The Convention, the Court, and Second-Order Compliance
During the last two decades, the system of human rights protection set up under the European Convention on Human Rights (ECHR, or the "Convention"), with the European Court of Human Rights (ECtHR, or the "Court")—frequently heralded as being one of the most effective international courts in existence—at its center, has come under stress. One source of that stress has been the rapid growth of the number of individual applications lodged since the late 1990s, which have vastly increased the workload of the Court. While the resources allocated to the Court from among the Council of Europe's budget also increased, they did so at a disproportionately lower rate that proved insufficient to process the mounting number of applications in a timely manner, resulting in a massive backlog of pending cases that at its peak exceeded 150,000 applications. Several factors interacted to produce this development, among them the doubling of the number of states party to the ECHR as part of the enlargement of the Council of Europe (COE)—the ECtHR's parent organization—following the end of the Cold War, direct access to the Court after the 1998 institutional reforms under Protocol No. 11, certain jurisprudential developments that made applying to Strasbourg appear attractive to potential litigants, as well as the very mundane fact that it had simply become more widely and better known among potential beneficiaries of rights litigation and their lawyers in the member states. To reduce the backlog, further institutional reforms were initiated with Protocol No. 14 which, inter alia, tightened admissibility requirements and provided for inadmissibility decisions by single-judge formations. These reforms contributed to reducing the number of applications pending before a judicial formation to 64,850 at the end of 2015.
A second source of stress for the ECHR system arose as a growing number of member states began to challenge the boundaries of the Court's authority more vocally and persistently than in previous instances of disagreement with the Court's judgments. This "diffusion of critical discourse" may find its most prominent manifestations in the debates within the United Kingdom of adopting a domestic bill of rights to shield against the influence of the ECtHR and in the Russian Constitutional Court's 2015 judgment that the execution of ECtHR judgments could be refused when they are seen to contradict the Russian constitution, but criticisms have been voiced in many other countries as well. At the multilateral level, efforts to "rebalance the system [ . . . ] between law and politics and between the international and the national" have resulted in the 2012 Brighton and 2015 Brussels Declarations' foregrounding of the principle of subsidiarity, the future inclusion of an express reference to the national margin of appreciation in the Convention's preamble, and the Court's competence to give advisory opinions at the request of the member state's "highest courts and tribunals."
A third source of strain, in part correlated with the first two, concerns the problem of insufficient, delayed, or outright noncompliance with the Court's judgments. With the most recent institutional reforms after Protocol No. 14 now implemented, the 2015 Brussels Declaration pointed out that "emphasis must now be placed on the current challenges, in particular the repetitive applications resulting from the non-execution of Court judgments [ . . . ] and the difficulties of States Parties in executing certain judgments due to the scale, nature or cost of the problems raised." The COE's Steering Committee for Human Rights (CDDH) has similarly noted that while "the overwhelming majority of Court judgments are executed without any particular difficulty [ . . . ] the execution of some cases is problematic for reasons of a more political nature, while the execution of some other cases is problematic for reasons of a more technical nature due notably to the complexity of the execution measures or the financial implications of the judgment" and stressed "that the execution of Court judgments raising structural or systemic problems is key to alleviating the Court's burden and to preventing future similar violations." In 2015 the Parliamentary Assembly of the Council of Europe (PACE) which selectively monitors compliance with the Court's judgments of its own accord "note[d] with concern that the main challenges facing the Court, most notably the high number of repetitive applications and persistent human rights violations of a particularly serious nature, reveal a failure by certain High Contracting Parties to discharge their obligations under the Convention."
The COE's Committee of Ministers is charged with supervising the execution of the Court's judgments by respondent states (Article 46 (2) ECHR) and has had to deal with its own backlog of unexecuted cases, with the number of judgments pending execution at year's end hovering around 10,000 since 2010. The Committee's most recent report on supervising the execution of judgments classifies 89 percent of the 9,941 cases pending before it at the end of 2016 as "repetitive cases," that is, those relating to a convention violation found in an earlier judgment that has not yet been adequately remedied. This development of delayed and insufficient compliance is problematic not least because of the self-reinforcing effects it may have in that it "risks generating a vicious circle in which government officials point to public criticisms of the Court and compliance delays in other States to justify noncompliance in their own jurisdictions and to legitimize criticism of those who advocate adherence to the Court's rulings."
What is clear is that earlier claims that the Court's judgments "have [...] not only generally but always been complied with by the Contracting States concerned," with "no exceptions" and "without the need for 'enforcement,'" so that they could be considered "as effective as those of any domestic court," are no longer sustainable in an unqualified manner in the face of the mounting empirical evidence that at a minimum challenges, and at worst contradicts, such claims. Furthermore, the earlier assertions of perfect compliance have long suffered from a lack of systematic empirical verification. While there have been a good number of (usually legal) analyses of compliance with select individual judgments, or sets of judgments, as well as occasional country-wide studies, comprehensive assessments covering all judgments across all years and respondent states, or at least a representative sample thereof, and methodologically well-grounded inquiries into the causal factors producing the observable patterns of implementation and compliance have long been a desideratum in the literature.
More recently, this lacuna is beginning to be filled, not least due to the growing interest in studying the ECtHR among social scientists and an emerging focus on researching specifically the issues of compliance with the Court's judgments (what Roger Fisher had labeled "second-order" compliance), as well as with the Convention as such ("first-order compliance"). In a 2006 book on the ECHR's "achievements, problems and prospects," Steven Greer discusses certain problems especially of first-order (non)compliance from an aggregate as well as comparative point of view, without, however, subjecting any of the emerging hypotheses to systematic testing. Similarly, the 2008 edited volume A Europe of Rights examines the impact of the Convention and of the Court's jurisprudence on national legal systems through focusing on the ECHR's domestic "reception," with "reception" being understood as "how— that is, through what mechanisms—national officials confront, make use of, and resist or give agency to Convention rights." Rich in descriptive detail, the study's country reports "chart cross-national variance in the impact of the ECHR on national legal systems, and they provide materials for generating hypotheses that might explain this variance," without again, however, testing these hypotheses in other cases. In any event, while compliance with individual judgments is addressed as part of the country studies, neither its systematic assessment nor its causal explanation is the study's principal focus. The same holds true for a recent volume, authored by ECtHR and national judges as well as by some other legal practitioners, on the impact of the Convention and Court on the democratization process in Central and Eastern Europe.
In other work, Darren Hawkins and Wade Jacoby have mined Council of Europe data on compliance trends in support of their argument that "partial compliance" is a significant outcome with respect to a nontrivial number of ECtHR judgments, but they do not test any causal hypotheses to explain why such partial compliance obtains. A 2013 edited volume expressly foregrounds the causal questions concerning the implementation of ECtHR judgments, but in order to canvass possibilities rather than to "advanc[e] a consistent causal argument or a theory about the factors promoting and obstructing the domestic impact of and compliance with ECtHR case law." In a separately published article, two of the contributors to that volume found, on the basis of a sample of judgments and violations of Articles 8-11 ECHR against nine respondent states, that "the greater the legal infrastructure capacity and government effectiveness, the more expeditious the implementation of the ECtHR's rulings is likely to be."
In the first book-length study of compliance with human rights judgments by the ECtHR and its counterpart in the Americas, the Inter-American Court of Human Rights (IACtHR), Courtney Hillebrecht has foregrounded three causal pathways that might be expected to affect compliance with adverse judgments. Such compliance may occur, Hillebrecht argues, (1) because it enables states to credibly signal their commitment to human rights and build a reputation for compliance with human rights in particular and international law more generally; (2) because it allows domestic actors to set and advance their own human rights agendas in contexts where they face political or institutional constraints; or (3) because a democratic commitment to human rights and the rule of law makes states comply "begrudgingly," despite diverging substantive preferences. Empirically, Hillebrecht finds, inter alia, that compliance varies across remedial measures, with financial obligations being more readily complied with than, for example, the requirement to implement legislative reforms, and that compliance is positively correlated with the strength of domestic institutional constraints on the executive. Most recently, Sharanbir Grewal and Erik Voeten have found that new democracies tend to be quicker than established democracies in achieving compliance with comparable types of judgments.
This book adds to this emerging literature on compliance with ECtHR judgments. It makes two main contributions. First, it provides the first comprehensive assessment of the current state of compliance, as of March 2017, with all of the Court's judgments rendered up to the end of 2015, making use of the Committee of Ministers' own accumulated data on the execution of judgments. While raising certain questions as to the validity of the Committee's data as reliable indicators of compliance—which I address later on—the data set constructed for this assessment allows identifying actual compliance rates for countries, years, and types of violations across all judgments rendered through 2015. The quantitative evidence shows that 43.3 percent of the ECtHR's compliance-relevant judgments had not been sufficiently complied with as of March 10, 2017, and therefore remained under supervision with the Committee of Ministers, with some of them dating back all the way to 1996. While these aggregate statistics are significant in their own right, they remain limited with respect to the information they convey. Being concerned solely with the final outcome of the process of executing the Court's judgments, they do not say anything about whether the lack of compliance equates to outright noncompliance, partial compliance, or (so far) insufficient attempts to achieve full compliance, nor do they allow any inference as to the causal factors at work.
The second—and larger—part of this study tackles some of these issues. It examines both what may be called the "depth of compliance" issue as well as the question of how the observable compliance patterns can best be explained. Both compliance and noncompliance with the Court's judgments may be said to present puzzles from the vantage point of the two main theoretical orientations in international relations: rationalism and constructivism. As concerns compliance and rationalist theories, why would the governments of sovereign states comply with human rights judgments they disagree with, especially if they have popular majorities behind them, given that the Court and the Council of Europe (of whose institutional architecture the Court forms a part) have no means of enforcement nor any material incentives at their disposal, and that other states, in light of the largely domestically limited impact of international human rights regulation, for the most part have no incentive to do so either? Even if we accept that European democracies created the Convention system to lock in a particular from of liberal democracy, this does not automatically translate into a preference in favor of complying with each and every of the Court's substantive decisions, not least because the governmental actors that initiated ratification of the Convention and those that have to implement adverse judgments are usually not the same (except occasionally with respect to countries that only recently joined the Convention system). From a rationalist perspective, then, this leaves certain domestic channels of influence as the principal pathways of exerting influence on decision-makers to comply with the ECtHR's judgments. These pathways, I argue, may explain some, but by no means all, instances of compliance and thus remain only partially satisfactory in the ECHR context.
Constructivists, by contrast, might argue that widespread compliance should not be surprising, given that the parties to the ECHR are, for the most part, liberal democracies (or at least endeavor to be so), and that it is implicit in the identity of a liberal democracy that it accept the authority of a duly constituted court and comply with human rights judgments rendered against it; indeed, it is precisely through such compliance with human rights norms, among other behavioral expectations, that the very identity of a liberal democracy is reaffirmed and recreated. While such a norm- and identity-based approach may help illuminate compliance in the face of conflicting policy preferences on the part of the governmental actors that have to implement a given judgment, it runs into its own difficulties when faced with the phenomenon of judgments that fail to be complied with for many years. Furthermore, even if compliance occurs as a formal legal matter, a purely norm-based approach is underspecified in that it remains silent on the specific manner in which states implement ECtHR's judgments. As will become clear in the course of this study, these specific choices can matter a great deal for the actual impact of a given compliance decision on domestic law or policy and those affected by them—a differential impact that is not captured by merely postulating norm-following as such.
In light of these challenges for each position taken separately, I argue that the best explanation of the empirically observable patterns of compliance with the judgments of the European Court of Human Rights requires combining rationalist and constructivist logics of action into a joint model that takes both positions seriously. Specifically, I argue that to understand the compliance behavior of European liberal democracies, the reason why states comply with ECtHR judgments should be considered separately from the question of how they comply with them, at least analytically. Whereas the former is best explained, in the specific context here under consideration, by what may be thought of as a normative compliance pull exerted on European liberal democracies by the obligation to comply with the judgments of a duly constituted international court (the constructivist element), the decision of how to comply remains generally subject to a consequentialist logic based on political preferences and cost-benefit calculations (the rationalist element). In other words, governments will generally choose rationally among the available options within the normative constraint of having to comply formally with the judgment(s) rendered against them. Because there is no reason to expect that ECtHR judgments will trigger immediate preference reversals on the part of respondent governments, and because the issues decided by the ECtHR in its judgments against liberal democracies are, for the most part and with some notable exceptions, not of such a nature as to elicit widespread political or even electoral pressures to implement a judgment broadly, the standard course of action pursued by respondent states will be to execute judgments narrowly, guided by the preferences held prior to the judgment. These considerations result in the expectation that liberal democracies will always (eventually) comply with ECtHR judgments rendered against them, even without any meaningful enforcement beyond peer pressure, but that they will frequently do so in a restrictive, minimalist, or otherwise evasive manner and not infrequently with some delay.
These theoretical expectations are tested in the context of two case studies of two established democracies and longtime parties to the ECHR: the United Kingdom and Germany. With respect to teach of these two countries, I examine the manner in which they have complied with all of the ECtHR's judgments rendered against them until the end of 2010. Two considerations guided the case selection. First, established democracies are commonly expected to perform better with respect to complying with human rights—not least because human rights protection is a defining characteristic of most conceptions of democracy—and I was particularly interested in the manner in which they live up to that expectation. Second, the feasibility of conducting comprehensive case studies of all judgments rendered against a country required limiting the number of countries studied. Comprehensiveness is desirable to be able to observe remedial responses to all types of judgments in which violations have been found against a particular country, not just those deemed most important (by whatever standard) or a sample that might miss observations from the quite varied spectrum of issue areas and types of violations that can provide important insights for the present inquiry.
In this study, I take seriously both the internal logic of international law generally, and of the ECHR specifically, as well as the political forces that affect how law operates within a given political community and historical context. The work is this placed squarely at the intersection of research strands on human rights, on compliance, and on judicial politics. The focus on the European Court of Human Rights and the particular sociopolitical context within which it operates gives rise to some distinct themes and issues. First, the countries subject to the Court's jurisdiction are, or aspire to be, for the most part, liberal democracies, which leads to a change in focus both in terms of the human rights issues that arise within the system as well as with respect to the institutional mechanisms through which remedies may be sought and adopted. So far, much of the human rights literature in international relations has predominantly addressed physical integrity rights: the right to life, freedom from torture, and liberty and security of the person. Given the fundamental nature of these rights as well as the still-widespread violations thereof that can be witnessed the globe over, this is justifiable because it is here that human rights improvements are most urgently needed.
The issues addressed by the ECtHR, by contrast, go well beyond the important but limited set of physical integrity rights and include the full panoply of civil and political rights protected by the Convention and its protocols. While violations of the right to life and of freedom from torture are unfortunately not absent in contemporary Europe, they are, at least in the established liberal democracies, comparatively rare. Many cases instead concern what may be qualified as policy disputes that revolve around different interpretations of what a given right requires in specific circumstances, including such issue areas as the protection of private property, privacy rights, and procedural fairness. In many of these cases the stark language of human rights "abuses" or "crimes" in the form of intentional violations of clear treaty standards that suffuses the literature on physical integrity rights is inapposite. Rather, many of these cases are better characterized as disputes concerning rival interpretations as to what implications a particular norm has for law and policy in a given issue area, and it is only after the Court's judgment that what is to count as a violation is authoritatively established. Still, the fact that many of these decisions may be qualified as "fine-tuning sophisticated national democratic engines that [are], on the whole, working well" should not diminish the significance of these judgments. To the claimants and to others in similar circumstances in the respondent state and beyond, possible changes in policies on such issues as parental visitation rights, compensation for infringements of property rights, access to personal data, protection of privacy, or length of proceedings, and so forth, are of key importance. With physical integrity largely assured, other rights and their exercise move to the foreground and become the legally and politically relevant criteria according to which one's ability to live a "good life" as one sees fit is assessed.
In addition to greater clarity and precision that comes with the specificity of judicial decisions, exploring second-order compliance with judgments, rather than first-order compliance with the Convention, involves a shift in focus "from rights to remedies," that is, away from the scope of a right and the reason for its violation to the willingness of respondent states to adopt appropriate remedial measures to end that violation, prevent its recurrence, and compensate the victims, as necessary. The question of why state authorities chose a particular course of action that was subsequently found to be in breach of the Convention naturally remains relevant to the question of second-order compliance because it will likely affect the state's response; the implementation of the ECtHR's judgments, after all, frequently takes place in the same political and institutional environment that previously sustained the policy, action, or situation later found to constitute a violation. Because most of the Convention parties are democracies that have their own domestic systems of human rights protection in place, possess a modicum of bureaucratic infrastructure, and are by comparison with many of the human rights trouble spots in the world comparatively well-off economically, we should expect different opportunities for compliance with the Court's judgments—as well as different obstacles—than in the case of many developing countries that often lack some of the elements of a fully functioning state apparatus and political system. The state, after all, is not only the violator of human rights but also their most important protector.
Finally, in terms of research on judicial politics beyond the state, propelled particularly by interest in the Court of Justice of the European Union (CJEU, formerly known as the European Court of Justice [ECJ]) as a motor of European integration and by the global expansion of adjudication and other forms of institutionalized dispute settlement, the strategic environments within which international courts operate are partly similar, partly distinct. On the one hand, all courts share certain functional features that underpin their "court-ness," structurally, as dispute settlement institutions generally, and in contrast to other forms of third-party dispute settlers specifically. On the other hand, they also differ with respect to their specific institutional designs and procedures, with some differences being more consequential than others. Most important, the specific strategic environment within which courts operate—defined by the legal regime of which they are part, interactions with state and non-state users, and the relationship with other international institutions—is different for each court. Because the ECtHR is part of the intergovernmental COE, not the European Union (EU), it does not benefit from the direct effect and supremacy of EU law as does the CJEU. It remains formally an international court whose judgments do not have direct effect in domestic law, except to the extent that states themselves provide for it. Also, in terms of the states that fall under its jurisdiction, the post-Cold War enlargement of the COE and the increase in the number of ECHR parties has significantly diversified the range of issues the Court faces as well as the types of domestic institutional and political contexts within which its judgments have to be executed.
While it is not required for understanding the compliance dynamics discussed in this book to have specialist knowledge of the historical origins and development of the European Convention on Human Rights and of the procedures before its supervisory machinery, a basic understanding of how the Convention and Court have evolved and operate is useful. This section provides such an overview.
The European Court and Convention of Human Rights: A Brief Overview
When the Convention for the Protection of Human Rights and Fundamental Freedoms—the ECHR's official name—was signed on November 4, 1950, assessments of its supervisory mechanism in particular were mixed. While the call for a "Charter of Human Rights" and for a "Court of Justice with adequate sanctions for the implementation of this Charter" included in the 1948 Congress of Europe's "Message to Europeans" found resonance in some of the Convention drafts coming out of the Consultative (later renamed Parliamentary) Assembly of the newly founded COE, the governments of the member states could not agree on a mandatory supervisory mechanism with a strong court at its center. Instead, the original Convention provided for a more cautious, tripartite supervisory mechanism comprising a commission, a court, and the Committee of Ministers, the COE's executive decision-making body composed of government representatives. Incoming complaints were first to be screened for admissibility and assessed as to their merits by the European Commission of Human Rights (the "Commission"), whose reports, however, were not legally binding. Moreover, both the right of the Commission to receive complaints from individuals and the jurisdiction of the Court were made optional, each requiring separate declarations of acceptance, in addition to ratification, by the states party to the Convention (cf. Articles 25 and 46 ECHR ).
In the absence of a state's declaration accepting the individual complaints procedure, the Commission could deal only with interstate complaints against that state, while any complaints against states that had not accepted the jurisdiction of the Court were instead decided by the Committee of Ministers (Article 32 ECHR ). The Committee's decisions required a two-thirds majority of the members entitled to sit on the Committee (not only of those actually present and voting) and were legally binding. While this requirement denied veto power to the respondent state, the specific supermajority requirement also resulted in a few "nondecisions" where the votes for or against the finding of a violation failed to reach the required two-thirds threshold. Even when the Court's jurisdiction had been accepted by a respondent state, cases still needed to be actively referred to it, or else they would also be decided by the Committee of Ministers. Notably, under the original design, such referrals to the ECtHR could be initiated by the Commission or by a state party to the Convention with a stake in the case (see Article 48 ECHR ) within three months after the Commission had issued its report, but not by individual applicants themselves.
From 1959, when the (then part-time) Court was first set up after eight states had accepted its jurisdiction (as required by Article 56 ECHR ), until November 1998, when Protocol No. 11 entered into force, the Convention's control mechanism essentially operated under this original design. While many states accepted the right of individual petition and the jurisdiction of the Court at the time of ratification of the Convention or shortly thereafter, some states did so only with some delay. The United Kingdom, for example, was the first state to ratify the ECHR in 1951 but accepted the optional elements only in 1966; for Italy, the gap was eighteen years; for Turkey thirty-three years (acceptance of individual petitions) and thirty-six years (acceptance of the Court's jurisdiction). France ratified the Convention in 1974 and while accepting the jurisdiction of the Court at the same time, it submitted to the individual complaints procedure only in 1981. These delays, among other factors, explain why judgments involving some ECHR parties only appear later on in the Court's history. Only in 1990 was there identity for the first time between the number of Convention parties and those subject to individual complaints and the Court's jurisdiction. In the subsequent post-Cold War enlargement of the COE, it became a political expectation of all new member states—met without exception—to ratify the Convention and accept both optional clauses.
With the entry into force of reform Protocol No. 11 on November 1, 1998, the optional clauses were eliminated. Henceforth, individual complaints (Article 34 ECHR) and the Court's jurisdiction (Article 32 ECHR) became part and parcel of a fully judicialized Convention control system: the European Commission on Human Rights was abolished, as was the Committee of Minister's quasi-judicial function of deciding cases under Article 32 ECHR . From then on, the now full-time Court became the sole Convention body to decide on the admissibility and merits of complaints. Because some states had valued the opportunity to reargue their case before the Court or Committee in light of the Commission's report, the post-Protocol No. 11 system provides for the possibility of a request—by the applicant, the respondent state, or both—within three months of the date of a judgment by a chamber, composed of seven judges, for a rehearing of that case by the Grand Chamber (GC), consisting of seventeen judges, at the latter's discretion. When such a request is rejected, or none is submitted, the chamber judgment becomes final; when the request is accepted, the Grand Chamber's judgment will be the final decision in the case (Article 44 ECHR). This is significant because only final judgments trigger the obligation of having to abide by them (Article 46 (1) ECHR).
While Protocol No. 11 reformed key institutional and procedural aspects of the ECHR system, many elements also stayed the same. Most important, the Committee of Ministers continued to be responsible for supervising the execution of the ECtHR's judgments (Article 46 (2) ECHR) and for determining whether execution has been satisfactory in light of applicable compliance standards. This is a significant feature of the ECHR's institutional architecture, and I argue below that the Committee's assessments have become an acceptable proxy for determining compliance with the Court's judgments. Furthermore, at least in formal terms and in contrast to EU law and the output of the CJEU, the Court's judgments continue to have legal effects only at the international level (unless national law provides otherwise); that is, they do not directly invalidate, displace, or modify national laws and policies found to be in violation of the Convention. A proposal prepared by the European Movement had still included a provision that would have provided the Court with prescriptive powers, including the power to demand the "repeal, cancellation or amendment of the [domestic] act" found to be in violation of the Convention, but this provision had been rejected at the Convention's drafting stage and was not revived in the lead-up to Protocol No. 11. Instead, as under the original design, the Court can only award just satisfaction (effectively meaning financial compensation) for pecuniary and nonpecuniary damages (as well as costs and expenses) that result from a Convention violation "if the internal law of the High Contracting Party concerned allows only partial reparation to be made" (Article 41 ECHR, previously Article 50 ECHR ). Although in most cases awarded together with the decision on the merits, the Court sometimes finds the question of just satisfaction not yet ripe for decision and reserves it for a separate judgment.
In terms of substance, the ECHR protects the principal civil and political rights of the Western liberal tradition, with some additions and changes included in subsequent protocols (for quick reference, see Table 1). These rights are guaranteed to "everyone" within the Convention parties' "jurisdiction" (Article 1 ECHR), that is, independent, for the most part, of citizenship and not necessarily, even if usually, ending at a state's territorial borders. While merely suffering adverse consequences of a Convention party's actions abroad has been deemed insufficient to establish the Court's jurisdiction, when a Convention party exercises powers extraterritorially that are normally exercised by, or tantamount to, governmental authority, that party will be subject to the ECHR and responsible for violations. Article 15 ECHR provides for the possibility to derogate from Convention obligations "in time of war or other public emergency threatening the life of the nation," with the exception of a few rights, including the prohibition of torture under Article 3 ECHR.
All beneficiaries of a right can, in principle, submit applications charging violations to the ECtHR. Before the merits of an application can be considered, however, the complaint has to satisfy a number of admissibility criteria, which are laid down in Articles 34 and 35 of the Convention. Among these is the requirement that the applicant has been an actual victim of the alleged Convention violation or is at least a closely related person with a justifiable interest in the case (e.g., parents, caretakers) in situations where the rights holder is unable (deceased, disappeared, or otherwise incapacitated) to submit the complaint him- or herself (Article 34 ECHR). This victim requirement thus excludes both abstract assessments of the compatibility of domestic law and policy with the Convention as well as actio popularis suits on behalf of others. Further key requirements are that all available and reasonably effective domestic remedies must have been exhausted (with exceptions in cases of demonstrable ineffectiveness or intolerable hardships) and that the application must have been lodged within six months of the final domestic decision (Article 35 (1) ECHR). Furthermore, the complaint must not be manifestly ill-founded or incompatible with the Convention; that is, it must concern conduct legally attributable to a state party that has occurred (or has ongoing effects) after the Convention entered into force for that state, and it must implicate a right actually protected by the Convention (Article 35 (3) lit. a ECHR).
Protocol No. 14, in force since June 1, 2010, added a controversial new admissibility criterion that enables the Court to declare inadmissible applications if it finds that "the applicant has not suffered a significant disadvantage" (Article 35 (3) lit. b ECHR), subject to certain safeguards. This new criterion was seen as a necessary response to the Court's caseload crisis to enable it to focus on graver human rights violations. The combined significance of the admissibility criteria—with further restrictions on the horizon as part of Protocol No. 15—becomes evident when one considers that most applications fail to meet them. Over the period from 1959 through 2015, of the 674,155 applications decided by the Court, 28,674 applications were disposed of by way of a judgment, whereas 645,481 (95.7 percent) were declared inadmissible or struck out of the Court's list. While the latter category includes cases that may already have passed the admissibility hurdle but were discontinued as a result of a friendly settlement or unilateral declaration by the respondent state recognizing a Convention violation, a much larger share is due to inadmissibility issues. In 2015, of the 43,135 applications declared inadmissible or struck out, 4,628 were based on a friendly settlement or unilateral declaration, whereas most of the remainder (38,507, or 89.3 percent) failed to clear at least one of the admissibility hurdles.
The lack of speedy and broad compliance with the ECtHR's judgments contributes its own share to the Court's caseload. When legal or factual situations resulting in Convention violations are not tackled expeditiously and broadly, they will often result in many, sometimes thousands of, clone cases that further clog up the Court's docket. Indeed, this problem was the key motivation for including in Protocol No. 14 two new competences for the Committee of Ministers: the first intended to remove interpretive differences hindering supervision by allowing the Committee to ask the Court for an interpretation of the judgment concerned (Article 46 (3) ECHR), and the second enabling it to initiate "infringement proceedings" by asking the Court for its view on whether the state party in question has breached its obligation to comply with the judgments of the Court (Article 46 (4) ECHR). As the protocol's explanatory report noted, "Execution of the Court's judgments is an integral part of the Convention system. [ . . . ] The Court's authority and the system's credibility both depend to a large extent on the effectiveness of this process. Rapid and adequate execution has, of course, an effect on the influx of new cases: the more rapidly general measures are taken by States Parties to execute judgments which point to a structural problem, the fewer repetitive applications there will be." As one commentator has noted, it is "the slothfulness, or even recalcitrance, of states which have been the sources of hundreds or thousands of clone cases which have been plaguing the Court." Even though the Court has begun to join clone cases and to stay their consideration pending the execution of remedial measures as part of the pilot judgment procedure, they still, at a minimum, consume administrative resources. So far, however, the Committee of Ministers has refrained from making use of the new infringement procedure which its own rules see as apposite only in "exceptional circumstances" (despite calls by nongovernmental organizations [NGOs] to begin using it in cases that appear to meet this criterion).
The State of Compliance: A Quantitative Assessment
So how good—or bad—is the state of compliance with the Court's judgments? While most of the relevant raw data is publicly available, no comprehensive data set that could be readily used for descriptive or inductive statistical analysis has been previously assembled. Earlier assessments have either taken findings of violations by the Court as an indicator, with qualifications, of first-order noncompliance without researching second-order compliance rates; or have made inferences from the cases whose supervision was still pending before the Committee of Ministers during select years, without connecting that data back to the overall number of compliance-relevant judgments to calculate an actual compliance rate; or have limited their assessment to lead cases that reveal a new type of violation; or have looked at only a subset of states and violations of select articles. In this section, by contrast, while also using the Committee's final resolutions as a proxy for compliance, I present descriptive statistics based on just such a comprehensive data set that addresses the compliance status of all 16,368 compliance-relevant judgments issued by the Court until December 31, 2015.
The Value of Resolutions of the Committee of Ministers as Proxies for Compliance
As a matter of formal legal obligation, the duty to comply with adverse judgments and with the terms of friendly settlements, as well as the principal elements of that duty, are not (or in any event, no longer) contentious. The extent to which the Committee of Ministers' supervision of the "execution" of judgments can be taken as a suitable proxy for the existence and quality of compliance may be less so. The answer to the question of whether it is has to be provided by examining the manner in which the Committee has come to understand and exercise its supervisory function in practice. That practice has evolved over time. In the beginning, the Committee, as far as can be gleaned from the publicly available documents, discharged its duty in a rather restrained and "timorous" fashion and did little more than acknowledge the information submitted to it by the respondent state. This restrained approach was enshrined in the formal (and very few) rules for the application of former Article 54 ECHR  that the Committee adopted for the first time in 1976. During this early phase, there were a number of cases in which "the Committee appears not to have been especially strict with the state concerned, and where it might have been expected that more severe measures would have been appropriate." On several occasions, the Committee accepted as sufficient to end its supervision a state's promise not to repeat the impugned conduct or its announcement that legislation was currently being prepared and to be put before parliament when ready.
Over time, however, the Committee has strengthened the standards it applies during the supervision of the execution of judgments. Beginning in the early 1980s, the Committee made clear that its competence to review a judgment's execution extended to the adoption of general measures beyond the individual case, where relevant, and that such general measures were not merely welcomed when adopted, but were indeed required to end the Committee's supervision, a practice followed ever since. Furthermore, since the mid-1980s, the Committee stopped the practice of adopting final resolutions on the basis of promised or pending legislative projects and instead demanded evidence that the latter had actually been enacted before the former would be adopted. Similarly, when a state argued that the status of the Convention in domestic law would assure judgment-compliant interpretations and applications of relevant legal provisions by domestic courts, the Committee began requesting proof in the form of actual case-law to that effect. Moreover, in 1988 the Committee introduced a new instrument, the Interim Resolution, which it uses to address publicly—sometimes simply noting, sometimes criticizing—the current status of a judgment's execution without yet ending the supervisory process. What these changes make clear is that the Committee has stopped acting merely as the passive recipient of government information and now assesses with some thoroughness the information submitted against increasingly demanding standards. Supervision as a substantive compliance-control function has thus been in evidence since relatively shortly after the Committee began exercising its supervisory function, and this substantive understanding has been repeatedly reaffirmed politically, most recently in the 2012 Brighton and the 2015 Brussels Declarations.
The increased level of scrutiny exercised by the Committee is also reflected in the revised rules for the application of Article 46 (2) ECHR adopted in 2001, expanded in 2006, and most recently amended in January 2017, which state more expressly the requirements for compliance than the rather cursory 1976 rules did:
When supervising the execution of a judgment by the High Contracting Party concerned [ . . . ] the Committee of Ministers shall examine:
- whether any just satisfaction awarded by the Court has been paid, including as the case may be, default interest; and
- if required, and taking into account the discretion of the High Contracting Party concerned to choose the means necessary to comply with the judgment, whether:
- individual measures have been taken to ensure that the violation has ceased and that the injured party is put, as far as possible, in the same situation as that party enjoyed prior to the violation of the Convention;
- general measures have been adopted, preventing new violations similar to that or those found or putting an end to continuing violations.
The Committee, it is true, is first and foremost a political body composed of state representatives who are neither necessarily lawyers nor otherwise experts in human rights issues; what is more, because they are political delegates under direction of their state's government, there is the concern that as peer reviewers they "either lack incentives or have the wrong incentives when it comes to reviewing" the compliance behavior of other ECHR parties. The dangers of diplomatic logrolling and horse-trading are considerably mitigated, although not entirely eliminated, by the fact that the Committee has been assisted in its supervisory task by what is today the COE's Directorate General of Human Rights and Rule of Law, and, since the late 1990s, by the specialized Department for the Execution of the Judgments of the ECtHR within it, both of which are staffed with legally trained personnel. It is the Department that undertakes the first evaluation of any compliance-relevant information, both as to requirements and measures adopted, and on the basis of those assessments then makes recommendations to the Committee of Ministers. The Committee does retain the formal authority to decide on ending or continuing supervision of a judgment's execution, and while it has happened that it has decided to close a case despite the Department's doubts as to whether full compliance had been achieved, such occurrences are reportedly very rare. In any event, the participation of the Secretariat's legal staff assures that the execution of judgments receives principled legal evaluation before the Committee decides whether to end or continue supervision and "reinforces the removal of peer review activities from a political to a rule-bound domain," thus assuring the integrity and reliability of the Committee's compliance assessments. The political space for what might be thought of as "post-judgment compliance bargaining" is thereby being reined in.
In conclusion, while issues with respect to determining the normative requirements and assessing the actual measures adopted in achieving compliance continue to exist, it is fair to say that supervision by the Committee has "become quite rigorous" and that is has worked reasonably well for the most part. The standards that the Committee has come to apply set thresholds sufficiently high so as to make the Committee's final resolutions a reasonably reliable indicator of compliance with the ECtHR's judgments. With regard to the obligation to pay just satisfaction awards, a final resolution establishes full compliance. With regard to individual and general measures, while different views as to requirements under the Convention may arise, a final resolution establishes compliance at a reasonably high level, a level that has also been affirmed as sufficiently compliant by the Secretariat, even if broader, more extensive or additional measures might have been conceivable.
The Compliance Data Set
Using the Committee's resolutions as an indicator of compliance, it is possible to assess the state of compliance with the Court's judgments. To do so, I created a data set of all compliance-relevant judgments issued by the court between 1960 and 2015 and linked each entry with its current compliance status as of March 10, 2017.
I first generated a list of all judgments issued by the Court from its own chronological list of judgments as well as from HUDOC, the Court's online judgment database. Pieces of information collected include the core case identifiers (applicant name[s], respondent state, application number[s], date of judgment); the type of judgment (merits, just satisfaction, admissibility/preliminary objections, revision, interpretation, or striking-out judgments); the formation by which it was issued (committee, chamber, grand chamber); the judgment's finding (violation, no violation, friendly settlement, etc.); and the Convention article(s) violated (or, in the cases of striking-out judgments and findings of no violation, alleged to have been violated). In thirty-six instances, judgments involved more than one respondent state. Because the relevant unit of analysis is individual state responses, I split these up as if judgments had been issued against each state individually and only included the finding specifically for that state (generating additional observations in the data set). Next, I grouped together as a single observation those judgments that concerned separate aspects of the same case, as identified by application number. This concerns separate judgments on just satisfaction, subsequent revision judgments, and chamber judgments that had been successfully referred to the Grand Chamber under Article 43 ECHR (reducing the number of observations). I also removed thirty-one cases that had been successfully referred to the GC for a rehearing, but which were either still pending before the GC as of March 2017 or in which the GC issued its judgments only in 2016 and 2017, that is after, the cut-off year for this study. This cleaned-up data set for the years 1960-2015 comprises a total of 17,891 observations.
As the next step, I dropped from the data set all judgments that did not require any execution measures. This includes merits judgments that did not find a violation of the Convention and judgments addressing various procedural issues, such as those declaring an application inadmissible, addressing preliminary objections, or striking out an application from the Court's list due to reasons other than a friendly settlement or unilateral declaration and without making an award for costs and expenses. After eliminating all judgments that do not require any remedial response from respondent states, the data set includes 16,368 compliance-relevant observations: 15,441 judgments on the merits finding at least one violation of the Convention; 905 judgments recognizing friendly settlements between the parties in accordance with Article 39 ECHR (entailing the obligation of paying any agreed sums and sometimes of adopting substantive individual or general measures); 14 judgments ending proceedings in response to unilateral declarations by respondent states (and the promised undertakings therein); and 8 judgments that were struck off the Court's list for reasons other than a friendly settlement or unilateral declaration but that contained a financial award for costs and expenses.
Finally, I matched all judgments requiring some remedial response from states with their current compliance status, as determined by the Committee of Ministers. If a final resolution was adopted in a case, that case is considered sufficiently complied with and closed, and the date and number of the final resolution has been recorded in the data set. If supervision of the execution of a given judgment is still continuing, it is considered not yet fully complied with and has been coded as "pending." The data set includes all final resolutions adopted as of March 10, 2017 (that is, up to and including those adopted at the Committee of Ministers' first human rights meeting in 2017, held March 7-10).
Aggregate Compliance Status by Year of Judgment
Figure 1 depicts the aggregate compliance status of all judgments in the data set by year of judgment and as of March 10, 2017. It also clearly shows the stark increase in the Court's workload and output after the COE's enlargement in the 1990s and the reforms that instituted the full-time Court when Protocol No. 11 entered into force in 1999. Several aspects are worth mentioning. First, as revealed by the source data, viewed cumulatively across the entire time period, 9,279 judgments (56.7 percent) had been satisfactorily complied with, while 7,087 judgments (43.3 percent) were still pending before the Committee of Ministers due to insufficient compliance or lack of information as to the measures taken. Second, the compliance rate decreases when less time has passed between the year in which judgments have been issued and the present. This is generally in line with what one should expect: achieving and then assessing compliance are processes that take time, especially when general measures have to be adopted, implemented, and evaluated. Third, except for 2003, the rate of compliance with striking-out judgments has been slightly higher than for merits judgments (implied by the fact that the rate for merits judgments is below the overall rate for all judgments); this also is in line with expectations, given that such judgments require mostly only payment of compensation and, in the case of friendly settlements, reflect—ideally at least—mutually agreeable terms.
Fourth, and particularly important with respect to the assertions about compliance with the Court's judgments, is the observation that one has to go back all the way to 1995 to find a perfect compliance rate. Beginning with 1996, we find the first judgments that have remained under the supervision of the Committee of Ministers for over twenty years. While it is true that many of these judgments concern so-called clone cases that originate from the same systemic shortcomings—concerning, for example, actions of the Turkish and Russian security forces, Italy's problems with respect to the efficient administration of justice, or the failure of Ukrainian authorities to execute judgments of domestic courts rendered against the state—so that they overstate the number of distinct problems when viewed discretely (clone cases are individual, but not independent, observations), the fact remains that appropriate general measures that successfully remedy these systemic problems have either not yet been adopted or, when some such measures have been adopted, have not yet been (or shown to be) effective. As discussed in the country case studies below, even long periods of formal noncompliance do not necessarily mean that a respondent state has failed to take any remedial measures whatsoever. The more time passes between a judgment and compliance with it, however, the more this gap invites scrutiny of the government's commitment to bringing about compliance within a reasonable time and to devoting sufficient resources to that objective. While there are no strictly defined time periods within which compliance of particular types of violations that require individual or general remedial measures need to be accomplished—just satisfaction awards, by contrast, need to be paid within the time limit set by the Court (normally three months) or else incur default interest—the nontrivial number of judgments that have remained under the Committee's supervision in excess of five, ten, and even more years indicates serious issues with respect to some states' commitment to ensuring, or their ability to achieve, effective and swift compliance with the Court's judgments.
Compliance Status by Respondent State
Compliance patterns are crucially determined by national characteristics and, as the theory advanced in this book assumes, in the human rights domain, especially by a country's democratic credentials. To check the relationship between democracy and compliance, Figure 2 maps countries' aggregate compliance rates against their Polity IV polity scores, with the latter averaged across the number of years the state has been subject to the ECtHR's jurisdiction, that is, from the year of the acceptance of the Court's jurisdiction until 2015. The polity score is derived by subtracting a country's autocracy score from its democracy score, each of which ranges from 0 to 10; the resulting polity score can thus assume values between +10 (strongly democratic) to −10 (strongly autocratic). Both autocracy and democracy scores measure institutional regime aspects and do not include substantive compliance with human rights or civil liberties. In addition, the total number of compliance-relevant judgments involving a country as respondent is indicated by the size of its bubble, ranging from 20 for Montenegro, 144 for Finland, 387 for Hungary, and 802 for Greece, to 1,603 for Russia, 2,088 for Italy, and 2,986 for Turkey. Seven countries, accounting for 122 judgments, are not depicted due to lack of inclusion in the Polity IV data set, and the outlier Azerbaijan (average polity score of −7, 102 judgments, compliance rate of 1 percent) has been omitted for reasons of better graphic representation.
Overall, the results shown in Figure 2 show a positive correlation between polity score and compliance rate: the trend-line slopes upward, but its R2 value is modest, at .336, meaning that only about a third of the variation in compliance rates can be explained by a state's polity score. The graph also reveals that, with the exception of Italy, all states with compliance rates below 70 percent are states that became subject to the ECtHR's jurisdiction only after the end of the Cold War, and of these all except for Turkey (which accepted the jurisdiction of the Court in 1990) are formerly communist or socialist countries. This history is clearly not determinative, however, because several former Eastern Bloc countries—Slovenia, Poland, Lithuania, and the Czech and Slovak Republics—score high on both the polity variable and compliance rates. Three countries shown have perfect compliance rates of 100 percent: Denmark, Sweden, and Norway. Because the bulk of the judgments were rendered by the post-Protocol-11 Court, limiting the analysis to the period from November 1999 onward does not change the overall picture (only 3.7 percent of the judgments in the data set originated with the pre-reform Court).
The fact that quite a number of countries with high polity scores and compliance rates cluster together in the upper right-hand corner of the graph provides suggestive support for the expected positive impact of democracy on compliance. At the same time, the graph also shows that there is great variance in terms of compliance rates even among those countries that have consistently scored a perfect 10 on the polity variable while subject to the ECtHR's jurisdiction. In the end, this variance and the modest explanatory power of the polity score in a simple bivariate relationship is not surprising when one recalls that this score measures only a small set of institutional aspects that may be said to define a regime as more or less democratic, but that the modalities of the execution of adverse judgments are conditioned by many other potential factors, such as the specifics of domestic institutional arrangements and procedures, bureaucratic efficiency, resources, aspects of the legal system, prevailing political and cultural preferences, and so forth.
Plan of the Book
While the quantitative data presented above provide a good sense of the aggregate state of compliance with the Court's judgments, they necessarily offer only a snapshot of the compliance status at a particular point in time. By itself, that snapshot is incapable of revealing the underlying legal and political dynamics that have resulted in either compliance or the continuing absence thereof, and, just as important, whether noncompliance results from the intentional disregard of the Court's judgments and the absence of any remedial measures, or whether relevant measures have already been adopted that have, however, (so far) failed to remedy effectively (all of) the violations found by the Court. The compliance data set could be exploited further and be used for additional quantitative analysis by adding variables that can be expected to affect the speed and depth of compliance, such as the type of violation involved, the remedial measures required, various country-specific control variables, and a time dimension. For the purpose of providing empirical evidence for the main argument in this book, however, which is primarily qualitative in nature, additional cross-sectional and time-series analyses are not the most auspicious strategy.
Instead, in the pages that follow I employ a case-study approach to investigate the particular behavioral logic guiding compliance decisions by liberal democracies. To avoid charges of selection bias with respect to the judgments covered, I opt for a comprehensive approach that examines all ECtHR judgments issued against two liberal democracies—the United Kingdom and Germany—until the cut-off year 2010. The next chapter presents the hybrid constructivist-rationalist theory of compliance that I argue best explains the empirically observable patterns of compliance by liberal democracies with the judgments of the ECtHR. In the two main parts that follow, I then examine the resulting hypotheses against the empirical evidence of the remedial responses adopted by the United Kingdom (Part I) and by Germany (Part II), organized along the lines of the principal types of remedial measures required to comply with a given judgment and the issue area in which the violation falls. The conclusion summarizes and provides a final assessment of the sustainability of the theory in light of the empirical evidence.