508591

earch-article2014

EJT0010.1177/1354066113508591European Journal of International RelationsHillebrecht

Article

The power of human rights tribunals: Compliance with the European Court of Human Rights and domestic policy change

EJIR European Journal of International Relations 2014, Vol. 20(4) 1100­–1123 © The Author(s) 2014 Reprints and permissions: sagepub.co.uk/journalsPermissions.nav DOI: 10.1177/1354066113508591 ejt.sagepub.com

Courtney Hillebrecht University of Nebraska-Lincoln, USA

Abstract When international human rights tribunals like the European Court of Human Rights find states responsible for human rights abuses, they ask governments to pay reparation to the victims, engage in symbolic measures, and enact the policy changes necessary to ensure that the violations do not recur. This article considers the conditions under which states comply with these rulings, especially when the tribunals are unable and often unwilling to provide strict enforcement. This article extends current theories about the domestic politics of compliance with international human rights law to the case of the European Court of Human Rights. This article analyzes a new, hand-coded data set on states’ compliance with over 1000 discrete obligations handed down by the European Court of Human Rights that ask states to change their human rights policies. The results of these analyses suggest that robust domestic institutions, particularly executive constraints, are the key to compliance with the European Court of Human Rights. When domestic institutions enforce the Court’s rulings, the results can be significant changes in states’ human rights policies and practices. Keywords Compliance, domestic politics, European Court of Human Rights, human rights, international law, supranational courts

Corresponding author: Courtney Hillebrecht, Assistant Professor of Political Science, University of Nebraska-Lincoln, 526 Oldfather Hall, Lincoln, NE 68588, USA. Email: [email protected]

Hillebrecht

1101

1. Introduction In 2011, over 64,500 European constituents lodged complaints against their governments at the European Court of Human Rights (ECtHR) in Strasbourg, France. These citizens sought international legal relief for domestic violations of human rights. In its 60-year history, the ECtHR has handed down over 14,000 judgments (European Court of Human Rights, 2012).1 The ECtHR, like its counterparts in Africa and the Americas, represents an important avenue for redress, forcing states to pay reparation to victims, hold perpetrators accountable, and change their human rights policies and practices. For these regional courts to provide justice for victims of human rights abuse and prevent violations from recurring, however, member states must comply with their rulings. The ECtHR has had varied success in facilitating state compliance. For example, in 2008, Russia paid 3.7 million euros in reparation as a result of rulings handed down by the ECtHR. That same year, the United Kingdom changed its anti-terrorism laws following judgments issued by the ECtHR (Anon, 2001, 2009). Yet, for every example of compliance, there are others of non-compliance: Germany has failed to enact the legislative changes necessary to reduce delays in civil proceedings and Norway has not responded to the ECtHR’s calls to provide media outlets for small political parties.2 This article asks: under what circumstances do states comply with the ECtHR’s rulings? The ECtHR has a broad membership and heavy caseload, offering a wealth of data on which to test a variety of explanations for compliance with international law. This article focuses specifically on measures of non-repetition, which ask states to change their human rights policies and practices, going well beyond paying reparation to victims or issuing apologies. Measures of non-repetition can involve repealing old laws, drafting new legislation, or changing administrative practices. For example, in the case of M. v. Germany (Application No. 19359/04), the ECtHR found the state responsible for the retroactive extension of preventive detention for perpetrators of serious crimes. Accordingly, the Court asked Germany to enact measures of non-repetition necessary to prevent similar violations. In response, Germany ordered a judicial review of all similar cases; the Federal Constitutional Court declared unconstitutional the mandate that allowed the extension of preventive detention and ordered new legislation on the issue; and the state undertook comprehensive statutory reform. In some instances, such as these, compliance with measures of non-repetition can precipitate significant policy change (Council of Europe Execution of Judgments, n.d.). The results of the statistical analyses below show that when domestic institutions are able and willing to comply with the ECtHR’s rulings, the rulings can usher in significant changes in states’ human rights policies and practices. This article makes two contributions to the current literature on compliance with international human rights law. First, this article builds on existing theories regarding the domestic politics of compliance with international law and applies them to the case of the ECtHR. In particular, this article highlights the importance of domestic institutions, namely, the executive, independent judiciaries, legislatures, and civil society actors. Although the ECtHR provides significantly more oversight and enforcement

1102

European Journal of International Relations 20(4)

than other human rights instruments, this research suggests that the burden of inducing de facto compliance falls to domestic actors. The statistical analyses that follow find that domestic institutions, particularly those that constrain the executive, have a positive effect on compliance, while alternative factors, such as states’ human rights practices, international socialization, or coercion, are unable to explain compliance with the ECtHR. Second, this article examines the patterns of states’ compliance with the ECtHR using a new, hand-coded data set on states’ compliance records. These data disaggregate existing, qualitative information on compliance and look at states’ compliance with 1184 discrete compliance obligations requiring states to change their domestic policies and/or practices. Each ruling handed down by the ECtHR can contain anything from two to 20 discrete obligations and states pick and choose the obligations within each ruling with which they want to comply. This approach of evaluating compliance at the obligation level is most closely aligned with states’ practice. In doing so, it provides a comprehensive and nuanced view of compliance with the ECtHR. This article is meant to be a high-level overview of compliance with the ECtHR’s rulings on measures of non-repetition. As such, a few caveats are in order. First, despite some commonalities in the compliance process, individual leaders’ priorities and country- and case-specific characteristics matter. Further, this article examines states’ de jure compliance with discrete non-repetition measures, not with the spirit of human rights norms. The very purpose of measures of non-repetition is to change states’ laws and practices, thus improving the level of human rights protection. Compliance with measures of non-repetition means that states are changing their human rights laws to the satisfaction of the ECtHR and the Committee of Ministers of the Council of Europe. That said, states might not always act in accordance with their own laws and policies. This article aims only to explain the conditions under which states comply with the Court’s rulings on measures of non-repetition, not to understand if and when states’ behavior subsequently diverges from their reformed policies. Finally, this article only looks at the process of compliance after a ruling has been handed down and does not seek to address why certain human rights cases get taken to the ECtHR in the first place. As governments do not get to dictate which cases the tribunals hear or when, they must deal with adverse judgments as they are handed down. This article proceeds in five parts. Section 2 describes the tribunals, while Section 3 tackles the questions of when and how states comply with the tribunals’ rulings. Section 4 describes a new data set on compliance with the ECtHR. Section 5 uses this data set to test the relationship between compliance and domestic political institutions, as well as competing hypotheses. Section 6 concludes.

2.  The ECtHR Founded by the Council of Europe in the aftermath of the human atrocities of World War II, the ECtHR was designed to provide recourse for individuals when states failed to respect their human rights obligations. Policymakers and activists now also look to the way that states respond to adverse judgments at the ECtHR as barometers of their commitment to human rights. As the chair of the Russian Justice Initiative, a non-governmental

Hillebrecht

1103

organization (NGO) that deals with the ECtHR, said with regard to a case at the ECtHR involving the disappearance of two Chechens, ‘The Russian government’s commitment to human rights and the European Convention will now be apparent by the way they respond to these judgments’ (Human Rights Watch, 2009). Despite ongoing contestation and critique, the ECtHR is at the forefront of the international protection of human rights, and understanding its effect on human rights requires examining how and when states comply with its rulings (Goldhaber, 2007). Unlike the International Criminal Court or the International Criminal Tribunals for Rwanda and the Former Yugoslavia, which seek to determine individual criminal responsibility for massive human rights abuse, the ECtHR adjudicates on state practice. States are the defendants at the ECtHR and are held accountable for failing to uphold their international human rights commitments at home. One of the hallmarks of the ECtHR is that it allows individuals to bring petitions alleging human rights abuses against member states (Allain, 2000; Evans, 1967; Pasqualucci, 2003; Ryssdal and Martens, 1996). Indeed, over 99% of the cases brought to the ECtHR originate from individual constituents. The individual petitioning mechanism invests individual plaintiffs and the Court with a fair amount of power: the ability to dictate to states changes in their human rights policies and practices. Moreover, the Committee of Ministers of the Council of Europe monitors states’ compliance and provides a forum in which states can pressure each other to comply with the ECtHR’s rulings. Yet, the actual process of complying with the Court’s rulings happens almost strictly on the domestic stage. The issue of compliance, especially with measures of non-repetition, is a critical one, both for states and for the Court. Measures of non-repetition are mandates that require states to change their policies and their practices to avoid the repetition of violations. Yet, for obvious reasons — the difficulty of actually changing policies and practices — states often struggle to comply with these mandates. This means that the ECtHR sees many repeat violations. The ECtHR, which receives approximately 50,000 petitions alleging abuse each year, cannot manage its caseload and hear cases in a timely manner (Registry of the European Court of Human Rights, 2010). This problem of an overwhelming number of cases is driven largely by non-compliance with measures of non-repetition. In 2011, nearly 80% of new cases were considered repeat cases, meaning that they deal with issues on which the Court has already ruled (Committee of Ministers of the Council of Europe, 2011). While the ECtHR and the Council of Europe have implemented a new mechanism, Protocol 14, to more effectively screen these repeat cases, Protocol 14 treats the symptoms, not the cause of the repeat cases (Buyse, 2010; Committee of Ministers of the Council of Europe, 2010; Council of Europe, 2004). The real losers in this cycle of non-compliance with non-repetition measures are the victims. When states fail to stop human rights violations and make the structural changes needed to ensure that such violations do not recur, it is victims, past and future, who suffer the most. It stands, then, that the key to the Court’s effectiveness in protecting human rights is in states’ compliance with their rulings, particularly with measures of non-repetition.

1104

European Journal of International Relations 20(4)

3.  Compliance with international law: The ECtHR in theoretical context The literature on international relations and international law has provided a number of explanations for compliance with international law, including: (1) explanations that focus on the endogenous or epiphenomenal nature of international law; (2) theories that understand compliance as the result of coercion or enforcement from the top down; (3) theories that point to the importance of international law in shaping new norms and socializing states into these norms; and (4) explanations that point to domestic politics (Hafner-Burton, 2012; Simmons, 1998, 2010). While much of the literature on compliance with international law, and international human rights law in particular, focuses on compliance with international treaties, these same theoretical expectations can be applied to the case of compliance with the ECtHR. Realist and neo-realist scholars have long attributed compliance with international law to the codification of existing practice. Downplaying the role of international law, these scholars argue that states only join international institutions when they are ex ante compliant. In other words, states comply with international law because of an inherent selection bias: they only enter into international agreements with which they know they will comply (Downs et al., 1996; Fearon, 1998; Goldsmith and Posner, 2005).3 If we were to extend this theory to states that had already accepted the jurisdiction of human rights tribunals, we might expect that states with stronger human rights practices would be more likely to comply with the rulings than states with weaker human rights records. Similarly, we might expect such states to be handed down fewer rulings than those states with more human rights violations. Yet, as discussed below, neither the number of adverse judgments handed down by the tribunals nor compliance maps onto states’ human rights practices. A second competing explanation for compliance focuses on the role of coercion. This theoretical framework suggests that strong states or regional hegemons push their neighbors to comply and might use instruments such as trade agreements or, in the case of the Court, EU membership4 to incentivize compliance. This explanation has roots in a long line of realist IR theory, which suggests that international law in and of itself is epiphenomenal but can be used as a coercive tool (Goldsmith and Posner, 2005; Gruber, 2000; Ikenberry et al., 1988). More recently, however, scholars have pointed to the potential effects of aid and trade conditionality on human rights practices, and it would follow that such pressure might also affect compliance with human rights tribunals’ rulings (HafnerBurton, 2008, 2009; Richards and Gelleny, 2001). For coercion to be effective, however, there must be a hegemon willing and able to use coercive measures to facilitate compliance. While the EU and the dominant European powers might desire that other states comply with the ECtHR’s rulings, they very rarely, if ever, rely on truly coercive measures. Instead, they reach for naming and shaming and similar tools. A third competing hypothesis, advocated by social constructivists, argues that international law helps to shape states’ preferences and facilitates compliance through an iterative process of social learning (Checkel, 2005: 801–826; Goodman and Jinks, 2004). International organizations teach states the appropriate ways to behave, while NGOs, transnational advocacy groups, and other actors can name and shame states, encouraging

Hillebrecht

1105

them to change their behavior to align with international norms and laws (Finnemore, 1996; Finnemore and Toope, 2001). Recent scholarship has suggested that naming and shaming from high-profile NGOs, the media, and the United Nations, for example, does, indeed, have an effect on states’ human rights practices, albeit in limited and defined ways (Krain, 2012; Lebovic and Voeten, 2006, 2009). Lambert Abdelgawad (2008) even identifies naming and shaming as one of the Council of Europe’s most effective tools in facilitating compliance. In addition to these three main explanations for compliance with international law, a robust literature in International Relations has developed over the past decade which suggests that domestic political institutions, not international organizations or normative constructs, are the linchpin to explaining compliance with international law. As international human rights institutions have weak enforcement capabilities, the responsibility for enforcement falls to states and their domestic institutions. Executives, independent judiciaries, civil society, political competition, and other domestic actors fill the enforcement lacuna left by international institutions (HafnerBurton and Tsutsui, 2004; Hathaway, 2002, 2007; Neumayer, 2005; Simmons, 2009; Vreeland, 2008). While much of the existing research has focused on states’ initial decision to join human rights treaties or their subsequent compliance with those treaties, this literature is particularly insightful when it comes to explaining compliance with the ECtHR’s rulings. Indeed, the importance of domestic institutions is perhaps even stronger with respect to human rights tribunals like the ECtHR. While the ECtHR does provide more oversight than many other international human rights instruments and institutions, it is simply not capable of enforcing its own rulings (Carozza, 2003: 38–79; Lambert Abdelgawad, 2008; Ministers’ Deputies, 2004; Pasqualucci, 2003; Sitaropoulos, 2008). The reality of the ECtHR is that discharging the compliance responsibilities mandated by the Court and the Committee of Ministers requires cooperation from multiple domestic actors. The International Relations and international law literatures posit two main channels through which domestic institutions facilitate compliance. The first is by providing a screening mechanism. That is to say, the presence of strong domestic institutions means that states have a higher likelihood of being held accountable for their human rights commitments. By virtue of this raised expectation of enforcement, it holds that states that are likely to violate human rights or that would be unwilling to comply with the tribunals’ rulings would not accept the ECtHR’s jurisdiction in the first place. This screening mechanism would increase compliance rates by vetting out any potentially non-compliant states (Alter, 1998; Conant, 2002; Simmons, 2000; Simmons and Hopkins, 2005; Von Stein, 2005). This mechanism, while illustrative of states’ initial membership decisions regarding human rights treaties, no longer holds. The ECtHR’s membership reaches to all of the Council of Europe states, including countries such as Azerbaijan, Georgia, and Turkey. It is quite clear that not all members of the ECtHR intend to comply with the tribunals’ rulings, or even to pretend to care about them. The second mechanism by which domestic institutions facilitate compliance is by providing post hoc enforcement. For example, Hathaway (2002, 2007) explains that domestic institutions such as an open media, competitive elections, and an engaged civil

1106

European Journal of International Relations 20(4)

society can provide channels for the incorporation of international laws and norms into the domestic political sphere (Barkhuysen and Van Emmerik, 2005; Hathaway, 2002, 2007; Neumayer, 2005; Vreeland, 2008). These domestic institutions can also help expedite compliance through the formation of pro-compliance coalitions that bring together multiple domestic actors working toward compliance (Cardenas, 2007). Further, these domestic constraints limit executives’ ability to shirk their compliance obligations, often forcing the issue of compliance onto the domestic political agenda (Hillebrecht, 2012a, 2012b, forthcoming).

3.1.  Domestic political actors and compliance with the ECtHR In her analyses of international human rights treaties, Simmons (2009) identifies three ways in which international law affects domestic actors and subsequently facilitates human rights change: (1) altering the national agenda; (2) providing a source of law and litigation; and (3) mobilizing civil society (Simmons, 2009). While the rulings handed down by the ECtHR differ in a number of ways from human rights treaties, they can play a similar role in domestic politics: first, the ECtHR’s rulings and the measures of nonrepetition that follow from them can change states’ national human rights agendas; and, second, the rulings provide the political coverage — and courage — that domestic actors need to engage in human rights reform. Unlike the decision to sign or ratify a human rights treaty, states cannot dictate when the ECtHR will hand down a judgment against them, nor can they pre-empt what the content of that judgment will be. This means that the rulings can act as exogenous shocks, altering states’ human rights agendas. For example, in the United Kingdom, both the Blair and Brown governments sought to bury the issue of prisoners’ voting rights, but a ruling from the ECtHR made this question front-page news, forcing the government to address — and ultimately acquiesce to — prisoners’ voting rights.5 The Court’s rulings and the attendant obligations for states to change their policies and practices can pose a significant alteration to states’ human rights agendas, and perhaps even more significantly, can wrench control of the states’ human rights agenda away from the executive towards other domestic institutions, like the judiciary, legislature, or civil society. Second and relatedly, the Court’s rulings are generally perceived as legitimate external mandates for human rights reform. Through a number of mechanisms, ranging from its monopoly on highly specific and legalistic information to its explicit mandate that allows it to adjudicate on and reprimand states for their bad behavior, the ECtHR has developed and maintained a reputation for being a legitimate human rights actor with a clear and authentic interest in protecting human rights (Barnett and Finnemore, 2004; Çali et al., forthcoming; Franck, 1995; Huth, 2006). The perceived legitimacy of the ECtHR’s rulings provides political cover for domestic actors looking to advance domestic policy change. That is, domestic actors can justify changes in the domestic human rights agenda and alterations of human rights policies to domestic audiences by pointing to the Court’s legitimacy and moral authority. This, coupled with the specificity of the measures of non-repetition states must enact following an adverse judgment, makes the ECtHR’s rulings a tidy focal point around which domestic actors can mobilize for reform.

Hillebrecht

1107

Changing the domestic human rights agenda and providing political cover for rights reform are two key mechanisms by which the Court can have an impact, through compliance, on states’ human rights policies and practices, but neither of these processes happens automatically. Instead, the process of compliance depends on coalitions of domestic actors — executives, judiciaries, legislatures, and civil society — that are able and willing to push through human rights reform (Alter, 2013). While such a coalition of domestic actors might slow the compliance process by involving more voices and potential veto points, it increases the likelihood of compliance by providing post hoc enforcement of the tribunals’ rulings. Compliance coalitions reduce the possibility that one domestic actor will shirk its compliance obligations. At the same time, these coalitions engage in the actual work of changing policies and practices in order to comply with the ECtHR’s rulings. Executives play a unique role with respect to compliance with the ECtHR. They are, sometimes simultaneously, the actor most responsible for compliance with the ECtHR’s rulings and perhaps the actor with the most at stake with respect to compliance. When the ECtHR hands down rulings, it issues the rulings to the executive branch, often the foreign ministry or diplomatic representation in Strasbourg. Once the ruling has been handed down, executives must delegate the responsibility for compliance to other domestic actors or wait for domestic actors to take up the mantle of compliance independent of, or perhaps in spite of, the executive. Executives might have a number of incentives to comply with the ECtHR’s rulings, including, but not limited to, a personal commitment to upholding human rights, signaling a commitment to human rights, or leveraging the Court’s rulings to advance domestic policy change. Executives can also use the perceived legitimacy of the Court’s rulings as a way to convince skeptical domestic audiences to embark on human rights reform. At the same time, the tribunals’ rulings pose a threat to an executive’s agenda-setting authority by introducing human rights issues that might be outside of the executive’s top priorities. Further, by mobilizing domestic actors, the ECtHR’s rulings ultimately arm judiciaries, legislatures, and civil society actors with an externally legitimated blueprint for human rights reform that might be counter to executives’ own policy preferences or domestically unpopular. As such, the rulings introduce a certain amount of risk into the domestic human rights policymaking process that could make even the most receptive executives somewhat wary of compliance. Further, even when executives are amenable to the measures of non-repetition suggested by the ECtHR and the Committee of Ministers, they generally cannot accomplish compliance with measures of non-repetition without the participation of other domestic actors, especially when those measures of non-repetition require states to repeal old laws and to pass new ones. Perhaps one of the most important allies in a compliance coalition is the judiciary. In fact, the literature on the implementation of the ECtHR’s rulings points to independent judiciaries as some of the most important indicators of successful compliance with the ECtHR’s rulings (Besson, 2008; Ferreres Comella, 2006; Kaboglu and Koutnatzis, 2008; Powell and Staton, 2009; Stone Sweet and Keller, 2008). Courts, both in the aggregate and individual judges, use the ECtHR’s jurisprudence to inform their own work, and many use the Court’s rulings as a yardstick against which they measure their country’s own human rights laws. Litigators can also use the ECtHR’s jurisprudence as legal tools

1108

European Journal of International Relations 20(4)

to argue cases on behalf of their clients (Besson, 2008; Ferreres Comella, 2006; Kaboglu and Koutnatzis, 2008; Powell and Staton, 2009; Stone Sweet and Keller, 2008). Similarly, legislatures can use the ECtHR’s mandates for measures of non-repetition as an impetus to pass new legislation. The perceived legitimacy of the Court’s rulings provides political cover for individual legislators who might embark on difficult human rights reforms, while the prospect of laws being subjected to future ECtHR scrutiny can be a way of facilitating legislative cooperation with the measures of non-repetition (Scribner and Slagter, 2011). Activists and NGOs can mobilize around the rulings in order to push for policy change, and measures of non-repetition provide clear plans for what that change should look like. Moreover, NGOs and the media can act as informational conduits, bringing transparency to the compliance process and putting pressure on state actors for compliance (Botelho Moniz and Maia Cadete, 2007; Hafner-Burton, 2008; Robinson, 2000; Santos, 2007; Sitaropoulos, 2008). NGOs also serve as legal and substantive experts, bridging the gap between the victims, the tribunals, and the state (CEJIL, 2008; Krsticevic, 2007). In these ways, civil society can facilitate compliance by providing pressure, expertise, and transparency (Dai, 2005). Domestic actors, ranging from the executive to the judiciary, from the legislature to NGOs, can work, often in concert, to provide post hoc enforcement of the ECtHR’s rulings. It follows from this discussion, then, that states with the most robust domestic institutions should be more likely to comply with human rights tribunals’ rulings. As compliance with measures of non-repetition is where the ECtHR and its counterparts can have the most impact in changing states’ human rights practices, these domestic actors hold the key to the Court’s ability to affect human rights on the domestic level.

4.  Measuring compliance with international human rights tribunals One of the main challenges in understanding when and how states comply with international human rights tribunals’ rulings is operationalizing compliance. While the ECtHR catalogues its expectations about compliance in the form of case reports, it has not done much to measure compliance in a comparable and consistent way. This is exacerbated by the structure of states’ compliance obligations. When the Court hands down adverse judgments, states face a spectrum of compliance obligations, ranging from paying individual reparations to changing their human rights policies, and any one case might involve numerous, discrete compliance obligations. When faced with this spectrum of demands, states often treat their compliance obligations like choices on a menu: picking and choosing the parts of the rulings with which they want to comply. It is rare, in fact, for states to comply with none or all of the discrete elements in a ruling. Instead, they tend to comply with the rulings in part (Hawkins and Jacoby, 2010). In order to mitigate these measurement problems, this article offers new data on compliance with the ECtHR’s rulings. Beginning with the unit of analysis, these data present a consistent and nuanced approach to measuring compliance. The unit of analysis here is the discrete obligations states are asked to fulfill within each ruling. The

Hillebrecht

1109

data used in this article identify each of the discrete compliance obligations that refer to measures of non-repetition and codes each obligation for compliance (Colandrea, 2007: 396–411; Hawkins and Jacoby, 2010; Hillebrecht, 2009, forthcoming; Loucaides, 2008). Compliance is a dichotomous variable, where each observation is coded as 1 if the state fulfilled that measure of non-obligation and 0 otherwise. As described above, the non-repetition measures ask states to change their domestic policies, amend their legislation, and overturn and update their case law and judicial practice. Such measures might include overturning amnesty laws, enacting new labor laws, or enforcing legislation regarding gender equality. The ECtHR generally defers to the Committee of Ministers and the states to determine the appropriate course of action following adverse judgments (Colandrea, 2007: 396–411; Loucaides, 2008). In its reporting on states’ implementation of the ECtHR’s rulings, the Committee of Ministers takes a number of steps to identify what the state has done and what the Committee of Ministers believes should be done in the future. Although not in a uniform or even organized fashion, the Committee of Ministers essentially identifies a list of measures that states should take before the Committee is willing to declare the states’ compliance obligations fulfilled. As the Committee of Ministers has the final authority to determine the cases in which states are fully compliant and which require more effort, their decisions and deliberations about which steps the states should take provide important information about the universe of obligations. The Committee of Ministers maintains a website on the state of execution (e.g. compliance) for the outstanding cases for each state. This website is updated periodically and contains highly specific information. For each case, the Committee of Ministers describes in detail the steps that the state has taken towards compliance and what steps the Committee of Ministers expects to see before declaring the case closed (Council of Europe, 2009). I used these documents to create a list of measures of non-repetition with which the Committee of Ministers expects states to comply. In total, this state of execution database contains 584 cases and 1184 discrete obligations pertaining to measures of non-repetition. A number of caveats are in order with respect to these data. First, because the state of execution database only includes cases that are pending, meaning cases with which states have not fully complied, there is the potential that this sample underestimates compliance. While potentially underestimating compliance, this approach allows for robust tests of hypotheses about compliance by making compliance the least likely scenario and thereby making it more difficult for the independent variable to have a statistically significant effect on compliance than if the data were biased toward observations of ‘easy’ compliance. Second, as noted above, the ECtHR does not specifically outline the discrete obligations. This difficulty was mitigated through a cautious reading of the case reports. Third, in recent years, the ECtHR has focused on eliminating what it calls ‘repeat’ or ‘clone’ cases. These cases deal with issues already addressed in other cases, and it follows that complying with the first, or ‘lead’, case would satisfy a state’s compliance obligations for the clone cases. To avoid double-counting compliance with those general measures, I did not include those obligations that the Committee of Ministers indicated were satisfied by virtue of compliance with another case.

1110

European Journal of International Relations 20(4)

The compliance data set also includes information on the types of abuse at the heart of each case. The abuses are divided into five categories: (1) physical integrity rights; (2) civil and political rights; (3) social and economic rights; (4) legal procedure and due process rights; and (5) property and privacy rights. Table 1 shows states’ compliance performances with respect to measures of non-repetition.

5.  When do states comply with the ECtHR? This section tests the central hypothesis of this article: stronger domestic institutions increase the likelihood of compliance with human rights tribunals’ rulings. I test this and competing hypotheses by estimating a series of logit models on the compliance data set explained in Section 4 above. As the discussion in Section 3 suggests, the main explanatory variable of interest is domestic political institutions. In order to capture the aggregate level of domestic institutions, I use three different measures of domestic institutions. The first is the xconst variable from the Polity IV data set. This is a composite variable that takes into account ‘the extent of institutionalized constraints on the decision-making powers of chief executives, whether individuals or collectivities’ (Marshall et. al., 2013: 23). Because domestic institutions effectively limit the executives’ ability to shirk the state’s compliance obligations and provide pro-compliance partners, this variable nicely captures the effect of domestic institutions on compliance. The variable ranges in value from 1 to 7, where a score of 1 suggests unfettered executive authority and a score of 7 indicates a highly constrained executive. As an alternative to the xconst variable, I use the general polity score from the Polity IV data set in robustness checks to test the possibility that democracy and democratic institutions more broadly defined also increase the likelihood of compliance. The polity variable ranges from −10 to 10, with −10 indicating a highly authoritarian regime and 10 indicating a highly democratic one (Marshall et al., 2009). In addition to these variables, I include robustness checks using the Democratic Accountability score from the International Country Risk Guide. This six-point measure evaluates how responsive a government is to its people. This variable ranges from 0, indicating a non-responsive regime, to 6, denoting a highly responsive one. This measure captures, again, the degree to which executives are hemmed in by the state’s larger institutional configuration, with more restricted executives less able to shirk their compliance responsibilities than those who are subject to less democratic accountability. To think about this differently, those governments that are subject to the most democratic accountability are likely to be those that also have the strongest domestic institutions (Marshall et al., 2009). The models also account for the main competing hypotheses, explained above: (1) that compliance is the result of the codification of existing behavior and that states with the strongest human rights practices are more likely to comply with the tribunals’ rulings; (2) that compliance is the result of coercion from important international partners, such as the European Union (EU); and (3) that compliance is the result of socialization in the international community. To control for existing human rights conditions, the models below use the empinx and physint measures from the Cingranelli and Richards Human Rights Database. Empinx is a composite variable, taking into account governments’ respect for freedom

1111

Hillebrecht Table 1.  Compliance rates. Country

Not complied

Complied

Total

Percent complied

Albania Armenia Austria Azerbaijan Belgium Bosnia and Herzegovina Bulgaria Croatia Cyprus Czech Republic Estonia Finland France Georgia Germany Greece Hungary Iceland Ireland Italy Latvia Lithuania Luxemburg Macedonia Malta Moldova Netherlands Norway Poland Portugal Romania Russia Serbia Slovakia Slovenia Spain Sweden Switzerland Turkey Ukraine United Kingdom Total ECtHR

16 5 9 14 10 8 58 19 6 11 1 11 16 17 2 29 5 4 3 26 7 5 4 12 6 49 6 2 41 12 16 112 17 9 7 5 2 2 75 65 7 730

6 0 3 3 5 1 35 20 1 9 0 2 12 7 0 34 3 0 7 15 7 0 4 5 0 23 10 3 38 10 17 44 5 23 7 0 2 2 51 29 14 455

22 5 12 17 15 9 93 39 7 20 1 13 28 24 2 63 8 4 10 41 14 5 8 17 6 72 16 5 79 22 33 156 22 32 14 5 4 4 126 94 21 1184

27% 0% 25% 18% 33% 11% 38% 51% 14% 45% 0% 15% 43% 29% 0% 54% 38% 0% 70% 37% 50% 0% 50% 29% 0% 32% 63% 60% 48% 45% 52% 28% 23% 72% 50% 0% 50% 50% 40% 31% 67% 37%

1112

European Journal of International Relations 20(4)

of movement, freedom of speech, freedom of religion, workers’ rights, and political participation. This variable ranges from 0 to 10, with a score of 0 indicating no respect for these rights and 10 indicating full respect for these rights. Physint measures governments’ respect for four rights: the rights against torture, extrajudicial killing, political imprisonment, and disappearance. This variable ranges from 0 to 8, with 0 indicating no respect for these rights and 8 representing complete respect for these rights (Cingranelli and Richards, 1999: 407–418; 2006). These indices capture the rights protected by the European Convention on Human Rights, and, similarly, those rights most often violated by states. These variables measure human rights conditions at the time of the ruling. As it is important to capture the human rights conditions in which governments and domestic institutions decide to undertake compliance, it is critical to measure human rights conditions at the time of the ruling, not the time of the abuse, which, as noted above, could have been decades prior. A second competing explanation for compliance focuses on the role of coercion. This theoretical framework suggests that strong states or regional hegemons push their neighbors to comply and might use instruments such as trade agreements to incentivize compliance. I control for this hypothesis by including an indicator for states’ embeddedness in the international system: the percentage of Gross Domestic Product (GDP) accounted for by trade. This indicator comes from the World Bank’s 2009 World Development Indicator Database (World Bank, 2009). In addition to a state’s general embeddedness in the international system, the EU could play a singularly important role in dictating compliance with the ECtHR. Membership of the Council of Europe and the ECtHR are often preconditions for membership of the EU, and recent research suggests that pro-European governments seek to appoint activist judges to the ECtHR (Voeten, 2007, 2008). Over the past few years, the EU has debated the merits and possibilities of becoming an institutional member of the ECtHR, increasing the importance of complying with the human rights tribunals’ rulings for EU members. In order to take into account the possible effects of EU membership, the models include a dichotomous variable, EU Member. The variable indicates whether or not a country is a member of the EU at the time of the adverse judgment, with 1 indicating members and 0 indicating others. In order to take into account states’ socialization, the models include a variable that captures states’ membership in international human rights agreements, namely: the International Covenant on Civil and Political Rights; the International Covenant on Economic, Social and Cultural Rights; the Convention against Torture, and Other Cruel, Inhumane or Degrading Treatment or Punishment; the Convention on the Elimination of All Forms of Discrimination Against Women; the Convention on the Elimination of All Forms of Racial Discrimination; the Convention on the Rights of the Child; and the Convention on the Protection of the Rights of All Migrant Workers and their Families. The variable treatymem is a percentage of the institutions of which the state is a member at the time of the human rights tribunals’ rulings. Although measuring socialization directly is difficult, this variable can serve as a proxy, as it stands to reason that the more involved a state is in the international human rights system, the more likely the state is to

Hillebrecht

1113

be socialized into international norms (Caporaso et al., 2003; Checkel, 2005; Johnston, 2005). An additional control takes into account the country’s experience with democracy. Moravscik (2000) finds that democratizing states join human rights tribunals in order to signal their commitment to the international human rights regime and to lock in prodemocratic, pro-human rights policies (Caporaso et al., 2003; Johnston, 2005; Moravscik, 2000). Although Moravscik is writing about states’ initial decisions to become members of human rights tribunals, this theoretical proposition has clear extensions to subsequent decisions about compliance. According to this logic, newly democratized or democratizing states might have the clearest incentive to comply with the tribunals’ rulings in order to demonstrate their commitment to upholding human rights to domestic and international audiences and to use international law to buttress domestic human rights advances. In order to account for the effects of democratization on compliance, the models include a variable, demlength, that is a count of the number of years that a country has been a democracy. Using the Polity IV scores for regime change (regtrans), regime duration (durable), and democracy (democ), this variable counts the number of years between a country’s transition to democracy and the date of the tribunals’ judgments (Marshall et al., 2009). Implementing the tribunals’ rulings takes time, particularly for those measures that involve implementing new legislation or changing existing human rights laws and practices. In order to account for the time it takes for compliance, the models include the variable timelag. This variable is a count of the number of years between the adverse judgment and the time of data collection. In the case of the ECtHR, the time of data collection is 2009, when the case reports documenting the current state of compliance were issued. While this variable artificially imposes an end-date on compliance (e.g. the time of data collection) and states could comply with the tribunals’ rulings at any point after the judgment is handed down, it still accounts for the potentially lengthy compliance process. The models also include a measure of GDP per capita, and consistent with contemporary practice, take the natural log of this measure (World Bank, 2009). Compliance with the tribunals’ rulings can be exceedingly costly, and controlling for GDP per capita ensures that states are not priced out of compliance because of insufficient financial resources. The models also include dummy variables for each of the types of violation associated with the case, Physical Integrity, Political and Civil, Social and Economic, Legal Procedure and Due Process, and Privacy and Property, with Privacy and Property the excluded category. The models use the value for the year the judgment was handed down. For example, if the ECtHR issued a judgment in 2003, all of the explanatory variables reflect the conditions in the state in 2003. As the theoretical framework in Sections 2 and 3 explains, it is important to capture the conditions under which governments weigh the prospects of compliance, not the conditions at the time of the violation, especially as there could be — and often is — a considerable gap in time and circumstance between when the judgment is handed down and the time of the abuse. I estimate a set of logit models, with robust standard errors clustered by country.6 The basic form of logit models is:

1114

European Journal of International Relations 20(4)

Probability ( Compliance = 1) = β 0 + β1Domestic Institutions + β 2 Physical Integrity Rights + β3 Empowerment Rights + β 4 Treaty Membership + β5 Trade as Percent of GDP + β 6 Length of Democracy + β 7 EU Member + β8 Physical Integrity Violation + β9 Civil / Political Rights Violations + β10 Law /Procedural Rights Violation +

β11Time Lag + e Models 1 and 2 use xconst as the indicator for the strength of domestic institutions. Model 2 does not include the empowerment rights index as missing data for that index reduces the sample size and tracks closely with the strength of states’ democratic institutions. Models 3 and 4 use states’ Polity scores and democratic accountability scores, respectively, to measure the robustness of domestic institutions. Models 5, 6, and 7 look only at rulings involving physical integrity rights violations, legal and procedural rights violations, and property and privacy violations, respectively. These models take three of the most common rights violations and test the assumption that domestic institutions have an effect of compliance across these different issue areas. In keeping with the hypothesis that compliance is endogenous to states’ human rights practices, Model 5 only uses states’ physical integrity scores, while Models 6 and 7 only use states’ empowerment index scores. Finally, Model 8 looks at compliance only with respect to strong democracies, defined by Polity as scoring 6 or higher on the polity scale. By only examining compliance in strong democracies, this model takes into account the potential bias that states with strong domestic institutions are more likely to address human rights abuses at home, rather than through the European human rights system. Table 2 shows that the data set actually includes more measures of non-repetition for states with stronger democratic institutions, thus suggesting that these states are not dealing with structural human rights abuses through domestic courts, while Model 8 statistically tests the assumption that domestic institutional constraints still matter for compliance for these strong democracies. Table 3 shows the results of all of these models. The results of the main model, as well as the various other specifications and robustness checks, are consistent with the expectations posited in Section 3 above. The variable xconst is positive and statistically significant. As domestic institutions are more robust, states are more likely to comply with the ECtHR’s rulings. Moreover, as the results of Table 2.  Executive constraints and measures of non-repetition. Strength of executive constraints Low                          High xconst score Number of non-repetition measures ordered

xconst = 2 17

xconst = 3 xconst = 4 xconst = 5 xconst = 6 xconst = 7 No obs 69 164 82 801

Hillebrecht

1115

Model 8 suggest, even in strong democracies, more robust domestic institutions have a positive and significant effect on compliance with measures of non-repetition. This finding holds across all model specifications, barring Model 5. Model 5 examines compliance in cases involving physical integrity violations. Among the Council of Europe member states — and certainly among the democracies and their constituents — physical integrity violations are perceived to be the most serious breach of a state’s human rights commitments. Executives are eager to repair any damage to their reputation that might follow an adverse judgment on physical integrity violations, thus making executive constraints less important in these cases.9 When it comes to measures of non-repetition following abuses of due process and property and privacy concerns, however, executive constraints do have a positive and statistically significant effect. While executive constraints and democratic institutions are statistically significant and help to account for states’ compliance with human rights tribunals’ rulings, other domestic factors have no statistically significant effect on compliance. For example, in contrast to received wisdom about the endogeneity of compliance with states’ human rights practices, states’ existing human rights standards have either no statistically significant effect or a weak, negative effect on compliance with the measures of non-repetition handed down by the ECtHR. Similarly, the length of democracy has little bearing on compliance, suggesting, perhaps, that new and established democracies alike must face the challenges and prospects of compliance. Other competing explanations suggest that the impetus for compliance might come from outside the state, whether in the form of normative pressure, coercion, or reciprocity. The results suggest that the likelihood of compliance does not increase as states join more human rights treaties. Even the EU variable, measuring EU membership, is not statistically significant. These results indicate that external factors do not drive compliance and tacitly support the idea that compliance is a domestic process. Two of the variables that capture case-specific characteristics (timelag and physical integrity) are positive and statistically significant in many of the model specifications. These demonstrate, as one might expect, that the more time a state has to comply, the more likely it is to enact the domestic policy changes necessary to comply with the tribunals’ rulings. These results also suggest that states are more likely to address cases dealing with physical integrity rights than, for example, law or procedural rights. Of course, implementing structural changes to states’ legal systems are enormous undertakings, and states might have few political incentives for such overhauls. It is important to understand not just the statistical significance of these results but also the substantive effects of changes in domestic political constraints on the likelihood of compliance. Given the sample of data used in this analysis, the likelihood of a state complying with any given obligation is 51%. A one-point increase (improvement) in constraints on the executive increases the likelihood of compliance by nearly 10% (9.80%). Table 4 shows the predicted probability of compliance with varying levels of executive constraints, with all other variables at their sample mean. What do these statistical and substantive results say about compliance with international human rights tribunals’ rulings? Ultimately, the estimation results presented in Table 3 suggest that domestic institutions can play an important role in determining states’ compliance with the ECtHR’s rulings. More specifically, the results suggest that

.55 (.59) .00 (.00) .51 (.12)*** .00 (.00) −.19 (.26) .87 (.23)*** –

.13 (.18) .18 (.04)*** −8.67 (1.50)*** 870

−.04 (.22) .12 (.06)** −7.76 (2.41)*** 639

.40 (.09)*** −.09 (.05)* –

.39 (.12)*** −.16 (.09)* .05 (.09) .79 (.79) .00 (.00) .44 (.19)** −.00 (.00) −.11 (.44) .71 (.25)*** –

Model 2: Measures of non-repetition (xconst) without empinx

***indicates p