Why Governments Cede Sovereignty: Evidence From Regional Human Rights Courts

Why Governments Cede Sovereignty: Evidence From Regional Human Rights Courts Abstract This study examines a government’s decision to cede authority over fundamental questions of policy to international organizations. Explanations for delegation to international institutions focus on the benefits of “locking in” domestic political institutions and signaling intentions to the public. This line of reasoning suggests that partly democratic and democratizing governments have the strongest incentive to participate in international human rights institutions. I evaluate this claim by analyzing patterns of participation in two regional human rights courts: the European Court of Human Rights and the Inter-American Court of Human Rights. The decision to join a regional court is a more valid indicator of delegation than those used in previous empirical studies and to date has not been examined in systematic fashion. I find strong evidence in support of the claim that governments grant authority to institutions to credibly commit to maintaining democracy. Introduction Why would a government cede authority over fundamental questions of policy to an international institution? This is a puzzling question, and in the realm of human rights, it becomes more puzzling. In contrast to most international agreements, human rights agreements attempt to constrain states’ behavior toward their own citizens rather than toward one another and so by definition represent a greater encroachment on state sovereignty. In addition, compared with other policy areas, the efficiency gains from creating these agreements are not obvious: governments who wish to observe the human rights of their citizens do not gain anything additional by jointly observing human rights (Koremenos 2007; Simmons 2009). That said, participation in many human rights institutions does not entail large sovereignty costs. Most analyses of the decision to join these institutions focus on UN-based human rights treaties for which monitoring and enforcement provisions are notoriously weak (Cole 2005; Goodliffe and Hawkins 2006; Hathaway 2007; Cole 2009; Powell and Staton 2009; Simmons 2009). In terms familiar to scholars of international cooperation, these agreements generally entail high levels of obligation but very little delegation (Abbott et al. 2000). The most severe form of sovereignty loss is delegation to an international human rights court (Bradley and Kelley 2008). This requires not only that a government agree to follow a set of externally created rules that limit their domestic political authority but also that they authorize a body they have limited control over to issue public judgments about whether they are complying with those rules. That is, submitting to the authority of an international court entails a higher level of delegation than treaty commitments and, in the area of human rights, delegation necessarily encroaches on sovereignty (Abbott and Snidal 2000; Hathway 2008). Explanations for international delegation focus on the benefits of constraining government behavior in the long term. This is a familiar reason for delegation, which is invoked, for example, in the literatures on property rights and economic growth (North and Weingast 1989) and delegation of monetary policy to central banks (Rogoff 1985; Lohmann 1992; Bernhard, Broz, and Clark 2002). International human rights courts in particular have been theorized as commitment devices for newly democratic governments who wish to signal a long-term commitment to democracy. Moravcsik (2000) argues that states create these courts in order to “lock in” good practices and constrain the behavior of potential successors. Here, the government is worried about antidemocratic opponents coming to power and subverting fledgling democratic institutions. Mansfield and Pevehouse (2006) build on this argument, claiming that democratizing states grant authority to international organizations (IOs) in order to raise the costs of reneging on reforms and credibly signal, to domestic and international audiences, their intent to remain democratic. Both these explanations view delegation as a mechanism for “self-binding” (Alter 2008), meaning the intent is to constrain government behavior, and both focus on the capacity of institutions to make undemocratic behavior costly by monitoring behavior and providing information about compliance. This line of reasoning has two implications: (1) delegation to human rights regimes should be attractive to newly democratized governments and (2) delegation to human rights regimes should be attractive to governments with weakly democratic institutions. Moravcsik (2000), Mansfield and Pevehouse (2006, 2008), and Hafner-Burton, Pevehouse, and Mansfield (2015) test these hypotheses, but the design of these studies prevents firm conclusions. Most importantly, all fail to create a good match between the concept of delegation and its operationalization. Specifically, they examine policy decisions that do not necessarily entail granting an institution real authority to monitor and publicize antidemocratic behavior. I elaborate this point below and provide a new analysis to evaluate the lock-in/signaling explanation for delegation to international institutions. I do this by examining delegation to human rights courts in the Council of Europe (hereafter the CE) and the Organization of American States (hereafter the OAS). Beginning in 1950, CE members had the option of accepting the jurisdiction of the European Court of Human Rights (hereafter the European Court) and have had no choice about this matter since 1998. The European Court is authorized to make legally binding, public decisions regarding compliance with the European Convention on Human Rights (hereafter the European Convention). Since 1969 members of the OAS have had the option of accepting the jurisdiction of the Inter-American Court of Human Rights (hereafter the Inter-American Court), which is authorized to make judgments regarding compliance with the American Convention on Human Rights (hereafter the American Convention). Accepting the jurisdiction of these courts clearly involves delegation to a relatively strong institution tasked explicitly with monitoring and publicizing human rights abuses. This means that acceptance of a regional court’s jurisdiction is a strong indicator of delegation to an institution that promotes human rights and democracy. Also, Europe and the Western hemisphere have both experienced a large number of democratic transitions since the creation of their respective human rights courts. This variation allows me to examine the reasons for acceptance of each court’s jurisdiction in a systematic fashion over a relatively long time period. Although delegation to international institutions has been the subject of much academic writing (see, for example, Hawkins et al. 2006), systematic empirical research in this area is thin (Bradley and Kelley 2008). I contribute to this literature by examining the link between democracy and commitments to international human rights institutions with varying levels of legalization. I find strong evidence consistent with the claim that uncertainty about the future of democracy creates incentives to join international human rights courts. I contrast these results with an analysis of ratification of the International Covenant on Civil and Political Rights (hereafter the ICCPR) between CE and OAS members. The ICCPR is a universal treaty drafted by the United Nation which is nearly identical to the European and American Conventions in terms of content but which does not delegate monitoring/enforcement power to the same extent as either of these treaties. I demonstrate that patterns of participation are different for human rights courts and treaties, which suggest that the design of international institutions affects how useful they are as commitment and signaling devices and is something that must be accounted for in empirical analysis. I follow the analysis with a discussion of the implications of these findings for future research. Why Delegate to International Human Rights Institutions? As mentioned above, most empirical research on human rights law examines UN treaties, which create legal obligations but do not rank as highly on the delegation dimension of legalization as human rights courts. UN treaties do create committees that serve as monitoring bodies, but these committees are limited in their information gathering capacity and, more importantly, do not have any legal authority to issue public, binding decisions concerning alleged treaty violations. This institutional feature of the treaty regime leads to participation that may not reflect a genuine commitment to protecting human rights. Research suggests that ratification is often not a sincere act; many studies find a weak (or even negative) unconditional relationship between ratification and respect for human rights (Hathaway 2002; Neumayer 2005; Hill 2010).1 States with questionable practices may ratify to receive an “expressive benefit” (Hathaway 2002), which refers to the approbation governments sometimes receive from communicating to other states that they accept the prevalent norms of the international community (Risse, Ropp, and Sikkink 1999). Studies informed by the global institutionalist school (Meyer et al. 1997) similarly argue that this “decoupling” results from a combination of (global) cultural pressure to express support for human rights norms and weak international monitoring and enforcement (for example, Cole 2005; Hafner-Burton and Tsutsui 2005). In line with these arguments, there is evidence that institutions with stricter monitoring procedures mitigate the decoupling of rules and practice. For example, Cole (2012) finds that the adoption of optional protocols, which grant more monitoring power to human rights committees than the treaties themselves, is actually associated with better human rights practices. This suggests that variation in the authority of delegated bodies across human rights regimes is related to patterns of participation in predictable ways, a point that I elaborate below. Others focus on a lack of domestic enforcement to explain the disconnect between treaty ratification and human rights behavior. Hathaway (2007) and Powell and Staton (2009) find that governments with poor practices but weak domestic courts are less hesitant to ratify, whereas those with poor practices and strong courts are more cautious. Since states that face a low likelihood of domestic enforcement are more likely to join, treaties have limited impact on states that ratify. Perhaps it is unsurprising that willful evasion of international and domestic enforcement mechanisms is common among states that ratify human rights treaties. After all, human rights agreements encroach on sovereignty to a greater extent than other kinds of agreements. However, evading meaningful legal constraints is not as feasible where human rights courts are concerned. Membership in a court involves a greater degree of delegation than membership in a treaty; submitting to a court’s jurisdiction creates the possibility of public, legally binding judgments that the government is violating its human rights obligations, which is not true of a treaty. Of course, governments can always refuse to comply with court rulings just as they can refuse to comply with treaty provisions. But noncompliance with a court ruling entails publicly defying a legal body whose authority the government has formally recognized. This is not to say that states cannot circumscribe the authority of international courts. Scholars of judicial politics have extensively cataloged the strategies governments may adopt to limit the autonomy of domestic courts, such as screening/appointing judges based on ideology and writing precise legislation that limits judicial interpretation. But these strategies are not always effective in the domestic context and are more difficult to employ in the international context due to the relative difficulty of changing international law and more limited control over appointments (Alter 2006). Additionally, judges on human rights courts can still interpret rules in unanticipated ways and sometimes do so in ways that increase their autonomy. For example, Hawkins and Jacoby (2006) document how the European Court, without a directive from member states, used procedural rules to enhance individual access to the court. These institutional differences between treaties and courts should produce different patterns of participation. Since courts create a greater likelihood of human rights abuses being exposed and discussed publicly, states should be cautious about making insincere commitments that lead to a decoupling of legal rules and practices. In line with this observation, explanations for participation in courts tend to focus on the conditions that make it attractive to submit to meaningful legal constraints. The argument made by Moravcsik (2000) appeals to an uncertain political environment to explain delegation to human rights regimes. Governments in recently democratized countries have good reason to believe that their successors will not share their preferences for democratization and, to reduce the risk of a reversal, may support the creation of a human rights courts in order to “lock in” liberal institutions. A human rights court presents these governments with an opportunity to “establish reliable judicial constraints” on successors by “placing interpretation in the hands of independent authorities managed in part by foreign governments” (Moravcsik 2000, 228). For these governments, the benefits of biasing policy toward their ideal point outweigh the sovereignty costs incurred by delegating to an international court. The benefits are too small for consolidated democracies, who have already locked in liberal institutions, and autocrats do not prefer to create or lock-in liberal institutions. The implication of this argument is that stable democracies and stable dictatorships will have the least incentive to join, producing an “inverse-U-shaped” relationship between the strength of democratic institutions and a government’s willingness to participate in a human rights court (Moravcsik 2000, 232). Mansfield and Pevehouse (2006, 2008) also appeal to the uncertainty present in transitioning regimes to explain why governments join international institutions. They make a valuable theoretical contribution by pointing out that IO membership helps to tie the hands of the current (and perhaps future) government and credibly signals to interested parties (the domestic public, international investors, etc.) the current government’s willingness to preserve democracy in the long term. This calls attention to a related incentive that leaders have for delegating to international institutions: by creating a cost for violating the rules, joining an institution can signal a willingness to follow the rules. Mansfield and Pevehouse (2006, 2008) present their argument as an explanation for IO membership generally, but the mechanism they seem to have in mind involves delegation (as defined by Abbott et al. 2000) to institutions that have an explicit mission to promote democracy and human rights. They claim that joining an IO credibly signals a commitment to democracy because “monitoring and enforcement are handled by a third party with the ability to publicly sound an alarm in the event that reform efforts falter” (Mansfield and Pevehouse, 2006, 141). They hypothesize that democratizing regimes have the strongest incentives to grant authority to IOs. To summarize, previous work examining why states incur sovereignty costs argues that political uncertainty resulting from democratizing/weak political institutions is the primary motivation for delegation to international human rights institutions. The implications of this argument are that transitions to democracy, as well as weakly democratic institutions, make delegation to human rights institutions more likely. Yet it is unclear whether the empirical record supports these hypotheses because these claims have been inadequately evaluated in previous work, a point that I elaborate below. Existing Evidence for the Lock-In/Signaling Theory The object of explanation for this study is the decision to accept the jurisdiction of a human rights court. Moravcsik (2000) examines preliminary negotiations about the features of the European Court and demonstrates that the earliest supporters of granting the court the powers of individual petition2 and compulsory jurisdiction tended to be newly democratic states. Although the evidence seems to support his argument, it is based on a relatively small number of cases. It is also limited to a short time frame rather than subsequent patterns of participation that unfolded over a longer period of time. Most importantly, this study does not examine actual delegation but rather support for the creation of a regime to which one could later delegate. Examining acceptance of the European Court’s jurisdiction would be a better, more direct test of the empirical implications of this argument. Explanations for delegation are best tested by examining actual participation in institutions rather than support for strong institutions in principle. It is not clear why supporting the creation of a court would raise the cost of reversing democracy if a state had not accepted its jurisdiction, though to be fair, Moravcsik examines support for compulsory jurisdiction, which would entail acceptance for all countries if adopted. But compulsory jurisdiction was not adopted when the European Convention was opened for ratification in 1950, which meant CE governments could (until 1998) ratify the European Convention without accepting the court’s jurisdiction. And in fact two of the seven governments that Moravcsik lists as initially supportive of compulsory jurisdiction, Italy and France, waited until 1973 and 1974, respectively, to accept jurisdiction. The United Kingdom accepted jurisdiction several years earlier (1966), even though its initial recalcitrance during negotiations about granting the court extended powers is well documented and discussed by Moravcsik. Additionally, Denmark, Belgium, the Netherlands, and Luxembourg, four of the seven established democracies listed by Moravcsik, were among the first to accept jurisdiction (1953, 1955, 1954, and 1958, respectively). Thus, the evidence on this count is not entirely convincing,3 and examining the available evidence more systematically is difficult, given the institutional changes that occurred in the CE beginning in 1994. In 1994, Protocol 11 to the European Convention, which would create compulsory jurisdiction once ratified by the requisite number of states, became open for ratification. This means that the largest wave of democratic transitions to occur in Europe since the court’s creation occurred at roughly the same time that the CE began moving toward compulsory jurisdiction, which makes it difficult to separate the effects of democratization from institutional changes in the CE itself. Although Moravcsik (2000) operationalizes delegation too narrowly, Mansfield and Pevehouse (2006, 2008) do so too broadly and measure IO membership generally rather than delegation to institutions devoted to democracy and human rights. These studies show that democratizing states join, on average, greater numbers of IOs than nondemocratizing states, but the indicators of IO membership used in these studies are not ideal for measuring the kind of institutional commitment needed to make a credible commitment to democracy. The lock-in/signaling argument focuses on IOs’ ability to monitor and publicize state behavior that violates the organization’s standards. This suggests that examining commitments to highly legalized human rights institutions would be more appropriate than examining IO membership generally. Delegation to a human rights court requires a primary legal obligation and submission to public, legally binding, judicial decisions. Publicly pledging to follow the rules of an institution devoted explicitly to many of the rights associated with democracy, and granting a court the authority to publicly declare behavior noncompliant, creates a much clearer cost for violating those rights than does membership in various IOs. Membership in IOs that have little to do with human rights or democracy does not obviously create such a cost, but the indicators used in these studies mostly measure institutions that have little to do with these issues, which means that the measure is very noisy. Further, summing the number of IOs that a state belongs to, as Mansfield and Pevehouse (2006, 2008) do, assumes that each of these IOs has an equivalent effect on the cost of undemocratic behavior. This is a strong assumption given the diversity of IOs included in their data, which include such organizations as the International Office of Epizootics and the Inter-State School of Hydraulic and Rural Engineering for Senior Technicians (Pevehouse, Nordstrom, and Warnke 2004). In a more recent study, Hafner-Burton, Pevehouse, and Mansfield (2015) offer an improved test by examining human rights-related organizations (defined as organizations whose stated goals include any type of human rights program) and developing a scale to measure the level of legalization created by a particular organization. But this measure also groups together IOs that are very heterogeneous in terms of both the level of delegation that they require (many involve no monitoring or oversight whatsoever) and the extent to which their goals are explicitly human rights-related. For example, the Organization for Economic Cooperation and Development (OECD) and North Atlantic Treaty Organization (NATO) are counted as “human rights IOs” alongside the European and Inter-American Courts. This indicator also assumes a priori a weighting scheme for the various institutional features of these organizations, which are all assumed to contribute equally to an institution’s level of legalization. This means that this measure is also noisy.4 Examining participation in regional human rights courts provides a simple but strong empirical test of explanations for delegation. This binary indicator is less noisy, and more valid, than the measures of IO membership employed by Mansfield and Pevehouse (2006, 2008) and Hafner-Burton, Pevehouse, and Mansfield (2015) or the negotiation positions examined by Moravcsik (2000) which, as noted above, do not correlate perfectly with acceptance of the European Court’s jurisdiction. This means that the inferences drawn from the analysis below are stronger than in previous studies and provide better evidence concerning the veracity of the lock-in/signaling argument. To demonstrate that this argument is best treated as an explanation for delegation to human rights regimes, I contrast participation in human rights courts with ratification of the ICCPR. The ICCPR is nearly identical in content to the European and American Conventions but does not create a court to monitor and enforce its provisions. The ICCPR creates the UN Human Rights Committee (hereafter HRC), which has some limited information gathering capacity, but its degree of legal authority is less than that of either the European Court or the Inter-American Court.5 Granting legal authority to the European Court or Inter-American Court clearly entails strong delegation to an institution devoted explicitly to human rights. Ratifying the ICCPR entails a nearly identical legal obligation, but little delegation. Thus, any differences in participation can be attributed to different levels of legalization across these institutions. Examining Evidence from Regional Human Rights Courts The CE was created in 1949 and today is comprised of forty-seven member states. Between 1950 and 1998, its members had the option of accepting the jurisdiction of the European Court, which after 1998, was no longer optional. There is a fair amount of variation in democracy among CE countries during this time period, due mostly to the democratic transitions in Southern Europe during the 1970s and Eastern Europe in the 1990s. As noted above, many of the transitions in the latter group occurred during the midst of the CE’s formal push toward compulsory jurisdiction. The OAS was formed in 1948, and its membership today consists of thirty-five countries. OAS members have had the option of accepting the jurisdiction of the Inter-American Court since 1969. Unlike the European Court, the jurisdiction of the Inter-American Court is not compulsory, which means that there is still some variation in delegation among OAS members. Twenty-one of the OAS’s thirty-five member states have accepted the court’s jurisdiction to date, with the last acceptance occurring in 2000 by Barbados. This, together with the fact that domestic political institutions across the Western hemisphere have fluctuated a great deal since 1969, makes the OAS an arguably better organization in which to test explanations for delegation. The key explanatory concept in the theory discussed above is democracy. Past work has examined two implications of this theory: (1) increases in/transitions to democracy make delegation to international human rights institutions more likely and (2) the relationship between democracy and delegation should be inverse-U-shaped. Testing these implications requires a measure of democracy that allows valid distinctions among autocracies, democracies, and weakly democratic or democratizing regimes. Most studies of the adoption of international human rights law treat democracy as an important explanatory concept, and many of these studies use the Polity scale or one of its components (Hathaway 2007; Simmons 2009; Simmons and Danner 2010; Chapman and Chaudoin 2012) as do the studies by Mansfield and Pevehouse (2006, 2008) and Hafner-Burton, Pevehouse, and Mansfield (2015). The Polity scale may be a reasonable indicator of democracy, but it is not ideal for testing these hypotheses. Analyses of the Polity data have uncovered problems with the scale’s construction and aggregation (Gleditsch and Ward 1997; Treier and Jackman 2008). Most importantly for this study, there is noise in the scale such that distinctions between states toward the high/low ends of the scale and those in the middle may not be valid. That is, in some cases, the scale does a poor job distinguishing between countries that have middling levels of democracy and those that have low/high levels. Thus, the measurement error that is known to exist in Polity makes it a poor choice for testing hypotheses about the behavior of weakly democratic states as this measurement error would call into question any inferences drawn about acceptance rates of weakly democratic states relative to states closer to the ends of the scale.6 I adopt instead the measure of democracy created by Pemstein, Meserve, and Melton (2010). This measure is derived from ten different commonly employed indicators of democracy. The authors employ a Bayesian version of factor analysis that allows for continuous and ordinal indicators (see Quinn 2004). The resulting measure is continuous and thus allows for distinctions between gradations of democracy. Perhaps the biggest advantage of using this indicator is that their model produces estimates of uncertainty (measurement error) around each country-year’s value of democracy that can be incorporated into subsequent analysis. In all the models below, I also include a measure of domestic legal system heritage. I use data from Powell and Mitchell (2007), which indicates whether a country’s legal system has primarily features of a common rather than a civil, law system. Recent work suggests that domestic legal systems are strong determinants of international legal commitments (Powell and Mitchell 2007; Mitchell and Powell 2009; Simmons 2009; Mitchell and Powell 2011). Accounting for legal system type is especially important in the OAS since many of the most consistently democratic countries in the OAS including the United States, Canada, the Bahamas, Jamaica, Trinidad, Barbados, and Belize were once British colonies and inherited common law legal systems. This means that common law heritage is likely correlated with democracy, as well as acceptance of the Inter-American Court’s jurisdiction, so it is important to account for this in the analysis. I also include a measure of civil war since democratizing and weakly democratic states frequently experience political violence (see, for example, Muller 1985; Hegre et al. 2001), and one study has found that states that have joined the International Criminal Court experienced lower levels of internal violence (prior to joining) than those that have not yet joined (Chapman and Chaudoin 2012). For this purpose, I employ the Uppsala Conflict Data Program/Peace Research Institute, Oslo (UCDP/PRIO) measure of violent conflict, which uses a threshold of twenty-five annual battle deaths to indicate the presence of civil war (Gleditsch et al. 2002; Themnér and Wallensteen 2012). Models and Estimation Below I present estimates from models using data on member countries from the CE and the OAS. For analysis of the European Court, the data include forty-one member states for the years 1950–1998. The data for the Inter-American Court models include twenty-nine member states for the years 1969–2004. In both populations, not all states were sovereign/independent for the entire time period, so some states enter the data later than 1950/1969. Once a state accepts the relevant court’s jurisdiction, it drops out of the population. The full population for the CE data includes 773 country-years, while the OAS population is comprised of 686 country-years. For the ICCPR models, the data begin when the treaty became open for ratification (1966). Once a state ratifies, it drops out of the population. The OAS data for the ICCPR analysis consist of twenty-eight member states for a total of 442 country-years, and the CE data consist of thirty member states and 464 country-years. The smaller population sizes for the ICCPR analyses are due primarily to the fact that most CE/OAS states ratified the ICCPR more quickly than they joined their respective human rights court. I employ Bayesian, hierarchical logistic regression models to analyze acceptance of a regional human rights court’s jurisdiction/ratification of the ICCPR. Each model controls for the effects of civil conflict and common law legal system, though I limit the discussion below to the results regarding democracy. Since my goal is to estimate the effect of democracy and democratic transitions on delegation to human rights courts, it is only necessary to control for factors that are correlated with both democracy and delegation. Limiting the covariates to those that have been theoretically linked to democracy and delegation also helps avoid a proliferation of highly correlated control variables, which can complicate analysis (see Achen 2002; Schrodt 2014). The models include random effects for country and a time counter7 and so are effectively discrete time versions of duration models with random effects (frailties). With the inclusion of time-varying covariates, estimation of parametric, continuous time duration models or the nonparametric Cox model becomes complicated. This is especially true in a Bayesian framework which I adopt here. These models also make the estimation of country-specific effects (frailties) more difficult. Thus, I employ the binary time-series cross-sectional design commonly used in international relations, which accounts for temporal dependence by simply including a time counter in a binary response (logit or probit) regression model (see Beck, Katz, and Tucker 1998; Carter and Signorino 2010). I estimate twelve regression models for each organization (six for the relevant court and six for the ICCPR). To make things explicit, each model can be written as yit∼Bernoulli(pit),logit(pit)=αi+Di,tβ+Ci,tγ+Ti,tδ,αi∼Normal(μi,σ2),μi=α0+Liη, where D is a measure of democracy and/or a transformation of that measure, C is an indicator of conflict occurrence in country i at time t, T is a time counter, and Li indicates whether country i has a common law legal system. To test the hypotheses discussed above, I estimate a model that includes democracy and its square, as well as models that include the change in democracy from a previous year (in addition to control variables). To allow for the fact that a change in democracy may not immediately result in delegation to a human rights court or ratification of the ICCPR, I estimate models that include the change from the previous year ( Di,t−Di,t−1), as well as models that include the change from the previous two years etc., up to five years previous. The model with the squared democracy term allows for the possibility that the probability of delegation is related to democracy in a nonmonotonic fashion as expected by the hypothesis that regimes “in the middle” are more likely to delegate. This expectation would be supported by a negative coefficient for the squared democracy term. The change variables are employed to test whether increases in democracy are associated with delegation to the court, which would support the hypothesis that transitioning regimes are more likely to delegate. The time counters are included because the probability of acceptance may vary over time as a function of unmeasured factors, so including T will reduce bias in parameter estimates (see Carter and Signorino 2010). In the context of a human rights court, such temporal effects are plausible as perceptions of the court’s autonomy and power are likely to change over time and influence the behavior of potential member states (see, for example, Hawkins and Jacoby 2006). To account for unobserved heterogeneity in these two groups of states, country-specific intercepts, α, are included in each model. This is important since countries like the United States and Brazil, for example, are being included in the same analysis as countries like Costa Rica and Barbados. There is thus good reason to believe that each state’s baseline probability of acceptance is not the same and may vary due to unmeasured factors. For example, the United States is notoriously averse to human rights law for reasons that may have nothing to do with democracy or legal system heritage. α is modeled as a function of L, which is a vector measuring common law heritage, in addition to α0, the “grand mean” across all countries. The variance of the country-specific effects, σ2, is estimated from the data. All models were estimated via Markov chain Monte Carlo (MCMC) simulation. Standard estimation techniques for logit models (such as maximum likelihood estimation), and the hypothesis tests which typically accompany them, rely on the notion of repeated random samples drawn from a larger population. Data on CE/OAS member states are not drawn at random from a larger population from which one could draw additional samples. The populations of interest here are CE and OAS member states, and the available data for these populations cannot be considered a random sample (see, for example, Western and Jackman 1994). Since the measure of democracy is estimated from a statistical model, one must account for the uncertainty (measurement error) in this indicator when using it as a covariate. To accomplish this, ten values of the democracy score are drawn randomly from the estimated posterior distribution for each country-year in the data, and each model is estimated ten times. The results are then aggregated across the ten estimates. For each model, two Markov chains were run for at least 20K iterations,8 and the last 10K values from each chain were retained to summarize the posterior distributions of the model parameters. Convergence was assessed with visual diagnostics and a Gelman–Rubin diagnostic test (Gelman and Rubin 1992). Further technical model details can be found in the Appendix. Results Coefficient estimates for democracy, democracy squared, and changes in democracy from models using the CE data are displayed in Figure 1. Estimates from models using the OAS data are shown in Figure 2. The figures show the average estimate (across ten estimates) for each coefficient as a dot, and the horizontal lines connect the minimum lower bound of the 90 percent credible interval to the maximum upper bound. Figure 1. View largeDownload slide Coefficient Estimates for CE models. The average estimates (across ten estimates) are shown as dots, minimum lower/maximum upper bound of the ten 90 percent credible intervals are connected by the horizontal lines Figure 1. View largeDownload slide Coefficient Estimates for CE models. The average estimates (across ten estimates) are shown as dots, minimum lower/maximum upper bound of the ten 90 percent credible intervals are connected by the horizontal lines Figure 2. View largeDownload slide Coefficient Estimates for OAS Models. The average estimates (across ten estimates) are shown as dots, minimum lower/maximum upper bound of the ten 90 percent credible intervals are connected by the horizontal lines Figure 2. View largeDownload slide Coefficient Estimates for OAS Models. The average estimates (across ten estimates) are shown as dots, minimum lower/maximum upper bound of the ten 90 percent credible intervals are connected by the horizontal lines For the European Court, the model with the squared democracy term supports the hypothesis that weakly democratic states are more likely to delegate than are strongly democratic or strongly autocratic states. This is indicated in Figure 1 by the negative coefficient for the squared democracy term, which means that the relationship between democracy and acceptance of a human rights court’s jurisdiction is nonmonotonic in the way anticipated by the argument, with the probability of acceptance increasing at first and then decreasing once the value of democracy is sufficiently high. The minimum upper bound for this coefficient is above zero. This stands in contrast to the estimate for the squared democracy term in the ICCPR model, where five/ten upper bounds are above zero. For the European Court, the evidence concerning transitions to democracy is less definitive. For the European Court model, three of five estimates are positive, but the minimum lower bounds are all below zero, and two of the estimates are actually negative. All the estimates for ICCPR Models 2–6 are positive, but here too, the minimum lower bounds are all below zero. Though the negative estimates for European Court Models 2 and 3 may appear odd, this is due to the fact that there are no cases where a CE member state accepted the jurisdiction of the court after a dramatic increase in democracy that occurred over 1–2 years. But there are several cases of acceptance that occur after changes that take place over 3–5 years as the positive coefficients for European Court Models 4–6 indicate. I discuss several of these cases in more detail below. For the Inter-American Court, the coefficient estimates, shown in Figure 2, more strongly support the hypotheses. Turning first to the coefficient for the squared democracy term, the maximum upper bound of the 90 percent credible interval is above zero, but across ten estimates, this is the only upper bound that crosses zero. Further, even for this one conservative estimate, 92 percent of the posterior distribution for the democracy squared coefficient is below zero. Interestingly, this result is stronger for the Inter-American Court, indicated by the larger (more negative) coefficient for the democracy squared term in Model 1 for the Inter-American Court. Here again, the coefficient estimate for the democracy squared term in the ICCPR model is much smaller, and eight/ten upper bounds are above zero. In the OAS, the claim that increases in democracy increase the probability of delegation also receives strong support. In the case of the Inter-American Court, all the coefficient estimates for Models 2–6 are positive, and the minimum lower bounds for Models 3–6 do not contain zero. In the OAS, even short-term increases in democracy are associated with participation in the Inter-American Court. This is because many cases of delegation to the Inter-American Court occurred almost immediately after democratic transitions as discussed in more detail below. As with the CE, none of the ICCPR models for the OAS indicate that ratification becomes more likely with increases in democracy. In short, the estimates from these models provide strong support for the lock-in/signaling argument as it concerns high delegation institutions, but little evidence that governments use low delegation institutions as commitment devices. A closer examination of the data and model results provides further support for the lock-in/signaling hypotheses as it concerns delegation to human rights institutions. The left panel of Figure 3 plots the probability of acceptance of the European Court’s jurisdiction, from Model 1, against the latent democracy variable.9 The mean probability from each of ten models is shown as a black line, with 90 percent credible intervals shown as gray lines. The distribution of the democracy variable in the CE since 1950 is shown in light gray in the background. The black ticks on the x-axis show the values of democracy at which states that have accepted the court’s jurisdiction did so. The figure indicates that the probability of acceptance is essentially zero among states toward the lower end of the democracy scale, rises as democracy increases, and then declines slightly at high values of democracy. Many cases of acceptance are toward the middle of the democracy scale rather than the extremes, though there are cases toward the high end of the scale. The distribution of the democracy score in the CE is bimodal, with a mode of autocratic country-years toward the left side of the graph and a larger mode of democratic country-years toward the right. It is clear from Figure 3 that autocratic governments do not delegate to the court: there are zero cases of acceptances in the mode on the left side of the graph, and the probability from all models is essentially zero until the democracy score approaches zero. The mode on the left is populated by the communist governments of Central and Eastern Europe before the collapse of the Soviet Union, the right-wing authoritarian governments in Southern Europe that fell in the 1970s, and Turkey in the early 1980s. Central and Eastern European countries, former Soviet republics, and former Yugoslav republics account for most of the cases of acceptance in the middle of the distribution. Figure 3 indicates that there are also cases of countries with relatively high scores on the democracy scale that accepted the court’s jurisdiction relatively early: the United Kingdom, Belgium, the Netherlands, Luxembourg, Iceland, Denmark, Norway, and Austria all accepted the court’s jurisdiction relatively quickly and at high levels of democracy. And Sweden, Italy, France, and Switzerland, though they did not accept jurisdiction quickly, did so at high levels of democracy. This is why, for some models, the probability of acceptance does not decline sharply at high values of democracy. Figure 3. View largeDownload slide Effect of Democracy on Acceptance of the European Court’s Jurisdiction (prior to 1998) and Ratification of the ICCPR. The horizontal axes indicate the latent democracy score; the vertical axes indicate the probability of accepting the court’s jurisdiction/ratifying the ICCPR. The solid black lines show how the probability of acceptance/ratification changes with the democracy score, based on estimates from ten models. 90 percent credible intervals for all models are shown as gray lines. The distribution of the democracy score in the council of Europe for the relevant time period is shown in gray. The black tick marks indicate the values of the democracy score at which governments who have accepted the court’s jurisdiction/ratified the ICCPR did so Figure 3. View largeDownload slide Effect of Democracy on Acceptance of the European Court’s Jurisdiction (prior to 1998) and Ratification of the ICCPR. The horizontal axes indicate the latent democracy score; the vertical axes indicate the probability of accepting the court’s jurisdiction/ratifying the ICCPR. The solid black lines show how the probability of acceptance/ratification changes with the democracy score, based on estimates from ten models. 90 percent credible intervals for all models are shown as gray lines. The distribution of the democracy score in the council of Europe for the relevant time period is shown in gray. The black tick marks indicate the values of the democracy score at which governments who have accepted the court’s jurisdiction/ratified the ICCPR did so The right panel of Figure 3 shows the effect of democracy on ratification of the ICCPR and makes clear that ratification patterns for the ICCPR in the CE do not support the lock-in/signaling hypothesis. Most notably, the probability of ratification is relatively high even at low levels of democracy due to the fact that many governments in the CE ratified the ICCPR during periods of autocracy. This is made clear by the fact that the black tick marks in the right panel extend much further to the left than those in the left panel. CE member states that ratified the ICCPR during periods of autocracy include Poland, Hungary, Czechoslovakia, Bulgaria, Romania, and the Soviet Union. Though all these countries joined the European Court prior to 1998, they did so only after experiencing large increases in democracy. For the Inter-American Court, again, the evidence is even stronger. The left panel of Figure 4 displays the effect of democracy on acceptance of the Inter-American Court’s jurisdiction. The distribution of democracy in the OAS is also bimodal. Here too, no cases of acceptance are found in the autocratic mode, which contains the military dictatorships that appeared in Latin American during the 1970s, as well as Cuba. Figure 4 indicates that, across most models, the probability of accepting the Inter-American Court’s jurisdiction drops quickly toward the high end of the democracy scale. This is because there very few cases of acceptance at high values of democracy. The OAS countries that score the highest and participate in the court are Costa Rica, Trinidad, and Barbados, the last two of which took quite some time to accept jurisdiction (twenty-two and thirty-one years, respectively). It is apparent from the density plot in Figure 4 that many country-years receive relatively high scores, but most of these country-years are for states that have not accepted the court’s jurisdiction to date or had accepted it before they reached their highest level of democracy. The first group includes the United States, Canada, the Bahamas, Jamaica, and Belize. The second group includes Mexico, Panama, Suriname, Peru, Chile, Argentina, and Uruguay. Figure 4. View largeDownload slide Effect of Democracy on Acceptance of the Inter-American Court’s Jurisdiction and Ratification of the ICCPR. The horizontal axes indicate the latent democracy score; the vertical axes indicate the probability of accepting the court’s jurisdiction/ratifying the ICCPR. The solid black lines show how the probability of acceptance/ratification changes with the democracy score, based on estimates from ten models. 90 percent credible intervals for all models are shown as gray lines. The distribution of the democracy score in the OAS between for the relevant time period is shown in gray. The black tick marks indicate the values of the democracy score at which governments who have accepted the court’s jurisdiction/ratified the ICCPR did so Figure 4. View largeDownload slide Effect of Democracy on Acceptance of the Inter-American Court’s Jurisdiction and Ratification of the ICCPR. The horizontal axes indicate the latent democracy score; the vertical axes indicate the probability of accepting the court’s jurisdiction/ratifying the ICCPR. The solid black lines show how the probability of acceptance/ratification changes with the democracy score, based on estimates from ten models. 90 percent credible intervals for all models are shown as gray lines. The distribution of the democracy score in the OAS between for the relevant time period is shown in gray. The black tick marks indicate the values of the democracy score at which governments who have accepted the court’s jurisdiction/ratified the ICCPR did so The right panel of Figure 4 displays the effect of democracy on ratification of the ICCPR. Again the evidence fails to support the hypothesis. As in the CE, the probability of ratifying the ICCPR for OAS member states is nontrivial at low levels of democracy because several governments ratified the ICCPR during periods of autocratic rule. One of the most striking examples is El Salvador in 1979, a year in which a (mostly) military junta controlled the government and the country was in the midst of an escalating civil war. Another example is Panama in 1977, when head of state Omar Torrijos ratified the treaty five years after altering the constitution to remove many constraints on his authority. Torrijos would rule Panama for four more years until his death in 1981. Also, for most models, the probability of ratification continues to rise at high levels of democracy. This is because the most consistently democratic countries in the OAS (the United States, Canada, the Bahamas, Jamaica, and Belize) have all ratified the ICCPR. Substantive effects from the European Court and Inter-American Court models, which include changes in democracy from previous years, are show in Figures 5 and 6.10 For each change model, these figures show, on the vertical axis, the change in the probability of delegation that results from a one-point change in the democracy score (that is, the difference between the probability of delegation if there was a change from a previous year and the probability of delegation if there was no change from a previous year).11 Ten estimates for each change model are shown as dots, and 90 percent credible intervals are shown as vertical lines. For the European Court (Figure 5), the estimates are all positive when examining changes from four or five years previous, but even here nearly all of the credible intervals contain zero. The results from the Inter-American Court (Figure 6) strongly support the hypothesis: for the first change model, seven out of ten credible intervals for the effect of an increase in democracy do not contain zero, and for the other models, none of the intervals contain zero. Figure 5. View largeDownload slide Effect of a One-Point Change in the Democracy Score (from a previous year) on Acceptance of the European Court’s Jurisdiction (prior to 1998). The horizontal axes index the ten estimates; the vertical axes indicate the change in the probability of accepting the court’s jurisdiction that results from a one-point increase in the democracy score. The dots show the mean change, with 90 percent credible intervals shown as vertical lines Figure 5. View largeDownload slide Effect of a One-Point Change in the Democracy Score (from a previous year) on Acceptance of the European Court’s Jurisdiction (prior to 1998). The horizontal axes index the ten estimates; the vertical axes indicate the change in the probability of accepting the court’s jurisdiction that results from a one-point increase in the democracy score. The dots show the mean change, with 90 percent credible intervals shown as vertical lines Figure 6. View largeDownload slide Effect of a One-Point Change in the Democracy Score (from a previous year) on Acceptance of the Inter-American Court’s Jurisdiction. The horizontal axes index the ten estimates; the vertical axes indicate the change in the probability of accepting the court’s jurisdiction that results from a one-point increase in the democracy score. The dots show the mean change, with 90 percent credible intervals shown as vertical lines Figure 6. View largeDownload slide Effect of a One-Point Change in the Democracy Score (from a previous year) on Acceptance of the Inter-American Court’s Jurisdiction. The horizontal axes index the ten estimates; the vertical axes indicate the change in the probability of accepting the court’s jurisdiction that results from a one-point increase in the democracy score. The dots show the mean change, with 90 percent credible intervals shown as vertical lines One more piece of evidence relates to the timing of delegation. The argument under examination predicts that governments should delegate during the midst of, immediately after, or shortly after a transition to a higher level of democracy. The results presented above from the models that include changes in democracy support that claim, and a closer examination of individual cases in the data reveals that delegation to these two regional courts has tended to occur at precisely these times. Figure 7 shows the timing of acceptance for six countries in the CE data.12 Displayed on the y-axis in each graph is the democracy score (shown as a black line), with 95 percent credible intervals (shown as gray lines). The dashed vertical line indicates the year in which the government in question accepted the jurisdiction of the court. In each case, delegation coincides very nearly with a pronounced increase in democracy. Figure 7. View largeDownload slide Timing of European Court acceptance for Six CE countries. The horizontal axis in each plot indicates time; the vertical axis indicates the democracy score. The democracy score for each year is shown as a black line, with 95 percent credible intervals shown in gray. The year in which the government accepted the jurisdiction of the court is indicated by a vertical, dashed line Figure 7. View largeDownload slide Timing of European Court acceptance for Six CE countries. The horizontal axis in each plot indicates time; the vertical axis indicates the democracy score. The democracy score for each year is shown as a black line, with 95 percent credible intervals shown in gray. The year in which the government accepted the jurisdiction of the court is indicated by a vertical, dashed line The six countries shown in Figure 7 illustrate the argument well. Among the most exemplary cases in the CE are Spain, Portugal, and Greece, all countries that transitioned to democracy in the 1970s. Spain and Portugal had both experienced long periods of authoritarian rule prior to democracy. The Spanish government accepted the jurisdiction of the court in 1979, four years after the death of Franco, two years after the 1977 parliamentary elections (the first meaningful elections held in Spain since 1936—before the civil war), and one year after redrafting the constitution. Portugal accepted the court’s jurisdiction in 1978, four years after Marcelo Caetano (successor to Salazar) was ousted by the military and two years after elections that resulted in the first democratic government in Portugal since before Salazar. Greece also accepted the court’s jurisdiction after a transition to democracy, though in this case it took several years. Greece held parliamentary elections in 1974 after seven years of military rule and accepted the court’s jurisdiction in 1979. The cases of Poland, Czechoslovakia, and Bulgaria also illustrate the argument well. Poland and Czechoslovakia both accepted the court’s jurisdiction within two years of their first parliamentary elections after the end of communism. These cases are also notable because both countries had newly elected presidents that were prominent pro-democracy dissidents under their respective communist regimes: Lech Walesa in Poland and Vaclav Havel in Czechoslovakia. Both presidents had well-established policy preferences that favored democracy and faced uncertain political environments, which make the lock-in/signaling hypothesis very plausible. The case of Bulgaria is also illustrative. The Bulgarian Communist Party won the election of 1990, held to form a constitutional assembly. Interestingly, the government accepted the jurisdiction of the court two years later, only after the country’s second noncommunist (and first openly anticommunist) prime minister took office. There are other former communist countries that also follow this pattern, most notably Hungary, Albania, and Romania (see Appendix). Also, several former communist countries delegated to the court at middling values of democracy, though their level of democracy at the time of acceptance was not increasing. However, many of these cases occur after 1994, when the CE was pushing to make acceptance of the court’s jurisdiction mandatory. Thus, while the evidence supports the hypothesis, it is more ambiguous than it would be if Protocol 11 to the European Convention had never been introduced. In the OAS, one can also find a clear record of decisions to join the Inter-American Court following closely after democratic transitions, and here delegation cannot be attributed as easily to pressure from within the institution itself. The timing of acceptance for six OAS countries is shown in Figure 8. These cases also illustrate the argument quite well, with acceptance sometimes occurring immediately after a transition to a higher level of democracy. Guatemala, for example, accepted the jurisdiction of the court in March of 1987, one year after Vinicio Cerezo, the country’s first democratically elected president in twenty years, assumed office. Cerezo was struggling to improve human rights conditions in the face of strong resistance from the military, and his government accepted the jurisdiction of the court only months before establishing the National Commission for Reconciliation, a policy decision that also signals a willingness to respect human rights. Honduras accepted the jurisdiction of the court in 1981, the same year that saw the installation of their first democratically elected president after roughly twenty years of military rule. In the case of Panama, the government submitted to the court’s jurisdiction in 1990, shortly after the US government deposed Noriega and only five months after the newly elected president took office. In Chile, the first democratically elected government after Pinochet accepted the jurisdiction of the court five months after the new president assumed office. In Argentina, the jurisdiction of the court was accepted in September of 1984. This was roughly nine months after the installation of the first democratic president to follow the military government that had ruled the country since 1976 and nine months after the establishment of the National Commission on the Disappearances of Persons. Finally, the Uruguayan government also accepted the court’s jurisdiction shortly after a notable democratic transition. Elections were held in Uruguay in late 1984 after roughly tweleve years of military rule. The new president assumed office in March of 1985, and the government accepted the Inter-American Court’s jurisdiction about seven weeks later. There are several other countries in the OAS that also joined following transitions, the most pronounced cases being Paraguay, Peru, Ecuador, Suriname, Nicaragua, El Salvador, and Haiti (see Appendix). Figure 8. View largeDownload slide Timing of Inter-American Court Acceptance for Six OAS Countries. The horizontal axis in each plot indicates time; the vertical axis indicates the democracy score. The democracy score for each year is shown as a black line, with 95 percent credible intervals shown in gray. The year in which the government accepted the jurisdiction of the court is indicated by a vertical, dashed line Figure 8. View largeDownload slide Timing of Inter-American Court Acceptance for Six OAS Countries. The horizontal axis in each plot indicates time; the vertical axis indicates the democracy score. The democracy score for each year is shown as a black line, with 95 percent credible intervals shown in gray. The year in which the government accepted the jurisdiction of the court is indicated by a vertical, dashed line In summary, the evidence concerning participation in the human rights courts of the CE and the OAS strongly supports the lock-in/signaling argument. The argument seems to explain well delegation in both of these regimes, and the evidence is stronger in the OAS than in the CE. In contrast, there is relatively little evidence to suggest that ratification of the ICCPR follows the pattern predicted by the theory, which suggests that governments view more legalized human rights institutions as better commitment devices and that the theory is best tested by examining such institutions. Discussion/Conclusion From the standpoint of most theories of international cooperation, the proliferation of human rights agreements is one of the more puzzling developments in international law. Some empirical work examines ratification for UN human rights treaties, but very little work examines delegation to strong human rights institutions. This study builds on the work of Moravcsik (2000), Mansfield and Pevehouse (2006, 2008), and Hafner-Burton, Pevehouse, and Mansfield (2015), who argue that states may create/join human rights institutions in order to prevent future leaders from subverting democracy and to signal to the public their commitment to democratic institutions and values. The empirical analysis presented here provides stronger evidence for these claims than previous studies for the simple reason that previous tests to not examine delegation to human rights institutions. If states participate in institutions to signal their intent to adhere to liberal democratic values, then this implies that delegation to human rights institutions specifically should be attractive, not support for the creation of an institution to which one could later delegate and not membership in international institutions broadly. Leveraging the European and Inter-American Courts, I find that the evidence weighs heavily in favor of the theories advanced by Moravcsik (2000) and Mansfield and Pevehouse (2006). In both organizations, delegation to the human rights courts has tended to occur among weakly democratic and democratizing states rather than strongly autocratic or strongly democratic states. Further, many cases of delegation occurred during, immediately after, or very shortly after transitions to higher levels of democracy. The evidence is even stronger in the OAS due to the fact that the most consistently democratic countries in the region have not yet accepted the court’s jurisdiction and that acceptance tended to happen more quickly after democratic transitions in the OAS than in the CE. This is strong evidence in favor of the claim that there is a causal relationship between domestic political institutions and delegation to human rights regimes. Delegation to a human rights court seems to be something a leader in a new or weak democracy can do to signal to the public her intent to respect human rights in the future. Joining a court may be one of several ways that leaders can signal a commitment to rights, such as the establishment of truth/reconciliation commissions as in the cases of Guatemala and Argentina. The fact that leaders have, in some cases, joined human rights courts while simultaneously establishing human rights commissions suggests that there are other kinds of behavior that might be relevant to the theory. The establishment of commissions and more permanent national human rights institutions, like ombudsman offices, may be worth examining in future studies.13 These results also indicate that the theory may have broader applicability than was originally anticipated. Moravcsik (2000, 246) claims that “the determinants of the evolution of human rights regimes are unlikely to be identical to the determinants of their founding.” If the evolution of the regime includes patterns of participation subsequent to its creation, then this claim appears to be incorrect. Just as governments in postwar Europe facing uncertain political environments were keen to create a strong court in the first place, governments with democratic aspirations that faced similarly uncertain environments long after the regime’s founding were eager to delegate authority to the court.14 The same is true of the Inter-American regime, where the decision to delegate seems to be explained by domestic political concerns about consolidating democracy, and where the evidence in favor of this theoretical argument is even stronger. These results are encouraging from a normative perspective. The literature on international human rights law has tended to report bad news. Specifically, most research examining compliance with UN treaties has found that ratification has little unconditional effect on the most repressive kinds of government behavior (Hathaway 2002; Hafner-Burton and Tsutsui 2005; Neumayer 2005; Hafner-Burton and Tsutsui 2007; Hill 2010; Lupu 2013). The analysis above suggests that “false positives,” governments that make legal commitments that they have no intention of upholding (Simmons 2009), are more likely for human rights treaties than for courts, which is one potential explanation for the nonfindings in the compliance literature. Too, studies of treaty ratification have found that states are avoiding obligation where it would entail genuine constraints (Hathaway 2007; Powell and Staton 2009), which also potentially explains these nonfindings. Delegation to courts follows a different pattern. In the case of the both the CE and the OAS, states have tended to submit to a court’s authority when it could conceivably have some impact on behavior, namely when democratic institutions are new or fragile. These are precisely the conditions under which international human rights law is theorized to have a meaningful impact on government behavior (Simmons 2009), which suggests that studying the effect of human rights court rulings conditional on domestic institutions would be useful. The findings above indicate that governments may be making sincere commitments when they delegate to human rights courts. At the very least, delegation to regional human rights courts seems to be a more sincere act than ratification of a UN treaty, suggesting more generally that “decoupling” of legal commitment and practices is less likely in more highly legalized institutions (Cole 2012). Additionally, recent work has shown that the decisions of human rights courts do impact government policy under certain circumstances (Helfer and Voeten 2014; Haglund 2014). Together with the results presented here, this is good news and suggests that further research on the reasons for delegation to, and compliance with, international human rights courts would be fruitful. Daniel W. Hill Jr. received his PhD in Political Science from Florida State University in 2012 and has been an Assistant Professor in the Department of International Affairs at the University of Georgia since 2012. His current main interests are the conceptualization and measurement of human rights practices, quantitative studies of compliance with international human rights law, and statistical forecasts for violent conflict. Footnotes 1But see Fariss (2014), who finds that this result is partially an artifact of data on human rights abuses. The standards that human rights monitors use have become more stringent over the years, so acts that would not be characterized as abuse in the past are characterized that way in the present, which produces artificial stagnation or decline in some indicators. 2Individual petition is an international legal principle that gives individuals the right to initiate judicial proceedings against states. 3See also Cole (2009), who finds that democratization is not correlated with the adoption of treaty protocols that delegate more authority to treaty monitoring bodies. Adopting optional protocols is not equivalent to accepting a court’s jurisdiction, but this finding is inconsistent with Moravcsik’s argument. 4A better approach to a simple additive scale that assumes a weighting scheme for its components would be a measurement model along the lines of that used to create the indicator of democracy used in this study (discussed below) (Pemstein, Meserve, and Melton, 2010). 5Notably, the HRC does not have the authority to issue public, legally binding decisions about compliance with the provisions of the ICCPR, relies largely on self-reporting by member states, and individuals have no standing to bring complaints against their government unless their government has ratified the Second Optional Protocol to the ICCPR. For an examination of the effectiveness of the HRC and similar UN treaty-based monitoring bodies in improving human rights conditions, see Krommendijk (2015). 6Treier and Jackman (2008) demonstrate that results from regression analyses using the Polity scale and its squared value do not hold when the measurement error is accounted for. 7Although it is common to include second- and third-order polynomials of time in such models, including polynomials did not improve the fit of the models presented below and caused complications during estimation (nonconvergence in some cases). Since the models allow for a monotonically increasing or decreasing hazard rate, they are analogous to Weibull models (with frailties). 8For some models, it was necessary to extend the MCMC run to as many as 150K iterations. 9When calculating the values shown in the figures, the other covariates were set to the following values: 0 for conflict, 0 for common law, and 20 for the time counter. 10As with the two previous figures, conflict was set to 0, common law to 0, and the time counter to 20. 11A one-point change is large, but in each organization, there are many cases where a country experienced a change this large over 2–5 years, and in some cases even over a single year, such as Argentina 1982–1983 and Spain 1976-1977. 12Plots for the other CE countries can be found in the Appendix. 13Recent data collection efforts by Conrad, Demeritt, and Moore (2014) have made such studies possible. See Welch (Forthcoming) for an analysis of the effect of national human rights institution (NHRIs) on human rights practices. 14See also Voeten (2007), who finds that governments with EU membership aspirations are more likely to appoint activist judges to the court in order to signal a commitment to democracy and human rights. Appendices A Model Details Each model estimated can be written as yit∼Bernoulli(pit)Logit(pit)=αi+Di,tβ+Ci,tγ+Ti,tδαi∼Normal(μi,σ2)μi=α0+Liη. To estimate the model’s parameters via MCMC, prior distributions must be specified for the country-year level coefficients β, γ, and δ, as well as the intercept (α0) and coefficient η in the hierarchical model for the country-specific intercepts, and the variance of these intercepts, σ2. β, γ, δ, α0, and η were all assigned normal priors with means 0 and variances of 10. σ2 was assigned a gamma prior with a scale parameter of 1 and a shape parameter of 10. Convergence was assessed with a Gelman–Rubin diagnostic (upper confidence intervals close to 1) and visual diagnostics (posterior densities unimodal and roughly normal). B Time-Series Plots for Other Countries Displayed below in Figures 9–12 are time-series plots for the democracy score (Pemstein Meserve and Melton, 2010) for each country in the sample. The black line tracks the point estimate for the democracy score, with 90 percent credible intervals shown in gray. Bear in mind that these intervals represent the uncertainty in these estimates. For any two time points where there is a considerable amount of overlap in these intervals, the levels of democracy at these points in time are not statistically distinguishable from one another. For example, while the graph for the United States shows an increase in democracy shortly after 1990, the fact that the intervals at this point in time overlap with the intervals at every other point indicates that an analyst would be wrong to judge the US level of democracy as being higher at this point than at any point in the past or the future. Figure 9. View largeDownload slide Timing of European Court acceptance for other CE countries Figure 9. View largeDownload slide Timing of European Court acceptance for other CE countries Figure 10. View largeDownload slide Timing of European Court acceptance for other CE countries Figure 10. View largeDownload slide Timing of European Court acceptance for other CE countries Figure 11. View largeDownload slide Timing of Inter-American Court acceptance for other OAS countries Figure 11. View largeDownload slide Timing of Inter-American Court acceptance for other OAS countries Figure 12. View largeDownload slide Timing of Inter-American Court acceptance for other OAS countries Figure 12. View largeDownload slide Timing of Inter-American Court acceptance for other OAS countries References Abbott Kenneth W. , Snidal Duncan . 2000 . “ Hard and Soft Law in International Governance .” International Organization 54 : 421 – 56 . Abbott Kenneth W. , Keohane Robert O. , Moravcsik Andrew , Slaughter Anne Marie , Snidal Duncan . 2000 . “ The Concept of Legalization .” International Organization 54 : 401 – 19 . Achen Christopher H. 2002 . “ Toward a New Political Methodology: Microfoundations and ART .” Annual Review of Political Science 5 : 423 – 50 . Alter Karen . 2006 . Delegation to International Courts and the Limits of Recontracting Power. In Delegation and Agency in International Organizations , edited by Hawkins Darren G. , Lake David A. , Nielson Daniel L. , Tierney Michael J. , 312 – 38 . New York : Cambridge University Press . Alter Karen . 2008 . “ Delegating to International Courts: Self-Binding vs. Other-Binding Delegation .” Law and Contemporary Problems 71 : 37 – 76 . Beck Neal , Katz Jonathan N. , Tucker Richard . 1998 . “ Taking Time Seriously: Time-Series-Cross-Section Analysis with a Binary Dependent Variable .” American Journal of Political Science 42 : 1260 – 88 . Bernhard William , Broz Lawrence , Clark William Roberts . 2002 . “ The Political Economy of Monetary Institutions .” International Organization 56 : 693 – 723 . Bradley Curtis A. , Kelley Judith G. . 2008 . “ The Concept of International Delegation .” Law and Contemporary Problems 71 : 1 – 36 . Carter David B. , Signorino Curtis S. . 2010 . “ Back to the Future: Modeling Time Dependence in Binary Data .” Political Analysis 18 : 271. Chapman Terrence L. , Chaudoin Stephen . 2012 . “ Ratification Patterns and the International Criminal Court .” International Studies Quarterly 57 : 400 – 9 . Cole Wade M. 2005 . “ Sovereignty Relinquished? Explaining Commitment to the International Human Rights Covenants, 1966–1999 .” American Sociological Review 70 : 472 – 95 . Cole Wade M. . 2009 . “ Hard and Soft Commitments to Human Rights Treaties, 1966–2000 .” Sociological Forum 24 : 563 – 88 . Cole Wade M. . 2012 . “ Human Rights as Myth and Ceremony? Reevaluating the Effectiveness of Human Rights Treaties, 1981–2007 .” American Journal of Sociology 117 : 1131 – 71 . Conrad Courtenay R. , Demeritt Jaqueline H. R. , Moore Will H. . 2014 . “The National Human Rights Institutions Data Project.” Accessed April 6, 2016. http://nhridata.weebly.com/data.html. Fariss Christopher J. 2014 . “ Respect for Human Rights Has Improved Over Time: Modeling the Changing Standard of Accountability .” American Political Science Review 108 : 297 – 318 . Gelman Andrew , Rubin Donald B. . 1992 . “ Inference from Iterative Simulation Using Multiple Sequences .” Statistical Science 7 : 457 – 72 . Gleditsch Kristian Skrede , Ward Michael D. . 1997 . “ Double Take: A Reexamination of Democracy and Autocracy in Modern Polities .” Journal of Conflict Resolution 41 : 361 – 83 . Gleditsch Nils Petter , Wallensteen Peter , Eriksson Mikael , Sollenberg Margareta , Strand Håvard . 2002 . “ Armed conflict 1946–2001: A New Dataset .” Journal of Peace Research 39 : 615 – 37 . Goodliffe Jay , Hawkins Darren G. . 2006 . “ Explaining Commitment: States and the Convention against Torture .” Journal of Politics 68 : 371 . Hafner-Burton Emilie M. , Tsutsui Kiyoteru . 2005 . “ Human Rights in a Globalizing World: The Paradox of Empty Promises .” American Journal of Sociology 110 : 1373 – 411 . Hafner-Burton Emilie M. , Tsutsui Kiyoteru . 2007 . “ Justice Lost! The Failure of International Human Rights Law to Matter Where Needed Most .” Journal of Peace Research 44 : 407 – 25 . Hafner-Burton Emilie , Pevehouse Jon , Mansfield Edward . 2015 . “ Human Rights Institutions, Sovereignty Costs, and Democratization .” British Journal of Political Science 45 : 1 – 25 . Haglund Jilliene N. D. 2014 . “Domestic Implementation of Supranational Court Decisions: The Role of Domestic Politics in Respect for Human Rights.” Paper presented at the Annual Meeting of the International Studies Association, Toronto, Canada. Hathaway Oona A. 2002 . “ Do Human Rights Treaties Make a Difference? ” Yale Law Journal 111 : 1935 – 2042 . Hathaway Oona A. . 2007 . “ Why Do Countries Commit to Human Rights Treaties? ” Journal of Conflict Resolution 51 : 588 – 621 . Hathaway Oona A. . 2008 . “ International Delegation and State Sovereignty .” Law and Contemporary Problems 71 : 115 . Hawkins Darren G. , Lake David A. , Nielson Daniel L. , Tierney Michael J. . 2006 . Delegation and Agency in International Organizations . New York : Cambridge University Press . Hawkins Darren G. , Jacoby Wade . 2006 . “How Agents Matter.” In Delegation and Agency in International Organizations , edited by Hawkins Darren G. , Lake David A. , Nielson Daniel L. , Tierney Michael J. , 199 – 228 . New York : Cambridge University Press . Hegre Håvard , Ellingsen Tanja , Gates Scott , Gleditsch Nils Petter . 2001 . “ Toward a Democratic Civil Peace? Democracy, Political Change, and Civil War, 1816–1992 .” American Political Science Review 95 : 33 – 48 . Helfer Laurence , Voeten Erik . 2014 . “ International Courts as Agents of Legal Change: Evidence from LGBT Rights in Europe .” International Organization 68 : 77 – 110 . Hill Daniel W. 2010 . “ Estimating the Effects of Human Rights Treaties on State Behavior .” The Journal of Politics 72 : 1161 – 74 . Koremenos Barbara . 2007 . “ If Only Half of International Agreements Have Dispute Resolution Provisions, Which Half Needs Explaining? ” The Journal of Legal Studies 36 : 189 – 212 . Krommendijk Jasper . 2015 . “ The Domestic Effectiveness of International Human Rights Monitoring in Established Democracies. The Case of the UN Human Rights Treaty Bodies .” Review of International Organizations 10 : 489 – 512 . Lohmann Susanne . 1992 . “ Optimal Commitment in Monetary Policy: Credibility Versus Flexibility .” The American Economic Review 82 : 273 – 86 . Lupu Yonaton . 2013 . “ Best Evidence: The Role of Information in Domestic Judicial Enforcement of International Human Rights Agreements .” International Organization 67 : 469 – 503 . Mansfield Edward D. , Pevehouse Jon C. . 2006 . “ Democratization and International Organizations .” International Organization 60 : 137 – 67 . Mansfield Edward D. , Pevehouse Jon C. . 2008 . “ Democratization and the Varieties of International Organizations .” Journal of Conflict Resolution 52 : 269 – 94 . Meyer John W. , Boli John , Thomas George M. , Ramirez Francisco O. . 1997 . “ World Society and the Nation-State .” American Journal of Sociology 103 : 144 – 81 . Mitchell Sarah McLaughlin , Powell Emilia Justyna . 2009 . “ Legal Systems and Variance in the Design of Commitments to the International Court of Justice .” Conflict Management and Peace Science 26 : 164 – 90 . Mitchell Sarah McLaughlin , Powell Emilia Justyna . 2011 . Domestic Law Goes Global: Legal Traditions and International Courts . New York : Cambridge University Press . Moravcsik Andrew . 2000 . “ The Origins of Human Rights Regimes: Democratic Delegation in Postwar Europe .” International Organization 54 : 217 – 52 . Muller Edward N. 1985 . “ Income Inequality, Regime Repressiveness, and Political Violence .” American Sociological Review 50 : 47 – 61 . Neumayer Eric . 2005 . “ Do International Human Rights Treaties Improve Respect for Human Rights? ” Journal of Conflict Resolution 49 : 925 – 53 . North Douglass C. , Weingast Barry R. . 1989 . “ Constitutions and Commitment: The Evolution of Institutions Governing Public Choice in Seventeenth-Century England .” The Journal of Economic History 49 : 803 – 32 . Pemstein Daniel , Meserve Stephen A. , Melton James . 2010 . “ Democratic Compromise: A Latent Variable Analysis of Ten Measures of Regime Type .” Political Analysis 18 : 426 – 49 . Pevehouse Jon , Nordstrom Timothy , Warnke Kevin . 2004 . “ The Correlates of War International Governmental Organizations Data Version 2.0 .” Conflict Management and Peace Science 21 : 101 – 19 . Powell Emilia Justyna. , Staton Jeffrey K. . 2009 . “ Domestic Judicial Institutions and Human Rights Treaty Violation .” International Studies Quarterly 53 : 149 – 74 . Powell Emilia Justyna , Mitchell Sarah McLaughlin . 2007 . “ The International Court of Justice and the World’s Three Legal Systems .” Journal of Politics 69 : 397 – 415 . Quinn Kevin M. 2004 . “ Bayesian Factor Analysis for Mixed Ordinal and Continuous Responses .” Political Analysis 12 : 338 – 53 . Risse Thomas , Ropp Stephen C. , Sikkink Kathryn , eds. 1999 . The Power of Human Rights: International Norms and Domestic Change . New York : Cambridge University Press . Rogoff Kenneth . 1985 . “ The Optimal Degree of Commitment to an Intermediate Monetary Target .” The Quarterly Journal of Economics 100 : 1169 – 89 . Schrodt Philip A. 2014 . “ Seven Deadly Sins of Contemporary Quantitative Political Analysis .” Journal of Peace Research 51 : 287 – 300 . Simmons Beth A. 2009 . Mobilizing for Human Rights: International Law in Domestic Politics . New York : Cambridge University Press . Simmons Beth A. , Danner Allison . 2010 . “ Credible Commitments and the International Criminal Court .” International Organization 64 ( 2 ): 225 – 56 . Themnér Lotta , Wallensteen Peter . 2012 . “ Armed Conflicts, 1946–2011 .” Journal of Peace Research 49 : 565 – 75 . Treier Shawn , Jackman Simon . 2008 . “ Democracy as a Latent Variable .” American Journal of Political Science 52 : 201 – 17 . Voeten Erik . 2007 . “ The Politics of International Judicial Appointments: Evidence from the European Court of Human Rights .” International Organization 61 : 669 – 701 . Welch Ryan M . Forthcoming . “ National Human Rights Institutions: Domestic Implementation of International Human Rights Law .” Journal of Human Rights . Doi: 10.1080/14754835.2015.1103166. Western Bruce , Jackman Simon . 1994 . “ Bayesian Inference in Comparative Research .” American Political Science Review 88 : 412 – 23 . © The Author (2016). Published by Oxford University Press on behalf of the International Studies Association. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Foreign Policy Analysis Oxford University Press

Why Governments Cede Sovereignty: Evidence From Regional Human Rights Courts

Foreign Policy Analysis , Volume Advance Article (3) – May 3, 2016

Loading next page...
 
/lp/ou_press/why-governments-cede-sovereignty-evidence-from-regional-human-rights-ZO31hkFomP
Publisher
Oxford University Press
Copyright
© The Author (2016). Published by Oxford University Press on behalf of the International Studies Association.
ISSN
1743-8586
eISSN
1743-8594
D.O.I.
10.1093/fpa/orw031
Publisher site
See Article on Publisher Site

Abstract

Abstract This study examines a government’s decision to cede authority over fundamental questions of policy to international organizations. Explanations for delegation to international institutions focus on the benefits of “locking in” domestic political institutions and signaling intentions to the public. This line of reasoning suggests that partly democratic and democratizing governments have the strongest incentive to participate in international human rights institutions. I evaluate this claim by analyzing patterns of participation in two regional human rights courts: the European Court of Human Rights and the Inter-American Court of Human Rights. The decision to join a regional court is a more valid indicator of delegation than those used in previous empirical studies and to date has not been examined in systematic fashion. I find strong evidence in support of the claim that governments grant authority to institutions to credibly commit to maintaining democracy. Introduction Why would a government cede authority over fundamental questions of policy to an international institution? This is a puzzling question, and in the realm of human rights, it becomes more puzzling. In contrast to most international agreements, human rights agreements attempt to constrain states’ behavior toward their own citizens rather than toward one another and so by definition represent a greater encroachment on state sovereignty. In addition, compared with other policy areas, the efficiency gains from creating these agreements are not obvious: governments who wish to observe the human rights of their citizens do not gain anything additional by jointly observing human rights (Koremenos 2007; Simmons 2009). That said, participation in many human rights institutions does not entail large sovereignty costs. Most analyses of the decision to join these institutions focus on UN-based human rights treaties for which monitoring and enforcement provisions are notoriously weak (Cole 2005; Goodliffe and Hawkins 2006; Hathaway 2007; Cole 2009; Powell and Staton 2009; Simmons 2009). In terms familiar to scholars of international cooperation, these agreements generally entail high levels of obligation but very little delegation (Abbott et al. 2000). The most severe form of sovereignty loss is delegation to an international human rights court (Bradley and Kelley 2008). This requires not only that a government agree to follow a set of externally created rules that limit their domestic political authority but also that they authorize a body they have limited control over to issue public judgments about whether they are complying with those rules. That is, submitting to the authority of an international court entails a higher level of delegation than treaty commitments and, in the area of human rights, delegation necessarily encroaches on sovereignty (Abbott and Snidal 2000; Hathway 2008). Explanations for international delegation focus on the benefits of constraining government behavior in the long term. This is a familiar reason for delegation, which is invoked, for example, in the literatures on property rights and economic growth (North and Weingast 1989) and delegation of monetary policy to central banks (Rogoff 1985; Lohmann 1992; Bernhard, Broz, and Clark 2002). International human rights courts in particular have been theorized as commitment devices for newly democratic governments who wish to signal a long-term commitment to democracy. Moravcsik (2000) argues that states create these courts in order to “lock in” good practices and constrain the behavior of potential successors. Here, the government is worried about antidemocratic opponents coming to power and subverting fledgling democratic institutions. Mansfield and Pevehouse (2006) build on this argument, claiming that democratizing states grant authority to international organizations (IOs) in order to raise the costs of reneging on reforms and credibly signal, to domestic and international audiences, their intent to remain democratic. Both these explanations view delegation as a mechanism for “self-binding” (Alter 2008), meaning the intent is to constrain government behavior, and both focus on the capacity of institutions to make undemocratic behavior costly by monitoring behavior and providing information about compliance. This line of reasoning has two implications: (1) delegation to human rights regimes should be attractive to newly democratized governments and (2) delegation to human rights regimes should be attractive to governments with weakly democratic institutions. Moravcsik (2000), Mansfield and Pevehouse (2006, 2008), and Hafner-Burton, Pevehouse, and Mansfield (2015) test these hypotheses, but the design of these studies prevents firm conclusions. Most importantly, all fail to create a good match between the concept of delegation and its operationalization. Specifically, they examine policy decisions that do not necessarily entail granting an institution real authority to monitor and publicize antidemocratic behavior. I elaborate this point below and provide a new analysis to evaluate the lock-in/signaling explanation for delegation to international institutions. I do this by examining delegation to human rights courts in the Council of Europe (hereafter the CE) and the Organization of American States (hereafter the OAS). Beginning in 1950, CE members had the option of accepting the jurisdiction of the European Court of Human Rights (hereafter the European Court) and have had no choice about this matter since 1998. The European Court is authorized to make legally binding, public decisions regarding compliance with the European Convention on Human Rights (hereafter the European Convention). Since 1969 members of the OAS have had the option of accepting the jurisdiction of the Inter-American Court of Human Rights (hereafter the Inter-American Court), which is authorized to make judgments regarding compliance with the American Convention on Human Rights (hereafter the American Convention). Accepting the jurisdiction of these courts clearly involves delegation to a relatively strong institution tasked explicitly with monitoring and publicizing human rights abuses. This means that acceptance of a regional court’s jurisdiction is a strong indicator of delegation to an institution that promotes human rights and democracy. Also, Europe and the Western hemisphere have both experienced a large number of democratic transitions since the creation of their respective human rights courts. This variation allows me to examine the reasons for acceptance of each court’s jurisdiction in a systematic fashion over a relatively long time period. Although delegation to international institutions has been the subject of much academic writing (see, for example, Hawkins et al. 2006), systematic empirical research in this area is thin (Bradley and Kelley 2008). I contribute to this literature by examining the link between democracy and commitments to international human rights institutions with varying levels of legalization. I find strong evidence consistent with the claim that uncertainty about the future of democracy creates incentives to join international human rights courts. I contrast these results with an analysis of ratification of the International Covenant on Civil and Political Rights (hereafter the ICCPR) between CE and OAS members. The ICCPR is a universal treaty drafted by the United Nation which is nearly identical to the European and American Conventions in terms of content but which does not delegate monitoring/enforcement power to the same extent as either of these treaties. I demonstrate that patterns of participation are different for human rights courts and treaties, which suggest that the design of international institutions affects how useful they are as commitment and signaling devices and is something that must be accounted for in empirical analysis. I follow the analysis with a discussion of the implications of these findings for future research. Why Delegate to International Human Rights Institutions? As mentioned above, most empirical research on human rights law examines UN treaties, which create legal obligations but do not rank as highly on the delegation dimension of legalization as human rights courts. UN treaties do create committees that serve as monitoring bodies, but these committees are limited in their information gathering capacity and, more importantly, do not have any legal authority to issue public, binding decisions concerning alleged treaty violations. This institutional feature of the treaty regime leads to participation that may not reflect a genuine commitment to protecting human rights. Research suggests that ratification is often not a sincere act; many studies find a weak (or even negative) unconditional relationship between ratification and respect for human rights (Hathaway 2002; Neumayer 2005; Hill 2010).1 States with questionable practices may ratify to receive an “expressive benefit” (Hathaway 2002), which refers to the approbation governments sometimes receive from communicating to other states that they accept the prevalent norms of the international community (Risse, Ropp, and Sikkink 1999). Studies informed by the global institutionalist school (Meyer et al. 1997) similarly argue that this “decoupling” results from a combination of (global) cultural pressure to express support for human rights norms and weak international monitoring and enforcement (for example, Cole 2005; Hafner-Burton and Tsutsui 2005). In line with these arguments, there is evidence that institutions with stricter monitoring procedures mitigate the decoupling of rules and practice. For example, Cole (2012) finds that the adoption of optional protocols, which grant more monitoring power to human rights committees than the treaties themselves, is actually associated with better human rights practices. This suggests that variation in the authority of delegated bodies across human rights regimes is related to patterns of participation in predictable ways, a point that I elaborate below. Others focus on a lack of domestic enforcement to explain the disconnect between treaty ratification and human rights behavior. Hathaway (2007) and Powell and Staton (2009) find that governments with poor practices but weak domestic courts are less hesitant to ratify, whereas those with poor practices and strong courts are more cautious. Since states that face a low likelihood of domestic enforcement are more likely to join, treaties have limited impact on states that ratify. Perhaps it is unsurprising that willful evasion of international and domestic enforcement mechanisms is common among states that ratify human rights treaties. After all, human rights agreements encroach on sovereignty to a greater extent than other kinds of agreements. However, evading meaningful legal constraints is not as feasible where human rights courts are concerned. Membership in a court involves a greater degree of delegation than membership in a treaty; submitting to a court’s jurisdiction creates the possibility of public, legally binding judgments that the government is violating its human rights obligations, which is not true of a treaty. Of course, governments can always refuse to comply with court rulings just as they can refuse to comply with treaty provisions. But noncompliance with a court ruling entails publicly defying a legal body whose authority the government has formally recognized. This is not to say that states cannot circumscribe the authority of international courts. Scholars of judicial politics have extensively cataloged the strategies governments may adopt to limit the autonomy of domestic courts, such as screening/appointing judges based on ideology and writing precise legislation that limits judicial interpretation. But these strategies are not always effective in the domestic context and are more difficult to employ in the international context due to the relative difficulty of changing international law and more limited control over appointments (Alter 2006). Additionally, judges on human rights courts can still interpret rules in unanticipated ways and sometimes do so in ways that increase their autonomy. For example, Hawkins and Jacoby (2006) document how the European Court, without a directive from member states, used procedural rules to enhance individual access to the court. These institutional differences between treaties and courts should produce different patterns of participation. Since courts create a greater likelihood of human rights abuses being exposed and discussed publicly, states should be cautious about making insincere commitments that lead to a decoupling of legal rules and practices. In line with this observation, explanations for participation in courts tend to focus on the conditions that make it attractive to submit to meaningful legal constraints. The argument made by Moravcsik (2000) appeals to an uncertain political environment to explain delegation to human rights regimes. Governments in recently democratized countries have good reason to believe that their successors will not share their preferences for democratization and, to reduce the risk of a reversal, may support the creation of a human rights courts in order to “lock in” liberal institutions. A human rights court presents these governments with an opportunity to “establish reliable judicial constraints” on successors by “placing interpretation in the hands of independent authorities managed in part by foreign governments” (Moravcsik 2000, 228). For these governments, the benefits of biasing policy toward their ideal point outweigh the sovereignty costs incurred by delegating to an international court. The benefits are too small for consolidated democracies, who have already locked in liberal institutions, and autocrats do not prefer to create or lock-in liberal institutions. The implication of this argument is that stable democracies and stable dictatorships will have the least incentive to join, producing an “inverse-U-shaped” relationship between the strength of democratic institutions and a government’s willingness to participate in a human rights court (Moravcsik 2000, 232). Mansfield and Pevehouse (2006, 2008) also appeal to the uncertainty present in transitioning regimes to explain why governments join international institutions. They make a valuable theoretical contribution by pointing out that IO membership helps to tie the hands of the current (and perhaps future) government and credibly signals to interested parties (the domestic public, international investors, etc.) the current government’s willingness to preserve democracy in the long term. This calls attention to a related incentive that leaders have for delegating to international institutions: by creating a cost for violating the rules, joining an institution can signal a willingness to follow the rules. Mansfield and Pevehouse (2006, 2008) present their argument as an explanation for IO membership generally, but the mechanism they seem to have in mind involves delegation (as defined by Abbott et al. 2000) to institutions that have an explicit mission to promote democracy and human rights. They claim that joining an IO credibly signals a commitment to democracy because “monitoring and enforcement are handled by a third party with the ability to publicly sound an alarm in the event that reform efforts falter” (Mansfield and Pevehouse, 2006, 141). They hypothesize that democratizing regimes have the strongest incentives to grant authority to IOs. To summarize, previous work examining why states incur sovereignty costs argues that political uncertainty resulting from democratizing/weak political institutions is the primary motivation for delegation to international human rights institutions. The implications of this argument are that transitions to democracy, as well as weakly democratic institutions, make delegation to human rights institutions more likely. Yet it is unclear whether the empirical record supports these hypotheses because these claims have been inadequately evaluated in previous work, a point that I elaborate below. Existing Evidence for the Lock-In/Signaling Theory The object of explanation for this study is the decision to accept the jurisdiction of a human rights court. Moravcsik (2000) examines preliminary negotiations about the features of the European Court and demonstrates that the earliest supporters of granting the court the powers of individual petition2 and compulsory jurisdiction tended to be newly democratic states. Although the evidence seems to support his argument, it is based on a relatively small number of cases. It is also limited to a short time frame rather than subsequent patterns of participation that unfolded over a longer period of time. Most importantly, this study does not examine actual delegation but rather support for the creation of a regime to which one could later delegate. Examining acceptance of the European Court’s jurisdiction would be a better, more direct test of the empirical implications of this argument. Explanations for delegation are best tested by examining actual participation in institutions rather than support for strong institutions in principle. It is not clear why supporting the creation of a court would raise the cost of reversing democracy if a state had not accepted its jurisdiction, though to be fair, Moravcsik examines support for compulsory jurisdiction, which would entail acceptance for all countries if adopted. But compulsory jurisdiction was not adopted when the European Convention was opened for ratification in 1950, which meant CE governments could (until 1998) ratify the European Convention without accepting the court’s jurisdiction. And in fact two of the seven governments that Moravcsik lists as initially supportive of compulsory jurisdiction, Italy and France, waited until 1973 and 1974, respectively, to accept jurisdiction. The United Kingdom accepted jurisdiction several years earlier (1966), even though its initial recalcitrance during negotiations about granting the court extended powers is well documented and discussed by Moravcsik. Additionally, Denmark, Belgium, the Netherlands, and Luxembourg, four of the seven established democracies listed by Moravcsik, were among the first to accept jurisdiction (1953, 1955, 1954, and 1958, respectively). Thus, the evidence on this count is not entirely convincing,3 and examining the available evidence more systematically is difficult, given the institutional changes that occurred in the CE beginning in 1994. In 1994, Protocol 11 to the European Convention, which would create compulsory jurisdiction once ratified by the requisite number of states, became open for ratification. This means that the largest wave of democratic transitions to occur in Europe since the court’s creation occurred at roughly the same time that the CE began moving toward compulsory jurisdiction, which makes it difficult to separate the effects of democratization from institutional changes in the CE itself. Although Moravcsik (2000) operationalizes delegation too narrowly, Mansfield and Pevehouse (2006, 2008) do so too broadly and measure IO membership generally rather than delegation to institutions devoted to democracy and human rights. These studies show that democratizing states join, on average, greater numbers of IOs than nondemocratizing states, but the indicators of IO membership used in these studies are not ideal for measuring the kind of institutional commitment needed to make a credible commitment to democracy. The lock-in/signaling argument focuses on IOs’ ability to monitor and publicize state behavior that violates the organization’s standards. This suggests that examining commitments to highly legalized human rights institutions would be more appropriate than examining IO membership generally. Delegation to a human rights court requires a primary legal obligation and submission to public, legally binding, judicial decisions. Publicly pledging to follow the rules of an institution devoted explicitly to many of the rights associated with democracy, and granting a court the authority to publicly declare behavior noncompliant, creates a much clearer cost for violating those rights than does membership in various IOs. Membership in IOs that have little to do with human rights or democracy does not obviously create such a cost, but the indicators used in these studies mostly measure institutions that have little to do with these issues, which means that the measure is very noisy. Further, summing the number of IOs that a state belongs to, as Mansfield and Pevehouse (2006, 2008) do, assumes that each of these IOs has an equivalent effect on the cost of undemocratic behavior. This is a strong assumption given the diversity of IOs included in their data, which include such organizations as the International Office of Epizootics and the Inter-State School of Hydraulic and Rural Engineering for Senior Technicians (Pevehouse, Nordstrom, and Warnke 2004). In a more recent study, Hafner-Burton, Pevehouse, and Mansfield (2015) offer an improved test by examining human rights-related organizations (defined as organizations whose stated goals include any type of human rights program) and developing a scale to measure the level of legalization created by a particular organization. But this measure also groups together IOs that are very heterogeneous in terms of both the level of delegation that they require (many involve no monitoring or oversight whatsoever) and the extent to which their goals are explicitly human rights-related. For example, the Organization for Economic Cooperation and Development (OECD) and North Atlantic Treaty Organization (NATO) are counted as “human rights IOs” alongside the European and Inter-American Courts. This indicator also assumes a priori a weighting scheme for the various institutional features of these organizations, which are all assumed to contribute equally to an institution’s level of legalization. This means that this measure is also noisy.4 Examining participation in regional human rights courts provides a simple but strong empirical test of explanations for delegation. This binary indicator is less noisy, and more valid, than the measures of IO membership employed by Mansfield and Pevehouse (2006, 2008) and Hafner-Burton, Pevehouse, and Mansfield (2015) or the negotiation positions examined by Moravcsik (2000) which, as noted above, do not correlate perfectly with acceptance of the European Court’s jurisdiction. This means that the inferences drawn from the analysis below are stronger than in previous studies and provide better evidence concerning the veracity of the lock-in/signaling argument. To demonstrate that this argument is best treated as an explanation for delegation to human rights regimes, I contrast participation in human rights courts with ratification of the ICCPR. The ICCPR is nearly identical in content to the European and American Conventions but does not create a court to monitor and enforce its provisions. The ICCPR creates the UN Human Rights Committee (hereafter HRC), which has some limited information gathering capacity, but its degree of legal authority is less than that of either the European Court or the Inter-American Court.5 Granting legal authority to the European Court or Inter-American Court clearly entails strong delegation to an institution devoted explicitly to human rights. Ratifying the ICCPR entails a nearly identical legal obligation, but little delegation. Thus, any differences in participation can be attributed to different levels of legalization across these institutions. Examining Evidence from Regional Human Rights Courts The CE was created in 1949 and today is comprised of forty-seven member states. Between 1950 and 1998, its members had the option of accepting the jurisdiction of the European Court, which after 1998, was no longer optional. There is a fair amount of variation in democracy among CE countries during this time period, due mostly to the democratic transitions in Southern Europe during the 1970s and Eastern Europe in the 1990s. As noted above, many of the transitions in the latter group occurred during the midst of the CE’s formal push toward compulsory jurisdiction. The OAS was formed in 1948, and its membership today consists of thirty-five countries. OAS members have had the option of accepting the jurisdiction of the Inter-American Court since 1969. Unlike the European Court, the jurisdiction of the Inter-American Court is not compulsory, which means that there is still some variation in delegation among OAS members. Twenty-one of the OAS’s thirty-five member states have accepted the court’s jurisdiction to date, with the last acceptance occurring in 2000 by Barbados. This, together with the fact that domestic political institutions across the Western hemisphere have fluctuated a great deal since 1969, makes the OAS an arguably better organization in which to test explanations for delegation. The key explanatory concept in the theory discussed above is democracy. Past work has examined two implications of this theory: (1) increases in/transitions to democracy make delegation to international human rights institutions more likely and (2) the relationship between democracy and delegation should be inverse-U-shaped. Testing these implications requires a measure of democracy that allows valid distinctions among autocracies, democracies, and weakly democratic or democratizing regimes. Most studies of the adoption of international human rights law treat democracy as an important explanatory concept, and many of these studies use the Polity scale or one of its components (Hathaway 2007; Simmons 2009; Simmons and Danner 2010; Chapman and Chaudoin 2012) as do the studies by Mansfield and Pevehouse (2006, 2008) and Hafner-Burton, Pevehouse, and Mansfield (2015). The Polity scale may be a reasonable indicator of democracy, but it is not ideal for testing these hypotheses. Analyses of the Polity data have uncovered problems with the scale’s construction and aggregation (Gleditsch and Ward 1997; Treier and Jackman 2008). Most importantly for this study, there is noise in the scale such that distinctions between states toward the high/low ends of the scale and those in the middle may not be valid. That is, in some cases, the scale does a poor job distinguishing between countries that have middling levels of democracy and those that have low/high levels. Thus, the measurement error that is known to exist in Polity makes it a poor choice for testing hypotheses about the behavior of weakly democratic states as this measurement error would call into question any inferences drawn about acceptance rates of weakly democratic states relative to states closer to the ends of the scale.6 I adopt instead the measure of democracy created by Pemstein, Meserve, and Melton (2010). This measure is derived from ten different commonly employed indicators of democracy. The authors employ a Bayesian version of factor analysis that allows for continuous and ordinal indicators (see Quinn 2004). The resulting measure is continuous and thus allows for distinctions between gradations of democracy. Perhaps the biggest advantage of using this indicator is that their model produces estimates of uncertainty (measurement error) around each country-year’s value of democracy that can be incorporated into subsequent analysis. In all the models below, I also include a measure of domestic legal system heritage. I use data from Powell and Mitchell (2007), which indicates whether a country’s legal system has primarily features of a common rather than a civil, law system. Recent work suggests that domestic legal systems are strong determinants of international legal commitments (Powell and Mitchell 2007; Mitchell and Powell 2009; Simmons 2009; Mitchell and Powell 2011). Accounting for legal system type is especially important in the OAS since many of the most consistently democratic countries in the OAS including the United States, Canada, the Bahamas, Jamaica, Trinidad, Barbados, and Belize were once British colonies and inherited common law legal systems. This means that common law heritage is likely correlated with democracy, as well as acceptance of the Inter-American Court’s jurisdiction, so it is important to account for this in the analysis. I also include a measure of civil war since democratizing and weakly democratic states frequently experience political violence (see, for example, Muller 1985; Hegre et al. 2001), and one study has found that states that have joined the International Criminal Court experienced lower levels of internal violence (prior to joining) than those that have not yet joined (Chapman and Chaudoin 2012). For this purpose, I employ the Uppsala Conflict Data Program/Peace Research Institute, Oslo (UCDP/PRIO) measure of violent conflict, which uses a threshold of twenty-five annual battle deaths to indicate the presence of civil war (Gleditsch et al. 2002; Themnér and Wallensteen 2012). Models and Estimation Below I present estimates from models using data on member countries from the CE and the OAS. For analysis of the European Court, the data include forty-one member states for the years 1950–1998. The data for the Inter-American Court models include twenty-nine member states for the years 1969–2004. In both populations, not all states were sovereign/independent for the entire time period, so some states enter the data later than 1950/1969. Once a state accepts the relevant court’s jurisdiction, it drops out of the population. The full population for the CE data includes 773 country-years, while the OAS population is comprised of 686 country-years. For the ICCPR models, the data begin when the treaty became open for ratification (1966). Once a state ratifies, it drops out of the population. The OAS data for the ICCPR analysis consist of twenty-eight member states for a total of 442 country-years, and the CE data consist of thirty member states and 464 country-years. The smaller population sizes for the ICCPR analyses are due primarily to the fact that most CE/OAS states ratified the ICCPR more quickly than they joined their respective human rights court. I employ Bayesian, hierarchical logistic regression models to analyze acceptance of a regional human rights court’s jurisdiction/ratification of the ICCPR. Each model controls for the effects of civil conflict and common law legal system, though I limit the discussion below to the results regarding democracy. Since my goal is to estimate the effect of democracy and democratic transitions on delegation to human rights courts, it is only necessary to control for factors that are correlated with both democracy and delegation. Limiting the covariates to those that have been theoretically linked to democracy and delegation also helps avoid a proliferation of highly correlated control variables, which can complicate analysis (see Achen 2002; Schrodt 2014). The models include random effects for country and a time counter7 and so are effectively discrete time versions of duration models with random effects (frailties). With the inclusion of time-varying covariates, estimation of parametric, continuous time duration models or the nonparametric Cox model becomes complicated. This is especially true in a Bayesian framework which I adopt here. These models also make the estimation of country-specific effects (frailties) more difficult. Thus, I employ the binary time-series cross-sectional design commonly used in international relations, which accounts for temporal dependence by simply including a time counter in a binary response (logit or probit) regression model (see Beck, Katz, and Tucker 1998; Carter and Signorino 2010). I estimate twelve regression models for each organization (six for the relevant court and six for the ICCPR). To make things explicit, each model can be written as yit∼Bernoulli(pit),logit(pit)=αi+Di,tβ+Ci,tγ+Ti,tδ,αi∼Normal(μi,σ2),μi=α0+Liη, where D is a measure of democracy and/or a transformation of that measure, C is an indicator of conflict occurrence in country i at time t, T is a time counter, and Li indicates whether country i has a common law legal system. To test the hypotheses discussed above, I estimate a model that includes democracy and its square, as well as models that include the change in democracy from a previous year (in addition to control variables). To allow for the fact that a change in democracy may not immediately result in delegation to a human rights court or ratification of the ICCPR, I estimate models that include the change from the previous year ( Di,t−Di,t−1), as well as models that include the change from the previous two years etc., up to five years previous. The model with the squared democracy term allows for the possibility that the probability of delegation is related to democracy in a nonmonotonic fashion as expected by the hypothesis that regimes “in the middle” are more likely to delegate. This expectation would be supported by a negative coefficient for the squared democracy term. The change variables are employed to test whether increases in democracy are associated with delegation to the court, which would support the hypothesis that transitioning regimes are more likely to delegate. The time counters are included because the probability of acceptance may vary over time as a function of unmeasured factors, so including T will reduce bias in parameter estimates (see Carter and Signorino 2010). In the context of a human rights court, such temporal effects are plausible as perceptions of the court’s autonomy and power are likely to change over time and influence the behavior of potential member states (see, for example, Hawkins and Jacoby 2006). To account for unobserved heterogeneity in these two groups of states, country-specific intercepts, α, are included in each model. This is important since countries like the United States and Brazil, for example, are being included in the same analysis as countries like Costa Rica and Barbados. There is thus good reason to believe that each state’s baseline probability of acceptance is not the same and may vary due to unmeasured factors. For example, the United States is notoriously averse to human rights law for reasons that may have nothing to do with democracy or legal system heritage. α is modeled as a function of L, which is a vector measuring common law heritage, in addition to α0, the “grand mean” across all countries. The variance of the country-specific effects, σ2, is estimated from the data. All models were estimated via Markov chain Monte Carlo (MCMC) simulation. Standard estimation techniques for logit models (such as maximum likelihood estimation), and the hypothesis tests which typically accompany them, rely on the notion of repeated random samples drawn from a larger population. Data on CE/OAS member states are not drawn at random from a larger population from which one could draw additional samples. The populations of interest here are CE and OAS member states, and the available data for these populations cannot be considered a random sample (see, for example, Western and Jackman 1994). Since the measure of democracy is estimated from a statistical model, one must account for the uncertainty (measurement error) in this indicator when using it as a covariate. To accomplish this, ten values of the democracy score are drawn randomly from the estimated posterior distribution for each country-year in the data, and each model is estimated ten times. The results are then aggregated across the ten estimates. For each model, two Markov chains were run for at least 20K iterations,8 and the last 10K values from each chain were retained to summarize the posterior distributions of the model parameters. Convergence was assessed with visual diagnostics and a Gelman–Rubin diagnostic test (Gelman and Rubin 1992). Further technical model details can be found in the Appendix. Results Coefficient estimates for democracy, democracy squared, and changes in democracy from models using the CE data are displayed in Figure 1. Estimates from models using the OAS data are shown in Figure 2. The figures show the average estimate (across ten estimates) for each coefficient as a dot, and the horizontal lines connect the minimum lower bound of the 90 percent credible interval to the maximum upper bound. Figure 1. View largeDownload slide Coefficient Estimates for CE models. The average estimates (across ten estimates) are shown as dots, minimum lower/maximum upper bound of the ten 90 percent credible intervals are connected by the horizontal lines Figure 1. View largeDownload slide Coefficient Estimates for CE models. The average estimates (across ten estimates) are shown as dots, minimum lower/maximum upper bound of the ten 90 percent credible intervals are connected by the horizontal lines Figure 2. View largeDownload slide Coefficient Estimates for OAS Models. The average estimates (across ten estimates) are shown as dots, minimum lower/maximum upper bound of the ten 90 percent credible intervals are connected by the horizontal lines Figure 2. View largeDownload slide Coefficient Estimates for OAS Models. The average estimates (across ten estimates) are shown as dots, minimum lower/maximum upper bound of the ten 90 percent credible intervals are connected by the horizontal lines For the European Court, the model with the squared democracy term supports the hypothesis that weakly democratic states are more likely to delegate than are strongly democratic or strongly autocratic states. This is indicated in Figure 1 by the negative coefficient for the squared democracy term, which means that the relationship between democracy and acceptance of a human rights court’s jurisdiction is nonmonotonic in the way anticipated by the argument, with the probability of acceptance increasing at first and then decreasing once the value of democracy is sufficiently high. The minimum upper bound for this coefficient is above zero. This stands in contrast to the estimate for the squared democracy term in the ICCPR model, where five/ten upper bounds are above zero. For the European Court, the evidence concerning transitions to democracy is less definitive. For the European Court model, three of five estimates are positive, but the minimum lower bounds are all below zero, and two of the estimates are actually negative. All the estimates for ICCPR Models 2–6 are positive, but here too, the minimum lower bounds are all below zero. Though the negative estimates for European Court Models 2 and 3 may appear odd, this is due to the fact that there are no cases where a CE member state accepted the jurisdiction of the court after a dramatic increase in democracy that occurred over 1–2 years. But there are several cases of acceptance that occur after changes that take place over 3–5 years as the positive coefficients for European Court Models 4–6 indicate. I discuss several of these cases in more detail below. For the Inter-American Court, the coefficient estimates, shown in Figure 2, more strongly support the hypotheses. Turning first to the coefficient for the squared democracy term, the maximum upper bound of the 90 percent credible interval is above zero, but across ten estimates, this is the only upper bound that crosses zero. Further, even for this one conservative estimate, 92 percent of the posterior distribution for the democracy squared coefficient is below zero. Interestingly, this result is stronger for the Inter-American Court, indicated by the larger (more negative) coefficient for the democracy squared term in Model 1 for the Inter-American Court. Here again, the coefficient estimate for the democracy squared term in the ICCPR model is much smaller, and eight/ten upper bounds are above zero. In the OAS, the claim that increases in democracy increase the probability of delegation also receives strong support. In the case of the Inter-American Court, all the coefficient estimates for Models 2–6 are positive, and the minimum lower bounds for Models 3–6 do not contain zero. In the OAS, even short-term increases in democracy are associated with participation in the Inter-American Court. This is because many cases of delegation to the Inter-American Court occurred almost immediately after democratic transitions as discussed in more detail below. As with the CE, none of the ICCPR models for the OAS indicate that ratification becomes more likely with increases in democracy. In short, the estimates from these models provide strong support for the lock-in/signaling argument as it concerns high delegation institutions, but little evidence that governments use low delegation institutions as commitment devices. A closer examination of the data and model results provides further support for the lock-in/signaling hypotheses as it concerns delegation to human rights institutions. The left panel of Figure 3 plots the probability of acceptance of the European Court’s jurisdiction, from Model 1, against the latent democracy variable.9 The mean probability from each of ten models is shown as a black line, with 90 percent credible intervals shown as gray lines. The distribution of the democracy variable in the CE since 1950 is shown in light gray in the background. The black ticks on the x-axis show the values of democracy at which states that have accepted the court’s jurisdiction did so. The figure indicates that the probability of acceptance is essentially zero among states toward the lower end of the democracy scale, rises as democracy increases, and then declines slightly at high values of democracy. Many cases of acceptance are toward the middle of the democracy scale rather than the extremes, though there are cases toward the high end of the scale. The distribution of the democracy score in the CE is bimodal, with a mode of autocratic country-years toward the left side of the graph and a larger mode of democratic country-years toward the right. It is clear from Figure 3 that autocratic governments do not delegate to the court: there are zero cases of acceptances in the mode on the left side of the graph, and the probability from all models is essentially zero until the democracy score approaches zero. The mode on the left is populated by the communist governments of Central and Eastern Europe before the collapse of the Soviet Union, the right-wing authoritarian governments in Southern Europe that fell in the 1970s, and Turkey in the early 1980s. Central and Eastern European countries, former Soviet republics, and former Yugoslav republics account for most of the cases of acceptance in the middle of the distribution. Figure 3 indicates that there are also cases of countries with relatively high scores on the democracy scale that accepted the court’s jurisdiction relatively early: the United Kingdom, Belgium, the Netherlands, Luxembourg, Iceland, Denmark, Norway, and Austria all accepted the court’s jurisdiction relatively quickly and at high levels of democracy. And Sweden, Italy, France, and Switzerland, though they did not accept jurisdiction quickly, did so at high levels of democracy. This is why, for some models, the probability of acceptance does not decline sharply at high values of democracy. Figure 3. View largeDownload slide Effect of Democracy on Acceptance of the European Court’s Jurisdiction (prior to 1998) and Ratification of the ICCPR. The horizontal axes indicate the latent democracy score; the vertical axes indicate the probability of accepting the court’s jurisdiction/ratifying the ICCPR. The solid black lines show how the probability of acceptance/ratification changes with the democracy score, based on estimates from ten models. 90 percent credible intervals for all models are shown as gray lines. The distribution of the democracy score in the council of Europe for the relevant time period is shown in gray. The black tick marks indicate the values of the democracy score at which governments who have accepted the court’s jurisdiction/ratified the ICCPR did so Figure 3. View largeDownload slide Effect of Democracy on Acceptance of the European Court’s Jurisdiction (prior to 1998) and Ratification of the ICCPR. The horizontal axes indicate the latent democracy score; the vertical axes indicate the probability of accepting the court’s jurisdiction/ratifying the ICCPR. The solid black lines show how the probability of acceptance/ratification changes with the democracy score, based on estimates from ten models. 90 percent credible intervals for all models are shown as gray lines. The distribution of the democracy score in the council of Europe for the relevant time period is shown in gray. The black tick marks indicate the values of the democracy score at which governments who have accepted the court’s jurisdiction/ratified the ICCPR did so The right panel of Figure 3 shows the effect of democracy on ratification of the ICCPR and makes clear that ratification patterns for the ICCPR in the CE do not support the lock-in/signaling hypothesis. Most notably, the probability of ratification is relatively high even at low levels of democracy due to the fact that many governments in the CE ratified the ICCPR during periods of autocracy. This is made clear by the fact that the black tick marks in the right panel extend much further to the left than those in the left panel. CE member states that ratified the ICCPR during periods of autocracy include Poland, Hungary, Czechoslovakia, Bulgaria, Romania, and the Soviet Union. Though all these countries joined the European Court prior to 1998, they did so only after experiencing large increases in democracy. For the Inter-American Court, again, the evidence is even stronger. The left panel of Figure 4 displays the effect of democracy on acceptance of the Inter-American Court’s jurisdiction. The distribution of democracy in the OAS is also bimodal. Here too, no cases of acceptance are found in the autocratic mode, which contains the military dictatorships that appeared in Latin American during the 1970s, as well as Cuba. Figure 4 indicates that, across most models, the probability of accepting the Inter-American Court’s jurisdiction drops quickly toward the high end of the democracy scale. This is because there very few cases of acceptance at high values of democracy. The OAS countries that score the highest and participate in the court are Costa Rica, Trinidad, and Barbados, the last two of which took quite some time to accept jurisdiction (twenty-two and thirty-one years, respectively). It is apparent from the density plot in Figure 4 that many country-years receive relatively high scores, but most of these country-years are for states that have not accepted the court’s jurisdiction to date or had accepted it before they reached their highest level of democracy. The first group includes the United States, Canada, the Bahamas, Jamaica, and Belize. The second group includes Mexico, Panama, Suriname, Peru, Chile, Argentina, and Uruguay. Figure 4. View largeDownload slide Effect of Democracy on Acceptance of the Inter-American Court’s Jurisdiction and Ratification of the ICCPR. The horizontal axes indicate the latent democracy score; the vertical axes indicate the probability of accepting the court’s jurisdiction/ratifying the ICCPR. The solid black lines show how the probability of acceptance/ratification changes with the democracy score, based on estimates from ten models. 90 percent credible intervals for all models are shown as gray lines. The distribution of the democracy score in the OAS between for the relevant time period is shown in gray. The black tick marks indicate the values of the democracy score at which governments who have accepted the court’s jurisdiction/ratified the ICCPR did so Figure 4. View largeDownload slide Effect of Democracy on Acceptance of the Inter-American Court’s Jurisdiction and Ratification of the ICCPR. The horizontal axes indicate the latent democracy score; the vertical axes indicate the probability of accepting the court’s jurisdiction/ratifying the ICCPR. The solid black lines show how the probability of acceptance/ratification changes with the democracy score, based on estimates from ten models. 90 percent credible intervals for all models are shown as gray lines. The distribution of the democracy score in the OAS between for the relevant time period is shown in gray. The black tick marks indicate the values of the democracy score at which governments who have accepted the court’s jurisdiction/ratified the ICCPR did so The right panel of Figure 4 displays the effect of democracy on ratification of the ICCPR. Again the evidence fails to support the hypothesis. As in the CE, the probability of ratifying the ICCPR for OAS member states is nontrivial at low levels of democracy because several governments ratified the ICCPR during periods of autocratic rule. One of the most striking examples is El Salvador in 1979, a year in which a (mostly) military junta controlled the government and the country was in the midst of an escalating civil war. Another example is Panama in 1977, when head of state Omar Torrijos ratified the treaty five years after altering the constitution to remove many constraints on his authority. Torrijos would rule Panama for four more years until his death in 1981. Also, for most models, the probability of ratification continues to rise at high levels of democracy. This is because the most consistently democratic countries in the OAS (the United States, Canada, the Bahamas, Jamaica, and Belize) have all ratified the ICCPR. Substantive effects from the European Court and Inter-American Court models, which include changes in democracy from previous years, are show in Figures 5 and 6.10 For each change model, these figures show, on the vertical axis, the change in the probability of delegation that results from a one-point change in the democracy score (that is, the difference between the probability of delegation if there was a change from a previous year and the probability of delegation if there was no change from a previous year).11 Ten estimates for each change model are shown as dots, and 90 percent credible intervals are shown as vertical lines. For the European Court (Figure 5), the estimates are all positive when examining changes from four or five years previous, but even here nearly all of the credible intervals contain zero. The results from the Inter-American Court (Figure 6) strongly support the hypothesis: for the first change model, seven out of ten credible intervals for the effect of an increase in democracy do not contain zero, and for the other models, none of the intervals contain zero. Figure 5. View largeDownload slide Effect of a One-Point Change in the Democracy Score (from a previous year) on Acceptance of the European Court’s Jurisdiction (prior to 1998). The horizontal axes index the ten estimates; the vertical axes indicate the change in the probability of accepting the court’s jurisdiction that results from a one-point increase in the democracy score. The dots show the mean change, with 90 percent credible intervals shown as vertical lines Figure 5. View largeDownload slide Effect of a One-Point Change in the Democracy Score (from a previous year) on Acceptance of the European Court’s Jurisdiction (prior to 1998). The horizontal axes index the ten estimates; the vertical axes indicate the change in the probability of accepting the court’s jurisdiction that results from a one-point increase in the democracy score. The dots show the mean change, with 90 percent credible intervals shown as vertical lines Figure 6. View largeDownload slide Effect of a One-Point Change in the Democracy Score (from a previous year) on Acceptance of the Inter-American Court’s Jurisdiction. The horizontal axes index the ten estimates; the vertical axes indicate the change in the probability of accepting the court’s jurisdiction that results from a one-point increase in the democracy score. The dots show the mean change, with 90 percent credible intervals shown as vertical lines Figure 6. View largeDownload slide Effect of a One-Point Change in the Democracy Score (from a previous year) on Acceptance of the Inter-American Court’s Jurisdiction. The horizontal axes index the ten estimates; the vertical axes indicate the change in the probability of accepting the court’s jurisdiction that results from a one-point increase in the democracy score. The dots show the mean change, with 90 percent credible intervals shown as vertical lines One more piece of evidence relates to the timing of delegation. The argument under examination predicts that governments should delegate during the midst of, immediately after, or shortly after a transition to a higher level of democracy. The results presented above from the models that include changes in democracy support that claim, and a closer examination of individual cases in the data reveals that delegation to these two regional courts has tended to occur at precisely these times. Figure 7 shows the timing of acceptance for six countries in the CE data.12 Displayed on the y-axis in each graph is the democracy score (shown as a black line), with 95 percent credible intervals (shown as gray lines). The dashed vertical line indicates the year in which the government in question accepted the jurisdiction of the court. In each case, delegation coincides very nearly with a pronounced increase in democracy. Figure 7. View largeDownload slide Timing of European Court acceptance for Six CE countries. The horizontal axis in each plot indicates time; the vertical axis indicates the democracy score. The democracy score for each year is shown as a black line, with 95 percent credible intervals shown in gray. The year in which the government accepted the jurisdiction of the court is indicated by a vertical, dashed line Figure 7. View largeDownload slide Timing of European Court acceptance for Six CE countries. The horizontal axis in each plot indicates time; the vertical axis indicates the democracy score. The democracy score for each year is shown as a black line, with 95 percent credible intervals shown in gray. The year in which the government accepted the jurisdiction of the court is indicated by a vertical, dashed line The six countries shown in Figure 7 illustrate the argument well. Among the most exemplary cases in the CE are Spain, Portugal, and Greece, all countries that transitioned to democracy in the 1970s. Spain and Portugal had both experienced long periods of authoritarian rule prior to democracy. The Spanish government accepted the jurisdiction of the court in 1979, four years after the death of Franco, two years after the 1977 parliamentary elections (the first meaningful elections held in Spain since 1936—before the civil war), and one year after redrafting the constitution. Portugal accepted the court’s jurisdiction in 1978, four years after Marcelo Caetano (successor to Salazar) was ousted by the military and two years after elections that resulted in the first democratic government in Portugal since before Salazar. Greece also accepted the court’s jurisdiction after a transition to democracy, though in this case it took several years. Greece held parliamentary elections in 1974 after seven years of military rule and accepted the court’s jurisdiction in 1979. The cases of Poland, Czechoslovakia, and Bulgaria also illustrate the argument well. Poland and Czechoslovakia both accepted the court’s jurisdiction within two years of their first parliamentary elections after the end of communism. These cases are also notable because both countries had newly elected presidents that were prominent pro-democracy dissidents under their respective communist regimes: Lech Walesa in Poland and Vaclav Havel in Czechoslovakia. Both presidents had well-established policy preferences that favored democracy and faced uncertain political environments, which make the lock-in/signaling hypothesis very plausible. The case of Bulgaria is also illustrative. The Bulgarian Communist Party won the election of 1990, held to form a constitutional assembly. Interestingly, the government accepted the jurisdiction of the court two years later, only after the country’s second noncommunist (and first openly anticommunist) prime minister took office. There are other former communist countries that also follow this pattern, most notably Hungary, Albania, and Romania (see Appendix). Also, several former communist countries delegated to the court at middling values of democracy, though their level of democracy at the time of acceptance was not increasing. However, many of these cases occur after 1994, when the CE was pushing to make acceptance of the court’s jurisdiction mandatory. Thus, while the evidence supports the hypothesis, it is more ambiguous than it would be if Protocol 11 to the European Convention had never been introduced. In the OAS, one can also find a clear record of decisions to join the Inter-American Court following closely after democratic transitions, and here delegation cannot be attributed as easily to pressure from within the institution itself. The timing of acceptance for six OAS countries is shown in Figure 8. These cases also illustrate the argument quite well, with acceptance sometimes occurring immediately after a transition to a higher level of democracy. Guatemala, for example, accepted the jurisdiction of the court in March of 1987, one year after Vinicio Cerezo, the country’s first democratically elected president in twenty years, assumed office. Cerezo was struggling to improve human rights conditions in the face of strong resistance from the military, and his government accepted the jurisdiction of the court only months before establishing the National Commission for Reconciliation, a policy decision that also signals a willingness to respect human rights. Honduras accepted the jurisdiction of the court in 1981, the same year that saw the installation of their first democratically elected president after roughly twenty years of military rule. In the case of Panama, the government submitted to the court’s jurisdiction in 1990, shortly after the US government deposed Noriega and only five months after the newly elected president took office. In Chile, the first democratically elected government after Pinochet accepted the jurisdiction of the court five months after the new president assumed office. In Argentina, the jurisdiction of the court was accepted in September of 1984. This was roughly nine months after the installation of the first democratic president to follow the military government that had ruled the country since 1976 and nine months after the establishment of the National Commission on the Disappearances of Persons. Finally, the Uruguayan government also accepted the court’s jurisdiction shortly after a notable democratic transition. Elections were held in Uruguay in late 1984 after roughly tweleve years of military rule. The new president assumed office in March of 1985, and the government accepted the Inter-American Court’s jurisdiction about seven weeks later. There are several other countries in the OAS that also joined following transitions, the most pronounced cases being Paraguay, Peru, Ecuador, Suriname, Nicaragua, El Salvador, and Haiti (see Appendix). Figure 8. View largeDownload slide Timing of Inter-American Court Acceptance for Six OAS Countries. The horizontal axis in each plot indicates time; the vertical axis indicates the democracy score. The democracy score for each year is shown as a black line, with 95 percent credible intervals shown in gray. The year in which the government accepted the jurisdiction of the court is indicated by a vertical, dashed line Figure 8. View largeDownload slide Timing of Inter-American Court Acceptance for Six OAS Countries. The horizontal axis in each plot indicates time; the vertical axis indicates the democracy score. The democracy score for each year is shown as a black line, with 95 percent credible intervals shown in gray. The year in which the government accepted the jurisdiction of the court is indicated by a vertical, dashed line In summary, the evidence concerning participation in the human rights courts of the CE and the OAS strongly supports the lock-in/signaling argument. The argument seems to explain well delegation in both of these regimes, and the evidence is stronger in the OAS than in the CE. In contrast, there is relatively little evidence to suggest that ratification of the ICCPR follows the pattern predicted by the theory, which suggests that governments view more legalized human rights institutions as better commitment devices and that the theory is best tested by examining such institutions. Discussion/Conclusion From the standpoint of most theories of international cooperation, the proliferation of human rights agreements is one of the more puzzling developments in international law. Some empirical work examines ratification for UN human rights treaties, but very little work examines delegation to strong human rights institutions. This study builds on the work of Moravcsik (2000), Mansfield and Pevehouse (2006, 2008), and Hafner-Burton, Pevehouse, and Mansfield (2015), who argue that states may create/join human rights institutions in order to prevent future leaders from subverting democracy and to signal to the public their commitment to democratic institutions and values. The empirical analysis presented here provides stronger evidence for these claims than previous studies for the simple reason that previous tests to not examine delegation to human rights institutions. If states participate in institutions to signal their intent to adhere to liberal democratic values, then this implies that delegation to human rights institutions specifically should be attractive, not support for the creation of an institution to which one could later delegate and not membership in international institutions broadly. Leveraging the European and Inter-American Courts, I find that the evidence weighs heavily in favor of the theories advanced by Moravcsik (2000) and Mansfield and Pevehouse (2006). In both organizations, delegation to the human rights courts has tended to occur among weakly democratic and democratizing states rather than strongly autocratic or strongly democratic states. Further, many cases of delegation occurred during, immediately after, or very shortly after transitions to higher levels of democracy. The evidence is even stronger in the OAS due to the fact that the most consistently democratic countries in the region have not yet accepted the court’s jurisdiction and that acceptance tended to happen more quickly after democratic transitions in the OAS than in the CE. This is strong evidence in favor of the claim that there is a causal relationship between domestic political institutions and delegation to human rights regimes. Delegation to a human rights court seems to be something a leader in a new or weak democracy can do to signal to the public her intent to respect human rights in the future. Joining a court may be one of several ways that leaders can signal a commitment to rights, such as the establishment of truth/reconciliation commissions as in the cases of Guatemala and Argentina. The fact that leaders have, in some cases, joined human rights courts while simultaneously establishing human rights commissions suggests that there are other kinds of behavior that might be relevant to the theory. The establishment of commissions and more permanent national human rights institutions, like ombudsman offices, may be worth examining in future studies.13 These results also indicate that the theory may have broader applicability than was originally anticipated. Moravcsik (2000, 246) claims that “the determinants of the evolution of human rights regimes are unlikely to be identical to the determinants of their founding.” If the evolution of the regime includes patterns of participation subsequent to its creation, then this claim appears to be incorrect. Just as governments in postwar Europe facing uncertain political environments were keen to create a strong court in the first place, governments with democratic aspirations that faced similarly uncertain environments long after the regime’s founding were eager to delegate authority to the court.14 The same is true of the Inter-American regime, where the decision to delegate seems to be explained by domestic political concerns about consolidating democracy, and where the evidence in favor of this theoretical argument is even stronger. These results are encouraging from a normative perspective. The literature on international human rights law has tended to report bad news. Specifically, most research examining compliance with UN treaties has found that ratification has little unconditional effect on the most repressive kinds of government behavior (Hathaway 2002; Hafner-Burton and Tsutsui 2005; Neumayer 2005; Hafner-Burton and Tsutsui 2007; Hill 2010; Lupu 2013). The analysis above suggests that “false positives,” governments that make legal commitments that they have no intention of upholding (Simmons 2009), are more likely for human rights treaties than for courts, which is one potential explanation for the nonfindings in the compliance literature. Too, studies of treaty ratification have found that states are avoiding obligation where it would entail genuine constraints (Hathaway 2007; Powell and Staton 2009), which also potentially explains these nonfindings. Delegation to courts follows a different pattern. In the case of the both the CE and the OAS, states have tended to submit to a court’s authority when it could conceivably have some impact on behavior, namely when democratic institutions are new or fragile. These are precisely the conditions under which international human rights law is theorized to have a meaningful impact on government behavior (Simmons 2009), which suggests that studying the effect of human rights court rulings conditional on domestic institutions would be useful. The findings above indicate that governments may be making sincere commitments when they delegate to human rights courts. At the very least, delegation to regional human rights courts seems to be a more sincere act than ratification of a UN treaty, suggesting more generally that “decoupling” of legal commitment and practices is less likely in more highly legalized institutions (Cole 2012). Additionally, recent work has shown that the decisions of human rights courts do impact government policy under certain circumstances (Helfer and Voeten 2014; Haglund 2014). Together with the results presented here, this is good news and suggests that further research on the reasons for delegation to, and compliance with, international human rights courts would be fruitful. Daniel W. Hill Jr. received his PhD in Political Science from Florida State University in 2012 and has been an Assistant Professor in the Department of International Affairs at the University of Georgia since 2012. His current main interests are the conceptualization and measurement of human rights practices, quantitative studies of compliance with international human rights law, and statistical forecasts for violent conflict. Footnotes 1But see Fariss (2014), who finds that this result is partially an artifact of data on human rights abuses. The standards that human rights monitors use have become more stringent over the years, so acts that would not be characterized as abuse in the past are characterized that way in the present, which produces artificial stagnation or decline in some indicators. 2Individual petition is an international legal principle that gives individuals the right to initiate judicial proceedings against states. 3See also Cole (2009), who finds that democratization is not correlated with the adoption of treaty protocols that delegate more authority to treaty monitoring bodies. Adopting optional protocols is not equivalent to accepting a court’s jurisdiction, but this finding is inconsistent with Moravcsik’s argument. 4A better approach to a simple additive scale that assumes a weighting scheme for its components would be a measurement model along the lines of that used to create the indicator of democracy used in this study (discussed below) (Pemstein, Meserve, and Melton, 2010). 5Notably, the HRC does not have the authority to issue public, legally binding decisions about compliance with the provisions of the ICCPR, relies largely on self-reporting by member states, and individuals have no standing to bring complaints against their government unless their government has ratified the Second Optional Protocol to the ICCPR. For an examination of the effectiveness of the HRC and similar UN treaty-based monitoring bodies in improving human rights conditions, see Krommendijk (2015). 6Treier and Jackman (2008) demonstrate that results from regression analyses using the Polity scale and its squared value do not hold when the measurement error is accounted for. 7Although it is common to include second- and third-order polynomials of time in such models, including polynomials did not improve the fit of the models presented below and caused complications during estimation (nonconvergence in some cases). Since the models allow for a monotonically increasing or decreasing hazard rate, they are analogous to Weibull models (with frailties). 8For some models, it was necessary to extend the MCMC run to as many as 150K iterations. 9When calculating the values shown in the figures, the other covariates were set to the following values: 0 for conflict, 0 for common law, and 20 for the time counter. 10As with the two previous figures, conflict was set to 0, common law to 0, and the time counter to 20. 11A one-point change is large, but in each organization, there are many cases where a country experienced a change this large over 2–5 years, and in some cases even over a single year, such as Argentina 1982–1983 and Spain 1976-1977. 12Plots for the other CE countries can be found in the Appendix. 13Recent data collection efforts by Conrad, Demeritt, and Moore (2014) have made such studies possible. See Welch (Forthcoming) for an analysis of the effect of national human rights institution (NHRIs) on human rights practices. 14See also Voeten (2007), who finds that governments with EU membership aspirations are more likely to appoint activist judges to the court in order to signal a commitment to democracy and human rights. Appendices A Model Details Each model estimated can be written as yit∼Bernoulli(pit)Logit(pit)=αi+Di,tβ+Ci,tγ+Ti,tδαi∼Normal(μi,σ2)μi=α0+Liη. To estimate the model’s parameters via MCMC, prior distributions must be specified for the country-year level coefficients β, γ, and δ, as well as the intercept (α0) and coefficient η in the hierarchical model for the country-specific intercepts, and the variance of these intercepts, σ2. β, γ, δ, α0, and η were all assigned normal priors with means 0 and variances of 10. σ2 was assigned a gamma prior with a scale parameter of 1 and a shape parameter of 10. Convergence was assessed with a Gelman–Rubin diagnostic (upper confidence intervals close to 1) and visual diagnostics (posterior densities unimodal and roughly normal). B Time-Series Plots for Other Countries Displayed below in Figures 9–12 are time-series plots for the democracy score (Pemstein Meserve and Melton, 2010) for each country in the sample. The black line tracks the point estimate for the democracy score, with 90 percent credible intervals shown in gray. Bear in mind that these intervals represent the uncertainty in these estimates. For any two time points where there is a considerable amount of overlap in these intervals, the levels of democracy at these points in time are not statistically distinguishable from one another. For example, while the graph for the United States shows an increase in democracy shortly after 1990, the fact that the intervals at this point in time overlap with the intervals at every other point indicates that an analyst would be wrong to judge the US level of democracy as being higher at this point than at any point in the past or the future. Figure 9. View largeDownload slide Timing of European Court acceptance for other CE countries Figure 9. View largeDownload slide Timing of European Court acceptance for other CE countries Figure 10. View largeDownload slide Timing of European Court acceptance for other CE countries Figure 10. View largeDownload slide Timing of European Court acceptance for other CE countries Figure 11. View largeDownload slide Timing of Inter-American Court acceptance for other OAS countries Figure 11. View largeDownload slide Timing of Inter-American Court acceptance for other OAS countries Figure 12. View largeDownload slide Timing of Inter-American Court acceptance for other OAS countries Figure 12. View largeDownload slide Timing of Inter-American Court acceptance for other OAS countries References Abbott Kenneth W. , Snidal Duncan . 2000 . “ Hard and Soft Law in International Governance .” International Organization 54 : 421 – 56 . Abbott Kenneth W. , Keohane Robert O. , Moravcsik Andrew , Slaughter Anne Marie , Snidal Duncan . 2000 . “ The Concept of Legalization .” International Organization 54 : 401 – 19 . Achen Christopher H. 2002 . “ Toward a New Political Methodology: Microfoundations and ART .” Annual Review of Political Science 5 : 423 – 50 . Alter Karen . 2006 . Delegation to International Courts and the Limits of Recontracting Power. In Delegation and Agency in International Organizations , edited by Hawkins Darren G. , Lake David A. , Nielson Daniel L. , Tierney Michael J. , 312 – 38 . New York : Cambridge University Press . Alter Karen . 2008 . “ Delegating to International Courts: Self-Binding vs. Other-Binding Delegation .” Law and Contemporary Problems 71 : 37 – 76 . Beck Neal , Katz Jonathan N. , Tucker Richard . 1998 . “ Taking Time Seriously: Time-Series-Cross-Section Analysis with a Binary Dependent Variable .” American Journal of Political Science 42 : 1260 – 88 . Bernhard William , Broz Lawrence , Clark William Roberts . 2002 . “ The Political Economy of Monetary Institutions .” International Organization 56 : 693 – 723 . Bradley Curtis A. , Kelley Judith G. . 2008 . “ The Concept of International Delegation .” Law and Contemporary Problems 71 : 1 – 36 . Carter David B. , Signorino Curtis S. . 2010 . “ Back to the Future: Modeling Time Dependence in Binary Data .” Political Analysis 18 : 271. Chapman Terrence L. , Chaudoin Stephen . 2012 . “ Ratification Patterns and the International Criminal Court .” International Studies Quarterly 57 : 400 – 9 . Cole Wade M. 2005 . “ Sovereignty Relinquished? Explaining Commitment to the International Human Rights Covenants, 1966–1999 .” American Sociological Review 70 : 472 – 95 . Cole Wade M. . 2009 . “ Hard and Soft Commitments to Human Rights Treaties, 1966–2000 .” Sociological Forum 24 : 563 – 88 . Cole Wade M. . 2012 . “ Human Rights as Myth and Ceremony? Reevaluating the Effectiveness of Human Rights Treaties, 1981–2007 .” American Journal of Sociology 117 : 1131 – 71 . Conrad Courtenay R. , Demeritt Jaqueline H. R. , Moore Will H. . 2014 . “The National Human Rights Institutions Data Project.” Accessed April 6, 2016. http://nhridata.weebly.com/data.html. Fariss Christopher J. 2014 . “ Respect for Human Rights Has Improved Over Time: Modeling the Changing Standard of Accountability .” American Political Science Review 108 : 297 – 318 . Gelman Andrew , Rubin Donald B. . 1992 . “ Inference from Iterative Simulation Using Multiple Sequences .” Statistical Science 7 : 457 – 72 . Gleditsch Kristian Skrede , Ward Michael D. . 1997 . “ Double Take: A Reexamination of Democracy and Autocracy in Modern Polities .” Journal of Conflict Resolution 41 : 361 – 83 . Gleditsch Nils Petter , Wallensteen Peter , Eriksson Mikael , Sollenberg Margareta , Strand Håvard . 2002 . “ Armed conflict 1946–2001: A New Dataset .” Journal of Peace Research 39 : 615 – 37 . Goodliffe Jay , Hawkins Darren G. . 2006 . “ Explaining Commitment: States and the Convention against Torture .” Journal of Politics 68 : 371 . Hafner-Burton Emilie M. , Tsutsui Kiyoteru . 2005 . “ Human Rights in a Globalizing World: The Paradox of Empty Promises .” American Journal of Sociology 110 : 1373 – 411 . Hafner-Burton Emilie M. , Tsutsui Kiyoteru . 2007 . “ Justice Lost! The Failure of International Human Rights Law to Matter Where Needed Most .” Journal of Peace Research 44 : 407 – 25 . Hafner-Burton Emilie , Pevehouse Jon , Mansfield Edward . 2015 . “ Human Rights Institutions, Sovereignty Costs, and Democratization .” British Journal of Political Science 45 : 1 – 25 . Haglund Jilliene N. D. 2014 . “Domestic Implementation of Supranational Court Decisions: The Role of Domestic Politics in Respect for Human Rights.” Paper presented at the Annual Meeting of the International Studies Association, Toronto, Canada. Hathaway Oona A. 2002 . “ Do Human Rights Treaties Make a Difference? ” Yale Law Journal 111 : 1935 – 2042 . Hathaway Oona A. . 2007 . “ Why Do Countries Commit to Human Rights Treaties? ” Journal of Conflict Resolution 51 : 588 – 621 . Hathaway Oona A. . 2008 . “ International Delegation and State Sovereignty .” Law and Contemporary Problems 71 : 115 . Hawkins Darren G. , Lake David A. , Nielson Daniel L. , Tierney Michael J. . 2006 . Delegation and Agency in International Organizations . New York : Cambridge University Press . Hawkins Darren G. , Jacoby Wade . 2006 . “How Agents Matter.” In Delegation and Agency in International Organizations , edited by Hawkins Darren G. , Lake David A. , Nielson Daniel L. , Tierney Michael J. , 199 – 228 . New York : Cambridge University Press . Hegre Håvard , Ellingsen Tanja , Gates Scott , Gleditsch Nils Petter . 2001 . “ Toward a Democratic Civil Peace? Democracy, Political Change, and Civil War, 1816–1992 .” American Political Science Review 95 : 33 – 48 . Helfer Laurence , Voeten Erik . 2014 . “ International Courts as Agents of Legal Change: Evidence from LGBT Rights in Europe .” International Organization 68 : 77 – 110 . Hill Daniel W. 2010 . “ Estimating the Effects of Human Rights Treaties on State Behavior .” The Journal of Politics 72 : 1161 – 74 . Koremenos Barbara . 2007 . “ If Only Half of International Agreements Have Dispute Resolution Provisions, Which Half Needs Explaining? ” The Journal of Legal Studies 36 : 189 – 212 . Krommendijk Jasper . 2015 . “ The Domestic Effectiveness of International Human Rights Monitoring in Established Democracies. The Case of the UN Human Rights Treaty Bodies .” Review of International Organizations 10 : 489 – 512 . Lohmann Susanne . 1992 . “ Optimal Commitment in Monetary Policy: Credibility Versus Flexibility .” The American Economic Review 82 : 273 – 86 . Lupu Yonaton . 2013 . “ Best Evidence: The Role of Information in Domestic Judicial Enforcement of International Human Rights Agreements .” International Organization 67 : 469 – 503 . Mansfield Edward D. , Pevehouse Jon C. . 2006 . “ Democratization and International Organizations .” International Organization 60 : 137 – 67 . Mansfield Edward D. , Pevehouse Jon C. . 2008 . “ Democratization and the Varieties of International Organizations .” Journal of Conflict Resolution 52 : 269 – 94 . Meyer John W. , Boli John , Thomas George M. , Ramirez Francisco O. . 1997 . “ World Society and the Nation-State .” American Journal of Sociology 103 : 144 – 81 . Mitchell Sarah McLaughlin , Powell Emilia Justyna . 2009 . “ Legal Systems and Variance in the Design of Commitments to the International Court of Justice .” Conflict Management and Peace Science 26 : 164 – 90 . Mitchell Sarah McLaughlin , Powell Emilia Justyna . 2011 . Domestic Law Goes Global: Legal Traditions and International Courts . New York : Cambridge University Press . Moravcsik Andrew . 2000 . “ The Origins of Human Rights Regimes: Democratic Delegation in Postwar Europe .” International Organization 54 : 217 – 52 . Muller Edward N. 1985 . “ Income Inequality, Regime Repressiveness, and Political Violence .” American Sociological Review 50 : 47 – 61 . Neumayer Eric . 2005 . “ Do International Human Rights Treaties Improve Respect for Human Rights? ” Journal of Conflict Resolution 49 : 925 – 53 . North Douglass C. , Weingast Barry R. . 1989 . “ Constitutions and Commitment: The Evolution of Institutions Governing Public Choice in Seventeenth-Century England .” The Journal of Economic History 49 : 803 – 32 . Pemstein Daniel , Meserve Stephen A. , Melton James . 2010 . “ Democratic Compromise: A Latent Variable Analysis of Ten Measures of Regime Type .” Political Analysis 18 : 426 – 49 . Pevehouse Jon , Nordstrom Timothy , Warnke Kevin . 2004 . “ The Correlates of War International Governmental Organizations Data Version 2.0 .” Conflict Management and Peace Science 21 : 101 – 19 . Powell Emilia Justyna. , Staton Jeffrey K. . 2009 . “ Domestic Judicial Institutions and Human Rights Treaty Violation .” International Studies Quarterly 53 : 149 – 74 . Powell Emilia Justyna , Mitchell Sarah McLaughlin . 2007 . “ The International Court of Justice and the World’s Three Legal Systems .” Journal of Politics 69 : 397 – 415 . Quinn Kevin M. 2004 . “ Bayesian Factor Analysis for Mixed Ordinal and Continuous Responses .” Political Analysis 12 : 338 – 53 . Risse Thomas , Ropp Stephen C. , Sikkink Kathryn , eds. 1999 . The Power of Human Rights: International Norms and Domestic Change . New York : Cambridge University Press . Rogoff Kenneth . 1985 . “ The Optimal Degree of Commitment to an Intermediate Monetary Target .” The Quarterly Journal of Economics 100 : 1169 – 89 . Schrodt Philip A. 2014 . “ Seven Deadly Sins of Contemporary Quantitative Political Analysis .” Journal of Peace Research 51 : 287 – 300 . Simmons Beth A. 2009 . Mobilizing for Human Rights: International Law in Domestic Politics . New York : Cambridge University Press . Simmons Beth A. , Danner Allison . 2010 . “ Credible Commitments and the International Criminal Court .” International Organization 64 ( 2 ): 225 – 56 . Themnér Lotta , Wallensteen Peter . 2012 . “ Armed Conflicts, 1946–2011 .” Journal of Peace Research 49 : 565 – 75 . Treier Shawn , Jackman Simon . 2008 . “ Democracy as a Latent Variable .” American Journal of Political Science 52 : 201 – 17 . Voeten Erik . 2007 . “ The Politics of International Judicial Appointments: Evidence from the European Court of Human Rights .” International Organization 61 : 669 – 701 . Welch Ryan M . Forthcoming . “ National Human Rights Institutions: Domestic Implementation of International Human Rights Law .” Journal of Human Rights . Doi: 10.1080/14754835.2015.1103166. Western Bruce , Jackman Simon . 1994 . “ Bayesian Inference in Comparative Research .” American Political Science Review 88 : 412 – 23 . © The Author (2016). Published by Oxford University Press on behalf of the International Studies Association. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

Foreign Policy AnalysisOxford University Press

Published: May 3, 2016

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off