worth pointing out that the negative media coverage variable is not highly correlated with the HRO shaming count (0.27) and intensity (0.23) variables. This should not be surpris- ing since the shaming variables are much more comprehensive than the negative media coverage variable in capturing the extent of human suffering in potential target countries. The negative media coverage variable relies only on two Northern media outlets and, more importantly, underestimates the extent of humanitarian problems by neglecting accounts of human rights abuses without the keywords ‘‘human rights.’’ The shaming measures also capture a very specific set of actors (HROs) instead of any general reports using these keywords. Similarly, the media exposure variable that accounts for the total number of news reports has a relatively low correlation with the HRO shaming count (0.25) and intensity (0.15) variables.

human rights shaming and interventions 221

economic coercion fails to address an ongoing humanitarian crises. The sanctions data are from Hufbauer et al. (2007). To account for any region- specific unobserved factors, we also include region dummy variables in the model, including Asia (reference category), Eastern Europe, Latin America, Middle East/N. Africa, and Sub-Saharan Africa.

Methodological Approach

We estimate the models using rare-events logistic regression (King and Zeng 2001). We use rare-event logistic regression since humanitarian interventions are rare events in the data set.10 We include a count variable, Past Intervention, which accounts for the years since the last humanitarian intervention in a country. This measure allows us to control for the possibility that a country that faced a humanitarian intervention in its recent history is a more likely target of a new intervention. This variable also helps us correct for the autoregressive process (temporal dependence) in the models, which is a common issue when cross-sectional time-series data are utilized (Beck and Katz 1995). All models are estimated using Huber-White corrected robust standard errors (clus- tered on countries) to deal with the heteroskedasticity problem. To reduce the possibility of simultaneity bias and make sure that the independent variables precede the dependent variable, we lag the right-hand side-control variables one year, except for our variable of IDEA-based Media Exposure, which is not lagged in order to correspond to the year of the IDEA-based HRO shaming variable in all models.

Findings

Table 1 reports the models examining the impact of HRO shaming on the onset of humanitarian mis- sions. The first two models are estimated using the count of HRO shaming events as the key independent variable, and the second two models are estimated using the HRO shaming intensity variable. In the first and third models, we use the media exposure variable to capture the overall number of news reporting on each country in a given year. In the second and forth models, we replace the media exposure variable with the negative media coverage measure. The results in

Table 1 lend strong support for our assertion that HRO shaming activities increase the probability of humanitarian interventions.

Among the control variables in Table 1, the media exposure (significant at .1 level in the third model), state capacity, ethnic fractionalization, and past in- tervention variables are statistically significant in the expected direction. One interesting finding we have is that presence of genocide/politicide in a country is more likely to cause a unilateral (i.e., non-IGO) rather than a multilateral intervention. It appears that individual countries are more conducive to resorting to armed forces to prevent mass killings and crimes against humanity than IGOs, which often need the support and approval of their member countries. The findings also suggest that negative media coverage, democracy, civil wars, oil producer, and presence of economic sanctions are not significantly correlated with the onset of humanitarian interventions. The insignificant result for the negative media coverage is consistent with the ‘‘CNN’’ effect literature that has shown that the effect of news coverage of human suffering on armed interventions is limited (Gilboa 2005; Jakobsen 2000; Robinson 2000). Based on the findings, it also appears that presence of sanctions in a country does not trigger armed intervention because intervener countries might choose economic coercion as a substitute to the use of military forces.

What do these results imply substantively? When we hold all other continuous variables at their means and the binary variables at their medians, we find that a one standard deviation increase in the mean values of the shaming count and intensity variables increase the predicted probability of humanitarian interven- tion by 60% and 80%, respectively. These numbers suggest that the shaming variables have significant substantive impact in predicting the onset of human- itarian missions. A one standard deviation increase in the average values of the human rights and media exposure variables, on the other hand, increases the probability of humanitarian intervention by 160% and 120%, respectively. The same amount of increase in the mean values of the state capacity and past intervention measures decrease the predicted proba- bility of intervention by 80% and 40%, respectively. A one standard deviation positive change in the mean value of the ethnic fractionalization variable decreases the predicted probability of intervention by 92%, while the same amount of change in the average value of the fractionalization squared variable increases the predicted probability of intervention by about 1220%, reflecting the non-monotonic effect of ethnic fractionalization on civil war processes

10 We also estimated the models using the generalized estimating

equation (GEE) technique. We find very similar results to the ones presented here (see the online appendix).

222 amanda murdie and dursun peksen

found elsewhere in the general literature (Collier and Hoeffler 1998).11

Extending the Analysis: Causality and Causal Process

The robust relationship between HRO shaming and humanitarian intervention shown in Table 1 provides some evidence in support of our main hypothesis. In order to further illustrate how HRO shaming can influence humanitarian intervention decision making,

we ran a number of causality and causal process robustness tests. First, we used a Granger causality test (Granger 1969). Granger causality is a statistical concept of causality denoting that a variable X causes a variable Y, if the past values of X predict Y even when the past values of Y are controlled for. It is an F-test on the coefficients (beta) and the null hypothesis is a variable X does not Granger-cause a variable Y. To run the test, we first estimated two logit models that include humanitarian intervention as the outcome variable and control for one- and two-year lagged values of the shaming variables separately. The models also include one- and two-year lagged values of the intervention variable. When we also include three- or four-year lagged values of the variables, the results from the test were very similar. We then ran two

TABLE 1 HRO Shaming and Humanitarian Interventions (Rare-Event Logit Models)

HRO Shaming (Count) HRO Shaming (Intensity)

HRO shaming (count) 0.615** (0.165)

0.664* (0.270)

HRO shaming (intensity) 0.340** (0.074)

0.273** (0.106)

Human rights abuses 0.860** (0.272)

0.919** (0.220)

0.941** (0.241)

0.963** (0.208)

Media exposure 0.458* (0.235)

0.445 (0.245)

Negative news media coverage -0.013 (0.068)

-0.016 (0.065)

State capacity -0.815** (0.237)

-0.847** (0.223)

-0.814** (0.235)

-0.764** (0.190)

Democracy -0.026 (0.030)

-0.026 (0.035)

-0.026 (0.029)

-0.021 (0.033)

Ethnic fractionalization -9.271** (3.085)

-9.222* (4.014)

-8.495** (3.159)

-9.119* (3.923)

Ethnic fractionalization squared 10.34** (3.522)

9.804* (4.643)

9.818** (3.562)

10.050* (4.548)

Civil war 0.456 (0.568)

0.741 (0.658)

0.352 (0.506)

0.607 (0.615)

Oil producer -0.177 (0.607)

0.202 (0.687)

-0.002 (0.544)

0.230 (0.707)

Economic sanctions 0.356 (0.266)

0.422 (0.344)

0.301 (0.263)

0.414 (0.332)

Genocide 0.559 (0.459)

0.386 (0.428)

0.668 (0.449)

0.470 (0.449)

Past intervention -0.026* (0.012)

-0.026 (0.019)

-0.027* (0.013)

-0.027 (0.019)

Region dummies Yes Yes Yes Yes Constant -14.13**

(2.634) -11.36** (2.196)

-14.39** (2.699)

-10.71** (1.824)

Observation 1,933 1,430 1,933 1,430

Note: Robust standard errors adjusted for clustering over country appear in parentheses. *Significant at .05. **Significant at .01. Regional dummies not shown to save space.

11A table of these predicted probabilities can be found in the online appendix.

human rights shaming and interventions 223

TABLE 2 HRO Shaming and Humanitarian Interventions (Conditional Mixed-Process Recursive Regression)

HRO Shaming (Count) HRO Shaming (Intensity)

DV: Humanitarian Intervention HRO shaming (count) 0.353**

(0.081) 0.345** (0.133)

HRO shaming (intensity) 0.178** (0.036)

0.113* (0.054)

Human rights abuses 0.448** (0.112)

0.428** (0.096)

0.478** (0.106)

0.405** (0.099)

Media exposure 0.233* (0.105)

0.240* (0.109)

Negative news media coverage -0.028 (0.028)

-0.027 (0.028)

State capacity -0.455** (0.104)

-0.429** (0.090)

-0.460** (0.105)

-0.381** (0.079)

Democracy -0.008 (0.013)

-0.004 (0.016)

-0.008 (0.013)

-0.002 (0.016)

Ethnic fractionalization -4.440** (1.351)

-4.308** (1.688)

-4.080** (1.416)

-3.951* (1.569)

Ethnic fractionalization squared 4.945** (1.535)

4.598* (1.958)

4.670** (1.596)

4.354* (1.802)

Civil war 0.252 (0.236)

0.409 (0.285)

0.205 (0.226)

0.280 (0.253)

Oil producer -0.082 (0.238)

0.060 (0.291)

-0.040 (0.234)

0.011 (0.278)

Economic sanctions 0.184 (0.130)

0.224 (0.158)

0.126 (0.121)

0.190 (0.139)

Genocide 0.246 (0.220)

0.144 (0.220)

0.326 (0.216)

0.183 (0.217)

Past intervention -0.013* (0.005)

-0.012 (0.007)

-0.015** (0.006)

-0.011 (0.007)

Region dummies Yes Yes Yes Yes

DV: HRO Shaming Human rights abuses 0.049*

(0.023) 0.026 (0.023)

0.101* (0.048)

0.068 (0.048)

Media exposure 0.092** (0.024)

0.117** (0.028)

0.127** (0.047)

0.170** (0.053)

Human rights INGOs -0.002 (0.002)

-0.003 (0.002)

-0.006 (0.004)

-0.009* (0.004)

Democracy -0.005 (0.007)

-0.007 (0.008)

-0.006 (0.012)

-0.005 (0.012)

Population 0.050 (0.029)

0.029 (0.029)

0.108 (0.060)

0.113 (0.059)

Economic openness 0.076* (0.033)

0.060 (0.035)

0.092 (0.087)

0.097 (0.089)

Civil war 0.030 (0.092)

0.041 (0.101)

0.005 (0.171)

0.019 (0.182)

Region dummies Yes Yes Yes Yes Observation 1,933 1,452 1,933 1,452

Note: Robust standard errors adjusted for clustering over country appear in parentheses. *Significant at .05. **Significant at .01. Regional dummies and constants not shown to save space.

224 amanda murdie and dursun peksen

ordinary least squares (OLS) models that use the shaming variables as the outcome measures. On the right-hand side of the equation, we control for one- and two-year lagged values of the intervention variable and account for one- and two-year lagged values of the shaming variables. The p-values of these tests clearly indicate that shaming Granger-causes humanitarian intervention but intervention does not Granger-cause shaming.12

To further address the concerns about the causal process connecting HRO shaming to humanitarian intervention, Table 2 reports the results of seemingly unrelated regressions (SUR) models that predict the determinants of humanitarian intervention and HRO shaming with a common error term. We use the conditional mixed-process (CMP) recursive estima- tor (Roodman 2011), which is a SUR estimator that allows us to add different types of response variables (continuous for HRO shaming and dichotomous for humanitarian intervention) in different equations.13

In the first equation, we predict the covariates of HRO shaming. Following the earlier research on the determinants of HRO shaming (Hafner-Burton and Ron 2012; Hill, Moore, and Mukherjee 2013; Ron, Ramos, and Rodgers 2005), we control for the fol- lowing covariates of HRO shaming: human rights abuses, media exposure, the number of Human Rights INGOs with a membership base within the state, democracy, population (logged), economic openness (total trade as a percentage of GDP), civil war, and the region dummies. In the second equation, we estimate the determinants of humanitarian intervention using the shaming variables and the controls included in the logit models reported above. Consistent with the findings from the rare-event logit models, the results in Table 2 suggest that the number and intensity of HRO shaming activities increase the probability of humanitarian intervention.

We also explored whether HRO shaming serves as a mediator between human rights abuses and

humanitarian intervention. The results from the causal mediation analysis (Imai et al. 2011) indicate that HRO shaming serves as a small but robust mediator between human rights abuses and human- itarian intervention: mediation through HRO sham- ing accounts for only about 6% of the total effect between human rights abuses and humanitarian in- tervention. These results further illustrate how HRO shaming can inform a concerned public about exist- ing human rights abuses, leading to an increased probability of armed intervention. Although these results are consistent with part of our theoretical story, we see the role of HROs as greater than just a mediator between existing levels of human rights abuses and the decision for intervention. Like the well-known and theorized ‘‘independent’’ (Robinson 2011, 2) CNN effect, our causal story centers on the role of HROs specifically and their ability to provide information about the possibility for changing human rights conditions and the continuation of deplorable humanitarian situations.14

Conclusion

Human rights organizations produce information that can be especially informative about human rights vulnerabilities in many locations throughout the world. Organizations are largely seen as credible actors, and their information can be influential in foreign policy. Although these statements are often rhetorically agreed to by a host of foreign policy experts, many doubt the ability of human rights organizations to actually matter on realpolitik issues, like the foreign policy decision of humanitarian intervention (Drezner 2007; Hafner-Burton 2008).

In this study, we argue and show empirically that HROs do matter for foreign policy decision making concerning humanitarian military action. Somewhat akin to the supposed CNN effect, HROs produce information which is poignant and vivid, starting a process of mobilization and opinion change which ultimately can increase the likelihood of humanitar- ian intervention into a civilian atrocity. For human rights advocates supporting a ‘‘responsibility to pro- tect’’ doctrine in foreign policy, this research provides some evidence of the process through which nonstate actors can influence the likelihood of an international community response to save lives.

12A summary table of these Granger tests can be found in the online appendix. Also in the online appendix, we present the results of Wald chi-squared tests of exogeneity following in- strumental variable probit models; these results also indicate that the variable HRO shaming is properly exogenous.

13Note that the results from the Granger causality tests discussed in the preceding paragraph indicate that causality runs from shaming to intervention and not vice versa. Thus, the CMP is a perfect estimator for our analysis since it is designed to model the causal relationship between two or more variables in a re- cursive manner. That is, the CMP fits the equations with clearly defined stages: X can be a modeled determinant of Z in the first equation, and Z can be a determinant of Y in the second equation. But Y cannot be a modeled determinant of X or Z.

14For the detailed report of the results from the causal mediation analysis, please see the online appendix.

human rights shaming and interventions 225

Our findings add to the cumulative knowledge of the determinants of humanitarian interventions and also contribute to the general intervention literature, which has not empirically linked HROs and other nonstate actors to intervention decision making writ large. This work also adds to the news media and foreign policy literature by stressing how a very partic- ular news source (shaming by HROs) will lead to an increase in the likelihood of humanitarian intervention. The empirical support to this argument reiterates recent calls within the news media effect literature to focus on particular actors and information networks and their influence while simultaneously controlling for a host of political and economic factors making interventions more likely (Gilboa 2005; Robinson 2011).

This work highlights the process through which HROs work to influence human rights and stop mass atrocities. The canonical theories of Keck and Sikkink (1998) and Risse, Ropp, and Sikkink (1999) rest on the idea that HRO actions will cause international com- munity responses. We provide a fully fleshed argument as to how HROs would matter for intervention decision making and provide large scale empirical evidence of this process. This would be in line with other recent work on how shaming by HROs and IGOs can in- fluence other foreign policy decisions (Lebovic and Voeten 2009; Murdie and Peksen 2013).

Worth noting, however, is that there is a growing literature on the influence of humanitarian interven- tions on human rights and the ending of mass killings (DeMerritt 2012; Krain 2005; Peksen 2012). Much of this work provides limitations on when interventions will actually help human rights and generally finds that not all interventions are structured in ways that will lead to improvements in a country’s human rights perfor- mance. These findings may highlight the importance of HRO efforts to directly work to influence the nature of military interventions, like Amnesty International tries to do by encouraging specific human rights program- ming in existing interventions.

In short, HROs are conduits of powerful stories of ongoing civilian atrocities and the possibility of future atrocities, some of which the world commu- nity could try to prevent. Through their production of information on repressive regimes, these seemingly powerless actors actually influence the movements of troops across borders for humanitarian purposes.

Acknowledgments

Author names are listed in alphabetical order.

References

Africa Rights (AR) Programmes. 2012. http://web.peacelink.it/ afrights/program.htm

Amnesty International (AI). 2012. ‘‘Deadly Reprisals: Deliberate Killings and Other Abuses by Syria’s Armed Forces.’’ June 14. http://www.amnestyusa.org/sites/default/files/2012-06-14_syria_ report_final_compressed.pdf (September 11, 2013).

Amnesty International (AI). 2013. ‘‘Sudan: Indiscriminate Bombing Exacerbates Humanitarian Crisis in Souther Kordofan.’’ April 17. http://www.amnesty.org/en/news/sudan-indiscriminate-bombing- exacerbates-humanitarian-crisis-southern-kordofan-2013- 04-17 (September 11, 2013).

Barry, Colin M., K. Chad Clay, and Michael E. Flynn. 2013. ‘‘Avoiding the Spotlight: Human Rights Shaming and Foreign Direct Investment.’’ International Studies Quarterly. Forthcoming.

Baum, Matthew A., and Tim J. Groeling. 2009. War Stories: The Causes and Consequences of Public Views of War. Princeton, NJ: Princeton University Press.

Beardsley, Kyle, and Holger 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.

Beck, Nathaniel, and Jonathan Katz. 1995. ‘‘What to Do (and Not to Do) with Time-Series Cross-Section Data.’’ American Political Science Review 89 (3): 634–48.

Bell, Sam K., Chad Clay, and Amanda Murdie. 2012. ‘‘Neighborhood Watch: Spatial Effects of Human Rights INGOs.’’ Journal of Politics 74 (2): 354–68.

Bob, Clifford. 2005. The Marketing of Rebellion: Insurgents, Media, and International Activism. Cambridge: Cambridge University Press

Boli, John, and George M. Thomas. 1999. ‘‘INGOs and the Organization of World Culture.’’ In Constructing World Culture: International Non-governmental Organizations Since 1875, eds. John Boli and George M. Thomas. Stanford, CA: Stanford University Press, 13–49.

Bond, Doug, Joe Bond, Churl Oh, J. Craig Jenkins, and Charles Lewis Taylor. 2003. ‘‘Integrated Data for Events Analysis (IDEA): An Event Typology for Automated Events Data Development.’’ Journal of Peace Research 40 (6): 733–45.

Choi, Seung-Whan. 2012. ‘‘What determines U.S. Humanitarian Intervention?’’ Conflict Management and Peace Science 30 (2): 121–39.

Collier, Paul, and Anke Hoeffler. 1998. ‘‘On Economic Causes of Civil Wars.’’ Oxford Economic Papers 50 (4): 563–73.

CR-1991-0111. United States Congressional Record. January 11, 1991. 102(1). http://congressional.proquest.com/congressional/ docview/t19.d20.cr-1991-0111?accountid514576 (September 11, 2013).

Davis, David R., Amanda Murdie, and Coty Garnett Steinmetz. 2012. ‘‘‘Makers and Shapers’: Human Rights INGOs and Public Opinion.’’ Human Rights Quarterly 34 (1): 199–224.

DeMeritt, Jacqueline H. R. 2012. ‘‘Delegating Death: Military Intervention and Government Killing.’’ Journal of Conflict Resolution. Forthcoming.

Drezner, Daniel W. 2007. All Politics is Global. Princeton, NJ: Princeton University Press.

Entman, Robert M. 2004. Projections of Power: Framing News, Public Opinion, and U.S. Foreign Policy. Chicago: University of Chicago Press.

226 amanda murdie and dursun peksen

Fearon, James D., and David D. Laitin. 2003. ‘‘Ethnicity, Insurgency and Civil War.’’ American Political Science Review 97 (1): 75–90.

Fitzduff, Mari, and Cheyanne Church. 2004. NGO at the Table: Strategies for Influencing Policies in Areas of Conflict. Lanham, MD: Rowman & Littlefield.

Fordham, Benjamin O. 2008. ‘‘Power or Plenty? Economic Interests, Security Concerns, and American Intervention.’’ International Studies Quarterly 52 (4): 737–58.

Forsythe, David P. 1988. Human Rights and US Foreign Policy: Congress Reconsidered. Gainesville: University Press of Florida.

‘‘Further Written Evidence from Amnesty International UK.’’ 2011. United Kingdom Parliament. Foreign Affairs Committee. December 18. http://www.publications.parliament. uk/pa/cm201213/cmselect/cmfaff/80/80awe19.htm (September 11, 2013).

Gartner, Scott Sigmund, and Gary M. Segura. 1998. ‘‘War, Casualties and Public Opinion.’’ Journal of Conflict Resolution 42 (3): 278–300.

Gartner, Scott Sigmund, Gary M. Segura, and Bethany A. Barratt. 2004. ‘‘War Casualties, Policy Positions, and the Fate of Legislators.’’ Political Research Quarterly 57 (3): 467–77.

Geeraerts, Gustaaf. 1995. ‘‘Analyzing Non-State Actors in World Politics.’’ Pole Paper Series 1 (4). http://poli.vub.ac.be/publi/ pole-papers/pole0104.htm (September 11, 2013).

Gibney, Mark, Linda Cornett, and Reed Wood. 2010. Political Terror Scale 1976–2008. Asheville: University of North Carolina Press.

Gilboa, Eytan. 2005. ‘‘Global Television News and Foreign Policy: Debating the CNN Effect.’’ International Studies Perspective 6 (3): 325–41.

Gleditsch, Nils Petter, Peter Wallensteen, Mikael Eriksson, Margareta Sollenberg, and Håvard Strand. 2002. ‘‘Armed Conflict 1946–2001: A New Dataset.’’ Journal of Peace Research 39 (5): 615–37.

Goldstein, Joshua S. 1992. ‘‘A Conflict-Cooperation Scale for WEIS Events Data.’’ Journal of Conflict Resolution 36 (2): 369–85.

Gourevitch, Peter A., and David A. Lake. 2012. The Credibility of Transnational NGOs: When Virtue Is Not Enough. Cambridge: Cambridge University Press.

Granger, Clive W. J. 1969. ‘‘Investigating Causal Relations by Econometric Models and Cross-Spectral Methods.’’ Econometrica: Journal of the Econometric Society 37 (3): 424–38.

Groeling, Tim J., and Matthew A. Baum. 2009. ‘‘Journalists’ Incentives and Media Coverage of Elite Foreign Policy Evaluations.’’ Conflict Management and Peace Science 26 (5): 437–70.

Hafner-Burton, Emilie M. 2008. ‘‘Sticks and Stones: Naming and Shaming the Human Rights Enforcement Problem.’’ International Organization 62 (4): 689–716.

Hafner-Burton, Emilie, and James Ron. 2012. ‘‘The Latin Bias: Regions, the Anglo-American Media, and Human Rights.’’ International Studies Quarterly. doi:10.1111/isqu.12023.

Hill, Daniel W., Will H. Moore, and Bumba Mukherjee. 2013. ‘‘Information Politics versus Organizational Incentives: When Are Amnesty International’s ‘Naming and Shaming’ Reports Biased?’’ International Studies Quarterly 57 (2): 219–32.

Holzgrefe, J. L. 2003. ‘‘The Humanitarian Intervention Debate.’’ In Humanitarian Intervention: Ethical, Legal, and Political Dilemmas, eds. J. L. Holzgrefe and R. O. Keohane. Cambridge: Cambridge University Press, 15–52.

Horowitz, Donald. 1985. Ethnic groups in conflict. Berkeley: University of California Press.

Hufbauer, Gary, Jeffrey Schott, Kimberly A. Elliott, and Barbara Oegg. 2007. Economic Sanctions Reconsidered: History and Current Policy. 3rd ed. Washington, DC: Patterson Institute for International Economics.

Imai, Kosuke, Luke Keele, Dustin Tingley, and Teppei Yamamoto. 2011. ‘‘Unpacking the Black Box of Causality: Learning About Causal Mechanisms from Experimental and Observational Studies.’’ American Political Science Review 105 (4): 765–89.

Jakobsen, Peter Vigo. 2000. ‘‘Focus on the CNN Effect Misses the Point: the Real Media Impact on Conflict Management is Invisible and Indirect.’’ Journal of Peace Research 37 (2): 131–43.

James, Patrick, and John R. O’Neal. 1991. ‘‘The Influence of Domestic and International Politics on the President’s Use of Force.’’ Journal of Conflict Resolution 35 (2): 307–32.

Kaldor, Mary. 2001. ‘‘A Decade of Humanitarian Intervention: The Role of Global Civil Society.’’ Global Civil Society 2001. London: LSE Global Governance.

Keck, Margaret E., and Kathryn Sikkink. 1998. Activists beyond Borders: Advocacy Networks in International Politics. Ithaca, NY: Cornell University Press.

King, Gary, and Langche Zeng. 2001. ‘‘Logistic Regression in Rare Events Data.’’ Political Analysis 9 (2): 137–63.

Krain, Matthew. 2005. ‘‘International Intervention and the Severity of Genocides and Politicides.’’ International Studies Quarterly 49 (3): 363–87.

Lebovic, James H., and Erik Voeten. 2009. ‘‘The Cost of Shame: International Organizations and Foreign Aid in the Punishing of Human Rights Violators.’’ Journal of Peace Research 46 (1): 79–97.

Marshall, Monty G., Ted Gurr, and Barbara Harff. 2012. ‘‘PITF – State Failure Problem Set: Internal Wars and Failures of Governance, 1955–2011.’’ http://www.systemicpeace.org/inscr/ PITFProbSetCodebook2011.pdf (September 11, 2013).

Marshall, Monty G., and Keith Jaggers. 2000. ‘‘Polity IV Project: Political Regime Characteristics and Transitions, 1800–2003.’’ http://www.systemicpeace.org/polity/polity4.htm (September 11, 2013).

Mueller, John E. 1973. War, Presidents and Public Opinion. New York: John Wiley.

Mullenbach, Mark J. 2005. ‘‘Deciding to Keep Peace: An Analysis of International Influences on the Establishment of Third- Party Peacekeeping Missions.’’ International Studies Quarterly 49 (3): 529–55.

Mullenbach, Mark J., and Gerald P. Matthews. 2008. ‘‘Deciding to Intervene: An Analysis of International and Domestic Influences on United States Interventions in Intrastate Disputes.’’ International Interactions 34 (1): 25–52.

Murdie, Amanda, and David R. Davis. 2012. ‘‘Shaming and Blaming: Using Events Data to Assess the Impact of Human Rights INGOs.’’ International Studies Quarterly 56 (1): 1–16.

Murdie, Amanda, and Dursun Peksen. 2013. ‘‘The Impact of Human Rights INGO Activities on Economic Sanctions.’’ Review of International Organizations 8 (1): 33–53.

‘‘Peacekeeping Operations in Kosovo Resolution.’’ 1999. United States. 145 Cong Rec H 1179. House. March 11. 106–1. http://congressional.proquest.com/congressional/docview/ t17.d18.c49965b41800186b?accountid514576 (September 11, 2013).

Pearson, Frederic S., and Robert A. Baumann. 1993. ‘‘International Military Intervention, 1946–1988.’’ Ann Arbor, MI: ICPSR [No. 6035].

human rights shaming and interventions 227

Peksen, Dursun. 2012. ‘‘Does Foreign Military Intervention Help Human Rights?’’ Political Research Quarterly 65 (3): 558–71.

Pickering, Jeffrey. 2002. ‘‘Give Me Shelter: Re-Examining Military Intervention and the Monadic Democratic Peace.’’ Interna- tional Interactions 28 (4): 293–324.

Pickering, Jeffrey, and Emizet N. F. Kisangani. 2009. ‘‘The International Military Dataset: An Updated Resource for Conflict Scholars.’’ Journal of Peace Research 46 (4): 589–99.

PRI. 2012. ‘‘Amnesty International Warns of Worsening Situation in Syria.’’ July 15. http://www.pri.org/stories/world/middle-east/ tt-amnesty-international-warns-of-worsening-situation-in-syria- 10657.html (September 11, 2013).

Ramos, Howard, James Ron, and Oskar N. T. Thoms. 2007. ‘‘Shaping the Northern Media’s Human Rights Coverage, 1986–2000.’’ Journal of Peace Research 44 (4): 385–406.

Regan, Patrick M. 2000. Civil Wars and Foreign Powers: Outside Intervention in Intrastate Conflict. Ann Arbor: University of Michigan Press.

Risse, Thomas, Stephen C. Ropp, and Kathryn Sikkink, eds. 1999. The Power of Human Rights: International Norms and Domes- tic Change. Cambridge: Cambridge University Press.

Roodman, David. 2011. ‘‘Fitting Fully Observed Recursive Mixed-Process Models with CMP.’’ Stata Journal 11 (2): 159–206.

Robinson, Piers. 2000. ‘‘The Policy-Media Interaction Model: Measuring Media Power during Humanitarian Crisis.’’ Journal of Peace Research 37 (5): 613–33.

Robinson, Piers. 2011. ‘‘The CNN Effect Reconsidered: Mapping a Research Agenda for the Future.’’ Media, War, & Conflict 4 (1): 3–11.

Ron, James, Howard Ramos, and Kathleen Rodgers. 2005. ‘‘Transnational Information Politics: NGO Human Rights Reporting, 1986–2000.’’ International Studies Quarterly 49 (3): 557–87.

Rost, Nicholas, and J. Michael Greig. 2011. ‘‘Taking Matters into their Own Hands: An Analysis of the Determinants of State-Conducted Peacekeeping in Civil Wars.’’ Journal of Peace Research 48 (2): 171–84.

Roth, Kenneth. 2011. ‘‘The Security Council Has At Last Lived Up to Its Duty.’’ Foreign Policy. March 19. http://www.hrw. org/news/2011/03/18/security-council-has-last-lived-its-duty (September 11, 2013).

Sané, Pierre. 2000. ‘‘Soldiers in the Name of Human Rights.’’ Spokesman Journal 69: 82–90.

Shavin, Gidon. 2012. ‘‘Amnesty International’s Credibility Problem over Israel.’’ Commentator. June 6. http://www. thecommentator.com/article/1279/amnesty_international_s_ credibility_problem_over_israel (September 11, 2013).

Singer, David J. 1988. ‘‘Reconstructing the Correlates of War Dataset on Material Capabilities of States, 1816–1985.’’ International Interactions 14 (2): 115–32.

Snyder, Richard C., H. W. Bruck, Burton Sapin, and Valerie Hudson. 2003. Foreign Policy Decision Making (Revisited). Basingstoke, UK: Palgrave Macmillian.

Sudan Tribune. 2012. ‘‘Darfur NGOs Urge Humanitarian Intervention amid Upsurge of Violence.’’ August 9. http:// www.sudantribune.com/spip.php?article43526 (September 11, 2013).

Waltz, Kenneth H. 1979. Theory in International Politics. New York: McGraw-Hill.

Wong, Wendy H. 2012. Internal Affairs: How the Structure of NGOs Transforms Human Rights. Ithaca, NY: Cornell University Press.

World Values Survey. 1981–2008. Official Aggregate V.20090901, 2009. World Values Survey Association. www.Worldvaluessurvey. org (September 11, 2013).

Amanda Murdie is an Assistant Professor in the Department of Political Science at the University of Missouri, Columbia, MO 65211.

Dursun Peksen is an Assistant Professor in the Department of Political Science at the University of Memphis, Memphis, TN 38152.

228 amanda murdie and dursun peksen