A comparison of the European and Inter-American Courts of Human Rights.

I.   INTRODUCTION                                       35
II.  DEFINING AND EXPLAINING COMPLIANCE:
     CLARIFYING THE CURRENT DEBATE                      39
III. COMPARING REGIMES:
     CHECKLIST VS. DELEGATIVE COMPLIANCE                43
     1. MEASURING COMPLIANCE                            44
     2. MEASURING COMPLIANCE: THE INTER-AMERICAN COURT  45
     3. MEASURING COMPLIANCE: THE EUROPEAN COURT        51
IV.  PARTIAL COMPLIANCE: GENERAL PATTERNS               55
     1. THE INTER-AMERICAN COURT                        56
     2. THE EUROPEAN COURT                              66
     3. TYPES OF PARTIAL COMPLIANCE:
        THE IACHR AND ECTHR COMPARED                    77
V.   CONCLUSIONS                                        83

I. Introduction

Expectations about the level of state compliance with international human rights norms vary widely, but tend to cluster around the extremes of high compliance or low compliance. Legal scholars such as Louis Henkin (1), and Abram Chayes and Antonia Chayes, (2) suggest that most states obey most laws most of the time. In the same vein, some political scientists suggest that when international institutions socialize states, the result is either stable compliance with international rules or an even deeper transformation of state interests to match international norms. (3) In contrast, other scholars are skeptical. Some suggest that international institutions are little more than cheap talk that reflect existing state preferences and practices. (4) Any observed compliance is the result of states designing easy rules that they already follow. Other scholars stress instead the large gaps between international rules and state behaviour, and argue that the independent effect of international institutions is negligible. (5) In the first skeptical version, international institutions are epiphenomenal, and in the second, they are redundant or even useless.

In this article, we will conceptualize and explore the middle ground between these opposing positions. Just as scholars of domestic governance systems have broken down the dichotomy between democracy and autocracy by examining imperfect democracies and varieties of autocracies, (6) we aim to break down the dichotomy between compliance and noncompliance by exploring partial compliance. While scholars are undoubtedly aware of the possibility of partial compliance, many write as if conditions of partial compliance are way stations on the path to full compliance. (7) Often, scholars suggest that the socialization of states by international institutions is a transformative experience, leading to the convergence of state interests. (8) In both views, partial compliance is thus merely transitional. No doubt, both patterns hold in many cases. We suggest, however, that partial compliance appears to be a relatively stable end point in many other cases--one that is more common than is often supposed.

This article will study the role of partial compliance in the context of the Inter-American Court of Human Rights (IACHR) and the European Court of Human Rights (ECtHR), (9) two adjudicative bodies with significantly different compliance regimes. The differences between the two courts are methodologically constraining, but they allow us to use the "most different cases" research strategy. According to this strategy, if an empirical phenomenon is present in political systems that are strikingly different, this increases confidence that it may be present in other systems as well. (10) The presence of partial compliance at both the IACHR and the ECtHR is therefore highly suggestive of the important role that partial compliance plays in international adjudication.

Through the cases of the IACHR and the ECtHR, this article explores the middle ground of partial compliance, defining its extent and contours and conceptualizing different types of partial compliance. The argument proceeds at two levels of analysis. At the regional level, we identify two different compliance regimes that nevertheless display shared outcomes of partial compliance. At a state level, we identify four different types of partial compliance--types that generally can be observed in both regions. A clear view of the extent and contours of partial compliance seems necessary before scholars can turn to the subsequent step of explaining those patterns, a step beyond the scope of the current article.

The next section will define our notion of partial compliance and place it within the broader International Relations literature on compliance and the effectiveness of legal norms. The third section will then outline how Europe and the Americas have built two quite different adjudicative institutions, with two contrasting compliance regimes. (11) The IACHR orders a series of clear, specific steps and then vigorously monitors compliance itself, often through multiple state-specific compliance orders. We will refer to this as a regime of "checklist compliance". In contrast, the ECtHR exercises what we call "delegative compliance", whereby its rulings will identify a violation, but not make orders on how to end the violation, compensate for its effects, or prevent future infringements. These decisions on the modalities of compliance are delegated to states and monitored by the Committee of Ministers, which itself is a political body. This difference between the two regimes is important from a methodological perspective: it affects the ways in which partial compliance can be observed and measured in each system and shapes the possibilities for comparison between them.

The fourth section will argue that despite these methodological constraints, partial compliance is observably widespread in both regimes. In the IACHR context, partial compliance is more common than either total compliance or non-compliance. We find that 83% of the cases for which we have compliance reports (n=81) can be coded as partial compliance, with complete non-compliance at 11% and full compliance at 6%. From these patterns, we also find preliminary clues as to the sources of partial compliance. The data supports the common sense notion that compliance is higher when it is at its least complicated. For example, states are more likely to comply with judgments requiring monetary compensation than with those requiring action, and the broader the action, the less likely they are to do anything. We also find some evidence that partial compliance can be more frequent than full compliance when the Court is less diligent in its monitoring; this supports the hypothesis that the Court can increase compliance through careful follow-up to initial state inaction. (12)

In the ECtHR context, we find that most states do comply fully with most judgments. Even in the three most recent years, arguably the busiest in the Court's history, full compliance was achieved in nearly 700 cases per year. At the same time, however, about twice that number of new judgments was being handed down each year. (13) It will be argued that there is far more scope for partial compliance in the last decade due simply to the great leap in cases "pending", that is, where the Court has found against a state, but the state has not yet shown evidence of full compliance. Of over 8,000 such cases that existed at the end of 2009, over 40% had been pending for over two years and some for much longer. Only about one third of the highest profile cases are closed each year (compared to about one half for all cases), and in 15 of 18 issue types, still pending cases outnumber cases closed in a given year. These are necessarily very indirect measures, and we treat them with caution. Looking in more detail at case studies of all leading cases in four countries, however, we find clear and direct evidence that 85 of the 90 pending cases we investigate are ones of partial compliance, rather than full compliance or non-compliance. We find also that the monitoring mechanism of the ECHR has undertaken substantial follow-up measures in the form of interim resolutions directed at states in 38% of the cases pending at the time of data collection. We show that virtually all of these are cases of partial compliance.

In the fifth section, we will explain how the patterns of partial compliance observed in both Europe and the Americas can be sorted into four types (that are not mutually exclusive): 1) split decisions, where states do some of what a court orders but not all; 2) state substitution, where states sidestep a court order, implementing an alternative response to the decision; 3) slow-motion, where states move so slowly that it is difficult to say that full compliance occurs; and 4) ambiguous compliance amid complexity, in which states face particularly daunting or demanding tasks.

In the concluding section, we will summarize and briefly elaborate on our major findings: despite the stark differences between the European and American human rights systems, states in each region consistently engage in partial compliance. They do so despite repeated efforts by international institutions to bring them to full compliance and despite the fact that their prior behaviour suggests they would prefer non-compliance. We approach compliance from a number of different analytical perspectives--examining compliance by state, by issue area, across time, and so forth--but always find remarkably durable levels of partial compliance.

II. Defining and Explaining Compliance: Clarifying the Current Debate

This section distinguishes between compliance and effectiveness, showing how the two differ, but then identifying a class of cases (examined in this paper) in which compliance is a useful proxy for effectiveness. In subsequent sections, this will allow us to proceed to the trickier task of identifying empirical examples of partial compliance. We will also in this section distinguish our concept of partial compliance from important recent work on the "reception" of ECtHR rulings, and show how it improves on and often complements the existing literature. Specifically, the major existing theories used to explain compliance can all be deployed with our concept of partial compliance, which should greatly facilitate subsequent tests with the more subtle dependent variables we suggest.

It is first important to distinguish between compliance and effectiveness. (14) For Kal Raustiala, (15) compliance is conformity between behaviour and a legal standard. Compliance could be the result of the rules and enforcement efforts or it could be sheer coincidence. To say an actor complies with the rule is not to imply that the rule caused the behaviour. Effectiveness, in contrast, is the degree to which a legal rule or standard induces the desired change in behaviour. Thus, international rules display some degree of effectiveness even when compliance is low (by inducing behavioural changes in some but not all), and international rules with high compliance can be totally ineffective (because they were drafted to fit preexisting behaviour, for example). (16)

While conceptually very useful, this distinction does put enormous data demands on researchers when, in the normal course of events, they try to distil cases of efficacy from the wider sample of compliance. There is, however, one set of circumstances in which, without very demanding assumptions, we can get a class of cases in which we can eliminate the category of pre-existing behaviours and thus treat compliance and effectiveness as rough equivalents. That class of cases is court rulings against states for violating their treaty obligations. When a country persists in behaviour long enough for an international court to rule against that country's practices, and the country subsequently changes its practices, we assume that the court's ruling helped trigger the change in behaviour, even if other factors may also have been important. Likewise, when a court orders a specific behaviour, such as the payment of monetary damages, to a particular individual, and the state complies, one can assume that the court's order played a role in the state's compliance. Compliance in these circumstances is very unlikely to be the result of chance: most international litigation takes years and costs states significant money to defend; it is therefore reasonable to assume that the state prefers to persist in the behaviour being challenged in court. Hence, any resulting behavioural changes after an adverse court ruling can suggest court effectiveness. This creates a class of cases where instances of compliance will be coextensive with those of effectiveness and where effectiveness can therefore be objectively measured through the proxy of state compliance.

If our approach shows compliance to be functionally similar to effectiveness in these cases, it also is narrower than a second competing dependent variable, namely the domestic "reception" of norms advanced by international courts. For example, a team led by Helen Keller and Alec Stone Sweet (17) studied the reception of ECtHR norms in nine matched pairs of European states. The study considered very broad questions about the way in which Court decisions were treated by all branches and levels of government, and even extended to whether the Court was a subject of wide media coverage and an important topic in law curricula. (18) In reporting its findings, this important study clearly recognized ECtHR frustrations for both ECtHR judges and the Council of Europe bureaucracy that some judgments are not fully complied with. With that said, its authors did not use the category of partial compliance, which focuses more specifically on state behaviour than the broader observation of domestic attitudes and institutional reform that is involved in the study of reception. We thus see our article as complementary to the literature on reception.

These observations about what compliance is (and is not) lead directly to a larger question: Why do states comply with international rules? Scholars have identified a variety of factors that push states either towards compliance or away from it. Yet they almost always conceptualize compliance as a fairly simple dichotomous variable: either compliance occurs, or it does not. We wish to provide a constructive way forward by conceptualizing compliance as a more complex phenomenon where states commonly engage in varying types of partial compliance. In the process, we suggest that each of the causal factors highlighted by compliance scholars could actually produce partial compliance rather than full compliance or no compliance. Partial compliance is thus a useful analytical category, complementary to the vast literature seeking to explain compliance.

Scholars have highlighted three types of factors influencing state compliance: international enforcement, management, and domestic politics. International enforcement refers to the imposition of penalties or rewards, both material and social--though scholars who focus on enforcement also tend to emphasize and prioritize the use of material rewards and sanctions. (19) In an influential piece, Beth Simmons (20) suggested that states are likely to comply with international commitments in order to maintain their good international reputations for predictable and law-abiding behaviour. States that enjoy such reputations are likely to be rewarded through mechanisms such as increased investment, while states lacking such reputations are likely to be punished in an opposite fashion. In the area of human rights, Emilie Hafner-Burton (21) has argued that international human rights agreements are more effective when states tie compliance to specific material incentives, such as trade integration. Many scholars focusing on international enforcement tend to combine this mechanism with either management (22) or domestic politics, (23) reviewed next.

The difficulty is that important international actors have mixed motives and records with respect to promoting human rights elsewhere. Multinational corporations, for example, do not necessarily want the developing countries in which they invest to comply fully with rights guarantees for labour unions. Powerful states like the United States reward allies who engage in anti-terrorist operations, which may involve less than full compliance with human rights norms. Hence, many states face mixed incentives for human rights compliance and may decide that partial compliance is optimal.

Another approach, most closely associated with Chayes and Chayes, (24) emphasizes the ways in which management problems obstruct compliance. Management problems are related to the nature of the international rules and the capabilities of states, rather than state motives and the rewards or punishments linked to rule-following. In some cases where non-compliance appears widespread, a closer examination may show that the international rules are quite ambiguous, which makes it difficult for states to comply with one particular interpretation of those rules. Another management problem occurs when states lack the technical expertise or economic capacity to implement international rules. Alternatively, non-compliance may simply be a timing issue: many international rules are quite difficult to implement and require a fair amount of time. This is an important issue with the ECtHR, which tends in official documents to attribute partial or non-compliance entirely to issues of timing, asserting that ultimately, all of its judgments are complied with. (25) Finally, an important management debate has revolved around the question of which types of international courts are better for compliance. (26)

Rather than leading to complete non-compliance, however, these same factors could easily produce partial compliance. Low-capacity state bureaucracies struggling to understand and comply with international rules are unlikely to be utterly incompetent and devoid of resources; such transparent ineptitude seems unlikely to survive politically. Chayes and Chayes do not paint such a cartoonish picture of bureaucracies. Instead, they view bureaucracies as sincere but under-resourced, especially in developing countries. Such circumstances are likely to produce half-measures that might better be classed as partial compliance rather than complete non-compliance.

A third approach to compliance focuses on domestic politics. For some scholars, compliance is a matter of calculating the costs and benefits of changing policy. Where policy changes are difficult for political actors because there are domestic incentives to defect from the international rules, compliance is likely to be low. (27) Relatedly, compliance might be the result of local factors that are difficult to observe, such as "political will." (28) In this view, compliance is likely to result from pre-existing domestic factors that led states to commit to particular international rules in the first place. Hence, compliance is not the result of international rules; rather, states that wish to behave in a particular way are likely to create and accept international rules that codify that behaviour. For example, Gerda Falkner and Oliver Treib (29) emphasize the importance of domestic "cultures of compliance," with the European Union member states falling into four compliance patterns. A more dynamic, actor-oriented approach suggests that compliance is more likely as domestic actors that favour compliance gain greater influence in the government. (30) Alternatively, other scholars have focused on domestic political structures, such as the nature of the linkages between international courts and domestic judicial systems or the robustness of domestic civil societies. (31)

On any given international compliance issue, the preferences of various powerful domestic groups are unlikely to be homogenous, and thus compliance is likely to be a matter of political contestation. Decision-makers attempting to maximize support and avoid costly political battles should be interested in a compromise position that embraces partial compliance. Alternatively, policy changes in most countries are relatively common as different coalitions gain and lose power; fluctuating policies can easily produce partial compliance as new regulations are layered on top of old.

This is a large and complex literature, yet much of it fails to conceptualize the key dependent variable, the nature of compliance. Compliance is often treated as a dichotomous term, but we wish to add nuance by exploring the middle ground of partial compliance. Moreover, many of the factors that scholars hypothesize to cause either compliance or non-compliance could just as easily produce partial compliance. International enforcement, state bureaucratic capacity and domestic political systems and actors do not uniformly point toward compliance or noncompliance. The competing pressures within each of these categories could very well produce partial compliance as decision-makers attempt to balance the various actors with heterogeneous compliance preferences and as they attempt to accomplish tasks with limited resources. The evidence presented in subsequent sections suggests that although partial compliance occurs to varying degrees and in different forms in the ECtHR and IACHR compliance regimes, in both it has been the most common outcome.

III. Comparing Regimes: Checklist vs. Delegative Compliance

We argue in this section that the ECtHR and IACHR exemplify two different types of compliance regimes, which we refer to as delegative compliance and checklist compliance, respectively. After briefly introducing these terms, we discuss the difficulties of measuring compliance in the context of the only other cross-regional comparative study of which we are aware. We then go into some detail on our compliance measures for the two different regimes. While the differences between them mean that we must measure compliance quite differently in the two cases, it is still possible to identify a range of compliance in each region and to make some judgment as to the distribution of compliance within this range.

The IACHR gives offending states a list of highly specific steps that they must undertake as remedies to adverse judgments. Each of these steps is known as a compliance order. Any given judgment contains what amounts to a checklist of multiple specific orders. The Court then uses that checklist to follow up on the state's behaviour in subsequent years by issuing a series of compliance reports in which it carefully examines the list of steps originally given to states. It makes a clear and specific judgment about whether a state has complied with each and every step and then continues to issue such reports indefinitely until compliance is complete.

The ECtHR works in a very different fashion. Though the Court is empowered by Art. 41 ECHR to specify precise monetary payments by states in order to provide "just satisfaction," the Court has no power to make other remedial orders that would legally bind the state in question. (32) Rather, the state, once notified that it is the object of an adverse ruling, is expected to pay the just satisfaction and then conceive and execute other steps to bring itself into compliance, both in the short and long term. (33)

We thus contrast IACHR "checklist" compliance with ECtHR "delegated" compliance. This broad distinction between the courts has an important implication for how partial compliance arises under each regime. The kinds of partial compliance observed in each case are quite different. For the ECtHR, partial compliance often emerges from cases in which states design remedies that take less than full account of the Court's judgment, a point long acknowledged by Court insiders. (34) As already noted above, the extent to which state remedies effectively respond to Court judgments is entirely a matter for the Committee to judge. (35) These Committee judgments often find fault with state responses. For the IACHR, partial compliance emerges when states accomplish some items on the checklist but not others. Each of these courts is thus likely to generate partial compliance in very different ways. Where missing items from the checklist might lead to protracted rounds of "institutional nagging" by the IACHR, this trend is less pronounced (though hardly absent) from the ECtHR setting, in which partial compliance is more often in the eye of the beholder (but where the Committee has the final say).

1. Measuring Compliance

These differences in court practices make it impossible to measure partial compliance in the same way in both regions. Other problems also beset attempts to measure compliance, such as the difficulties in judging whether a behaviour is consistent with Court preferences, whether slow progress counts as partial compliance or something else, and who gets to judge whether compliance has occurred. Yet these difficulties do not mean that scholars should give up on the enterprise. A large number of social science concepts are difficult to measure, especially in cross-regional perspective.

Perhaps as a result of these difficulties, systematic studies of state compliance with the rulings of international courts are in short supply. Eric Posner and John Yoo (36) have authored the only study we have found that explicitly measures compliance rates across different international courts. (37) When they reviewed the data on compliance with the IACHR in 2004, they found only one case of full compliance with a court ruling, though their measurement of compliance is fairly unclear and seems to be drawn in part from reading secondary sources. (38) Taking partial compliance into account, they gauged overall compliance with the IACHR to be 5%. However, they also found that compliance with judgments ordering monetary compensation was somewhat higher, at 23.6% full compliance. Posner and Yoo could not find good compliance data on the ECtHR. Although they did mention one measure, taken from the Court's own Survey of Activities, suggesting that compliance with ECtHR judgments (as measured by domestic law adjustment in the wake of an adverse decision) hovered around 64% between 1960 and 1995, (39) they expressed doubt that compliance is as high as this.

In contrast to Posner and Yoo, we will present our criteria and identify our data sources for judging various levels of compliance in some detail. We explore various levels of compliance--non-compliance, partial compliance and full compliance--at different levels of analysis: within particular cases, at the case level, and aggregating across cases. We also take different analytical cuts at the question of compliance by examining levels of compliance according to factors such as year, state, right violated, and the type of remedy ordered by the Court.

2. Measuring Compliance: The Inter-American Court

In order to measure compliance with the IACHR, it is first important to understand the workings of the Court. (40) It issues several forms of jurisprudence: decisions and judgments on contentious cases, advisory opinions, provisional measures, and reports on compliance with judgment. The compliance reports detail the measures taken by states in response to the Court's decisions on contentious cases. We rely primarily on those reports for our understanding of compliance levels. We ignore provisional measures and advisory opinions because the Court does not issue compliance reports on those rulings (relying instead on the Commission in the case of provisional measures) and hence compliance information is less systematic and not comparable with the data we consider here.

The Court's decisions and judgments are broken down into three categories: decisions on preliminary objections, decisions on the merits of the case, and decisions on reparations in the case. From its inception through June 2010, the Court issued decisions and judgments on 123 cases. During the Court's early years, however, the caseload was extremely light. The Court was officially established in 1979, but did not receive its first case until 1986; it issued its first judgment, decisions on preliminary objections in Velasquez-Rodriguez v. Honduras, in 1987. (41) More recently, the Court has carried a significantly higher caseload, processing more than a dozen new cases each year. (42) At the end of 2009, the Court was considering decisions on 14 new contentious cases and was monitoring compliance with 104 previously decided cases. (43)

The Court's procedures have changed in various ways over the years and so have the behavioural patterns of actors involved with the Court, including states, the Inter-American Commission on Human Rights, victims, and victims' advocates. (44) One interesting development is that states have increasingly admitted violations before the Court and taken partial or full responsibility for their actions. Such behaviour is analytically distinct from compliance: just because a state acknowledges responsibility does not mean that it will make the reparations ordered by the Court. Once states acknowledge responsibility (in those cases in which they do), the Court still proceeds to issue reparations judgments and to monitor compliance.

Each reparation decision includes several paragraphs ordering states to take a particular action. Each paragraph orders a discrete action. For example, one paragraph will order the payment of material damages; a separate paragraph will order the payment of moral damages; and a third paragraph will order the reimbursement of Court costs and expenses to the victim. We label each of these discrete paragraphs a "compliance order." A compliance order is our basic unit of analysis for the IACHR. To illustrate the nature of these orders, we replicate here verbatim the reparations section of one relatively recent Court decision of moderate complexity: (45)

6. The State must secure, within a reasonable time; that the domestic proceedings instituted in order to investigate and punish those responsible for the events in the instant case be operative, as set forth in paragraphs 245 to 248 herein.

7. The State must, within the term of six months, publish at least once in the Official Gazette and in another nationwide daily newspaper, Chapter VII-Proven Facts-of the instant Judgment, without its footnotes, as well as the operative paragraphs herein, as set forth in paragraph 249 herein.

8. The State must keep developing an education and training program for staff in health care, psychiatry, psychology, nursing, and for any person involved in mental health services, in particular, covering the principles that govern treatment to patients with mental illness, according to international standards and the provisions of the instant Judgment, as set forth in paragraph 250 herein.

9. The State must pay in cash to Albertina Viana-Lopes and Irene Ximenes-Lopes-Miranda, within the term of one year, as compensation for pecuniary damage, the amount fixed in paragraphs 225 and 226 herein, as set forth in paragraphs 224 to 226 herein.

10. The State must pay in cash to Albertina Viana-Lopes, Irene Ximenes-Lopes-Miranda, Francisco Leopoldino Lopes and Cosme Ximenes-Lopes, within the term of one year, as compensation for non-pecuniary damage, the amount fixed in paragraphs 238 herein, as set forth in paragraphs 237 to 239 herein.

11. The State must pay in cash, within the term of one year, as costs and expenses incurred in the domestic proceedings as well as in the international proceedings under the Inter-American system of protection of human rights, the amount fixed in paragraph 253 herein, which shall be delivered to Albertina Viana-Lopes, as set forth in paragraphs 252 to 253 herein.

According to Art. 68(1) of the American Convention, state parties must comply with all Court rulings in all cases in which they are parties. The Court typically suggests that states have six months to one year from the date reparations are issued to comply with the Court's judgment. (46) Yet unlike in the European system, the American Convention's adjudication mechanism does not include a formal mechanism for monitoring state compliance. The Court has thus taken it upon itself, beginning in 1996, to issue periodic reports monitoring compliance with its judgments. In November 2003, after this practice was challenged by state parties, the IACHR held that the issuance of compliance reports was implicitly within its authority as granted by the American Convention. (47) To evaluate a state's compliance, the Court asks victims' representatives, the Inter-American Commission of Human Rights, and the state to submit reports regarding the state's actions. (48) In some cases, the Court may request a private hearing of those same actors to determine state compliance. (49)

The IACHR issued its first compliance report on September 10, 1996. As of June 23, 2010, we believe the Court had published 191 compliance reports, of which we located 184. (50) Not all cases have compliance reports. In some cases, the Court acknowledged the state's preliminary objections and dismissed the case. Other cases are still pending merit and reparations judgments. Many of the cases whose reparations and judgments have been issued recently also do not have compliance reports.

We have compliance reports for 81 cases as of June 23, 2010, the cutoff date for our data. The Court issued 703 compliance orders for those 81 cases. Most of these cases have more than one compliance report. Some, such as the Loayza-Tamayo v. Peru case, (51) have as many as seven compliance reports. (52) The Court's additional compliance reports in each case do not impose new compliance orders, but rather evaluate state compliance with the original orders. We have been unable to identify a pattern as to how soon after a judgment and how often the Court publishes compliance reports. Although the length and specificity of compliance reports has increased through the Court's history, each report follows the same general format. In remarkable detail, the Court specifically reports whether or not the state has complied with each and every one of its compliance orders, paragraph by paragraph.

In the analysis that follows, we simply adopt the Court's perspective on whether compliance has occurred on each of its compliance orders. (53) While this introduces the Court's own possible institutional biases into our coding, the Court has access to a wealth of information about compliance that is either unavailable to others or would be prohibitively costly to gather independently. This includes transcripts of victims' statements, and state correspondence with the Court. We believe the Court's informational advantage justifies the risk of introducing its biases into the coding.

The Court reports compliance clearly in each compliance report for each of its orders. For each order, we code compliance dichotomously: comply or not. This is a methodological simplification because compliance with each discrete order could also be partial. The state could, for example, pay some of the individuals specified by a given compliance order but not others. The Court in fact sometimes does code states as having partially complied with discrete compliance orders. We could not be sure the Court consistently used this category, however, and hence we opted to code partial compliance as non-compliance at the level of compliance orders. We thus probably understate the extent of partial compliance because we treat partial compliance in this part of the analysis as an aggregate quality that occurs only as states combine full compliance with non-compliance across compliance orders within a given case. In reality, partial compliance also occurs at the lowest level of analysis in these discrete compliance orders as well, as we discuss in section V below. Additionally, we coded no compliance when the court requested further information because the state had not submitted a report, which occurred quite frequently. We believe this coding decision is justified because it seems likely that if a state has complied, it would want that fact to be known and would have submitted the relevant information. For each case, then, a state might be in full compliance, complete non-compliance, or partial compliance. A state will be in partial compliance if it complies with some of the Court's discrete compliance orders, but not all.

We then can analyze compliance in a variety of ways: by date of the case, by the number of compliance reports issued within a given case, by right violated, by state, and so forth. Most of these require little comment. One analytical category requiring more detail is the type of action required by the Court, what we call "type of judgment". Jo Pasqualucci (54) has posited a typology along this dimension for the Court's compliance orders, which we replicate here. We coded each compliance order according to Pasqualucci's typology, which is largely self-explanatory. The most confusing categories are "enjoyment of right violated" and "take action or refrain from action". The difference is that the former is aimed at restoring rights to the victim while the latter attempts to compensate for the violation. (55)

Type of Compliance Order          Example of judgment

Enjoyment of right or freedom     The State shall nullify any court,
violated                          government, criminal or police
                                  proceedings there may be against
                                  Luis Alberto Cantoral Benavides
                                  in connection with the events in this
                                  case and shall expunge the
                                  corresponding records ...

Remedy the consequences of the
violation

Investigate, identify, publicize  The State shall adopt ... all
& punish                          measures necessary to identify,
                                  prosecute and punish the physical
                                  perpetrators and instigators of the
                                  violations committed against Mr.
                                  Bernabe Baldeon-Garcia ...

Amend, repeal, or adopt domestic  The State should adopt the
laws or judgments                 legislative measures and
                                  any other kind of measures
                                  as necessary to adapt the
                                  Guatemalan legal system to the
                                  international standards on human
                                  rights, and give full effect
                                  to said standards at a
                                  domestic level ...

Take action or refrain from       The State shall name, within one
taking action                     year following notice of
                                  this Judgment, a street, park
                                  or school in the memory
                                  of Mr. Bernabe Baldeon-Garcia ...

Apologize                         The State shall make, within six
                                  months following notice
                                  of this Judgment, a public
                                  apology and acknowledgment of
                                  its international liability
                                  regarding the violations
                                  referred to herein, in
                                  the presence of the highest-raking
                                  State authorities ...

Pay fair compensation

Material damages                  The State shall pay... all members of
                                  the Baldeon-Yllaconza family, within
                                  one year, the compensation for
                                  pecuniary damage established ...

Moral damages                     The State shall pay ... all members
                                  of the Zaldeon-Yllaconza family,
                                  within one year, the compensation for
                                  nonpecuniary damage established ...

Cost and expenses                 The State shall pay, within one
                                  year, the costs and expenses
                                  incurred in domestic courts and in
                                  the international
                                  proceedings carried out within the
                                  Inter-American System for the
                                  Protection of Human Rights, pursuant
                                  to the amount established ...

Source: Pasqualucci, supra note 40; IACHR, "Monitoring Compliance",
supra note 50, reports for Baldeon v. Peru (February 7, 2008),
Cantoral-Benavides v. Peru (February 7, 2008), Bamaca Velasquez v.
uatemala (November 27, 2003).

Compliance with an order may take 10 years and the Court may have judged that non-compliance occurred in nine of the previous years, but once the Court determines that compliance has occurred, then we count it as compliance. Such a method possibly overstates compliance because some observers might prefer to consider delayed compliance as partial compliance or even non-compliance if the delays create additional problems.

This issue is especially applicable to older cases, and so we create a second measure labelled "resistance" to reflect the extent to which a state delays in complying with an order. The straightforward interpretation of our resistance measure is the average number of times the Court has found states failing to comply with each compliance order. Resistance is calculated as the number of instances of non-compliance divided by the number of compliance orders in that case. An instance of non-compliance occurs when the Court issues a compliance report and finds the state has not complied with the order. Because any given order could repeatedly be subject to such a finding, average resistance rates can easily go above one, and could, over time, become quite large. For example, if the Court gives 10 compliance orders in a given case and then issues one compliance report in each of the subsequent five years and the state repeatedly fails to comply with all of orders, then the resistance level is five (50 instances of non-compliance divided by 10 orders). If in the sixth year the state complies with all of the 10 orders, its compliance rate would be 100% but its resistance level would still be five. In this way we can incorporate long delays in compliance into the analysis. The resistance measure is a useful counterpart to the compliance measure in part because it can help reveal whether "full compliance" should be qualified by the observation that significant delays occurred on the path to full compliance.

3. Measuring Compliance: The European Court

The ECtHR, far busier than the IACHR, has a very different compliance regime. Most importantly, there is no "checklist" for states to comply with. Instead, we call the ECtHR regime one of "delegated compliance," and we describe it below. While it is more difficult to distil data about partial compliance from ECtHR and Committee of Ministers publications than it is for the IACHR, it is becoming increasingly possible. This section discusses our measures.

The ECtHR was established in 1959 by the ECHR and hears cases against states parties. As of 2010, the Convention and Court had 47 members--from plaintiffs who have exhausted all available domestic remedies. Crucially, the Court can only rule on whether an individual has had his or her rights violated by a state party to the Convention. Upon the publication of such judgments, the states, according to Art. 46, "undertake to abide by the decision of the Court in any case to which they are parties." In practice, this means the Court does not overrule the decisions of domestic courts, invalidate national laws, or even make specific orders for legislative reform. Instead, states generally must reason backwards from the violation to understand the appropriate remedy in a specific case, and the actions required to avoid similar future violations. The Court thus draws a line between finding that an individual's rights have been violated, and commenting on specific state practices. Put differently, the Court, in the words of one of its presidents, has not been "prescriptive" in its judgments. (56)

Several particular characteristics of the ECtHR impact patterns of state compliance. For example, unlike the highest courts in domestic legal orders, the ECtHR has no power to remand its cases to lower courts. (57) It is also quite unlike the European Court of Justice (ECJ), which has developed linkages to other actors that have become more uniformly substantial over time. (58) While the Council of Europe--the ECtHR's parent organization--tries to inform national officials, including judges, police, and bureaucrats, of ECtHR jurisprudence, formal links between the ECtHR and domestic courts vary widely across Europe. (59) Moreover, whereas rulings of the ECJ are generally superior to domestic law, European states have a wide variety of approaches to ECtHR law. For example, while ECtHR jurisprudence is superior to the national constitution in the Dutch case and has become co-equal with it in the Austrian case, in many other states the ECtHR's judgments hold a position in the legal hierarchy somewhere between constitutional and ordinary statutory law. (60) As the number of cases grows, the Court has suggested that elevating the legal status of its decisions would spare states the difficulty of "complex and lengthy legislative work." (61) Nevertheless the Court continues to presume that individual states are best suited to devise remedies compatible with their different traditions and legal principles.

Compliance with ECtHR judgments is monitored by the Committee of Ministers, comprising states parties' Ministers of Foreign Affairs (or their deputies). After an adverse ruling against a state, the Committee "invites" the state to report on the measures it has taken to address violations found by the Court. The Committee may, "where appropriate, adopt decisions or interim resolutions to express concern, encourage and/or make suggestions with respect to execution." (62) The Council of Europe's Parliamentary Assembly also issues rapporteur reports that regularly examine state implementation of ECtHR judgments, but the Committee remains the principal actor in monitoring compliance. (63)

When the Committee is convinced that just satisfaction has been paid and that appropriate individual and general measures put in place, the Committee will close the case. (64) Typically, just satisfaction takes the form of payment to the victim, often to remedy a combination of pecuniary losses, non-pecuniary losses (e.g. psychological damages), court costs, and interest payments, all of which the Court specifies. (65) In addition, the Court also invites the state to take either individual and/or general measures, which, with a few exceptions noted below, are designed by the states. (66) Individual measures are meant to put the victim into the same position enjoyed prior to the violation, while general measures are intended to prevent future cases of a similar nature from arising. For example, an individual measure might be a state decision (not an ECtHR order) to release an individual from jail, while a general measure might be a state decision to amend a law or practice that has resulted in findings of a violation in the past.

The complexity of measuring compliance with these different measures varies widely. Whereas just satisfaction payments are relatively easy to monitor, individual and general measures are a source of more ambiguity. As ECtHR caseload expands rapidly, the Committee is not able to track individual measures as carefully as those for just satisfaction. (67) The Committee recently began issuing annual reports, and the first three (2008-2010) all dwell extensively on this growing dilemma and sound warnings about, for example, the danger of states taking a "minimalist approach" to compliance that an overextended Committee may not be able to monitor adequately. (68) The Committee also has strong incentives to make sure that states take appropriate general measures, since these are critical preemptive tools against the flood of applications that arrive each month. The Committee does catalogue the individual and general measures chosen by states, and its monitoring reports on individual cases convey these steps in substantial detail. As we show below, the Committee often refuses to close cases even when just satisfaction has been paid and some individual and/or general measures have been taken by the states. The key point is that while compliance is "delegated" to the states, it is hardly the case that it is therefore straightforward for states to comply, a claim we demonstrate in the next section.

The growth of the Court's docket has made the Committee's monitoring tasks much harder. Like the IACHR, the ECtHR caseload was relatively light in its initial years. In its first two decades in existence, the Court ruled on only 84 cases. (69) Far more than the IACHR, however, the ECtHR caseload has exploded in recent years, reaching over 1000 rulings per year by 2004 and with a new case load that averaged just over 1300 per year from 2004 through 2009. Of these new cases, about 200 "leading cases" each year raised novel issues, while the remainder were so-called "clone" cases. (70) This surge has hampered both the Court's and the Committee's ability to monitor compliance. This is especially the case with individual measures that are harder than just satisfaction to oversee and yet less likely than general measures to provide docket relief for the Court. Thus, it is here in the details of individual cases that states may still have very substantial freedom to design their own remedies, including in ways that may not be in the spirit of the Court's judgment.

Testing this proposition is not easy: ECtHR compliance data is compiled in ways very different from that of the IACHR. Unlike the IACHR, there are no multi-pronged judgments that allow us to track several issues within a single case. Instead, there are single judgments that an individual's rights either have or have not been violated. Committee data on compliance have recently been made much more readily available, however, in the above mentioned series of Annual Reports begun in 2007. For cases where violations were found, Committee databases include three major categories--closed cases, pending cases, and interim resolutions--analyzed more fully below:

I. We look first at closed cases for which general and/or individual (71) measures were taken by the states and for which the Committee has been satisfied by state remedies. We code these cases as full compliance because the Committee has explicitly stated that it is satisfied with the state's response to the judgment of the Court.

II. Second, we look at pending cases of adverse judgments transmitted by the Court to the Committee for monitoring. These are a complex amalgamation of cases. In a few cases, where states refuse even to pay just compensation or take any individual or general measures, we code them as cases of non-compliance. If just compensation is paid late, however, we count this as partial compliance. Such cases amounted to 7% of cases in 2007, 5% in 2008, and 11% in 2009. (72) Yet many pending cases are ones in which full compliance will later result but where such full compliance has not been achieved at the time of the data being reported, so we also report on the length of time these cases have been pending. We then look to the Committee's Annual Reports to distil their views of which patterns of pending cases warrant further investigation. This exercise yields important clues about the extent and duration of partial compliance, but because it is based on the Committee's selection of pending cases (and not the full universe), we turn next to a database on execution of judgments that does allow us to code full, partial, and non-compliance from a different sample in the universe of pending cases. We count as non-compliance only those (very few) narratives that contain no evidence of compliance. Partial compliance can be attributed to the bulk of the cases, where there is clear evidence that states have taken some constructive steps but where the Committee asks for evidence of further action. Full compliance is coded for those older pending cases where the case narrative strongly suggests that adequate steps have occurred. Finally, some very new pending cases contain too little information for coding. We note these and then drop them from further consideration.

III. The third category, interim resolutions (IRs), gives a more accurate sense of the overall size of the partial compliance cases. Where data on pending cases requires the complex sorting just noted (mostly because many new pending cases will achieve full compliance relatively quickly), IRs typically occur in cases of longerterm partial or non-compliance. IRs are formal communications from the Committee to the states asking for evidence that a prior judgment has been complied with. (73) In practice, they occur when the Committee has good reason to suspect that full compliance is not imminent. We demonstrate that states almost invariably take some individual and/or general measures, and thus we treat IRs as evidence of partial compliance. The number of case with IRs is substantially lower than the numbers of pending cases but are far more likely to include problematic cases of partial, along with a very few cases of non-compliance.

IV. Partial Compliance: General Patterns

In this section, we survey the available data on compliance in the Americas and Europe. Our primary intent is to provide a nuanced and sophisticated yet big-picture overview of the shape of partial compliance in both regions. Hence, especially in the inter-American case, where more data is available, we analyze compliance through different categories, such as whether compliance varies by the type of order the Court issues, over time, or by state. With some exceptions, we find relatively little variance within these categories, suggesting that multiple paths lead to partial compliance. In the ECtHR, we find that partial compliance is extremely widespread among long-term (more than two years) pending cases. European states very often comply fully and quickly, and they rarely ignore Court judgments completely. But a very large number of cases result in partial compliance, a behavioural pattern concentrated in a number of problematic states but which is also displayed less frequently by many other states. We do not explicitly state or test hypotheses about why variation sometimes occurs in compliance levels. Where we observe variation in compliance levels (e.g., where one state exhibits more compliance than another), we suggest factors that may explain that variation but leave the systematic exploration of those hypotheses for future research.

1. The Inter-American Court

As of 23 June 2010, the IACHR has ordered states to engage in 703 discrete actions--what we label compliance orders--and states have complied with 50% of these. Many of these instances of compliance, however, only occurred after repeated requests by the Court. This is significantly higher than Posner and Yoo's reported compliance rate of around 5%. (74) The difference is partly a result of different units of analysis. Posner and Yoo apparently only reported the cases in which states had complied with every aspect of the Court's rulings. Using that measure, we find more similar results: full compliance has occurred in five of the 81 cases for which there are compliance reports, or 6% of the time. In nine of those 81 cases, the state has not complied with any compliance orders, for an absolute non-compliance rate of 11%. Thus, 83% of the cases should be coded as having partial compliance, though some of these partial compliance cases could of course be on their way to full compliance. By this unit of analysis, partial compliance is the most common outcome, observed in a significant majority of cases. In any given case, states rarely do all they are ordered to do. But by the same token, states rarely do nothing at all. Rather, they engage in partial compliance, i.e. complying with some compliance orders in any given case but not others.

Figure 1 offers one representation of partial compliance, by graphing the distribution of the percent of compliance orders with which a state complies in any given case. For example, 14% of the cases fall in the fifth decile where states have complied with 41-50% of the compliance orders in those cases. The distribution of compliance is fairly normal, with higher tails on both ends representing cases with complete (or nearly complete) non-compliance and complete (or nearly complete) compliance.

Figure 1: IACHR Percent Compliance by Case.

Percentage of total cases     Percentage of
with compliance reports n=81  orders complied with

0-10%                          5%
11-20%                        15%
21-30%                         9%
31-40%                        11%
41-50%                        14%
51-60%                         9%
61-70%                        16%
71-80%                         7%
81-90%                         5%
91-100%                       10%

Note: Table made from bar graph.

Rates of state compliance vary by the type of reparation required by the Court, as illustrated in figure 2a. States comply most with Court orders to pay moral damages (43%), material damages (40%), and Court costs and expenses (43%), and to apologize (40%). Posner and Yoo also found rates of compliance with orders for pecuniary remedies to be above average, but reported them at 24%. (75) Compliance rates are lowest with Court orders to amend, repeal or adopt domestic laws or judgments (7%). States comply 17-19% of the time with Court orders to undertake other sorts of activities such as punishing perpetrators or restoring rights to those who have had them violated. Figure 2b depicts resistance by type of compliance order. Resistance mostly mirrors compliance, where resistance is higher when compliance is lower. Yet there is some variation. For example, states resist paying material damages more than they resist apologizing even though compliance is ultimately the same.

Figure 2a: IACHR Compliance by type of reparation.

Judgment type                                           Compliance rate

Enjoyment of right or freedom violated n=48                         17%
Investigate, Identify, Publicize, & Punish n=107                    19%
Amend, repeal or adopt domestic laws or judgments n=38               7%
Take action or refrain from taking action n=210                     17%
Apologize n=31                                                      40%
Material Damages n=90                                               40%
Moral Damages n=110                                                 43%
Costs & Expenses n=68                                               43%

Note: Table made from bar graph.

Figure 2b: IACHR Resistance by Type of Reparation.

Judgment type                                      Resistance level

Enjoyment of right or freedom violated                         1.71
Investigate, Identify, Publicize, & Punish                     1.54
Amend, repeal or adopt domestic laws or judgments              1.84
Take action or refrain from taking action                      1.61
Apologize                                                      0.94
Material Damages                                               1.10
Moral Damages                                                  0.96
Costs & Expenses                                               1.00

Note: Table made from bar graph.

Overall, these figures provide some evidence that states comply when the costs are relatively low. It is probably easiest for states to pay monetary damages or apologize and walk away. Although the monetary cost for such damages can be higher than some of the other actions required of states, monetary costs probably do not require as many political capital expenses, coordination efforts, or reputational expenses as some of the other types of reparations. These more demanding reparations can involve punishing perpetrators, altering government behaviour in a way that ends the violations of rights, and changing rules and institutions. (76) These types of actions receive lower compliance.

Figure 3a lists compliance by date of the Court's judgment. It is difficult to identify a pattern: significant variation exists from year to year, and compliance rates after 2003 may be low because it typically takes a few years for states to comply. The same is true for figure 3b, which displays resistance levels by date of judgment. There may appear to be relatively low resistance starting in about 1999, but one might expect it to be lower because the Court has had less time to issue compliance reports, and therefore states have had less time to resist them.

Figure 3a: IACHR Compliance by judgment date.

Year reparation
judgment issued     Compliance rate

1989 n=1                100%
1990 n=0                   0
1991 n=0                   0
1992 n=0                   0
1993 n=0                   0
1994 n=0                 67%
1995 n=6                 67%
1996 n=9                 44%
1997 n=18                61%
1998 n=30                 20
1999 n=0                   0
2000 n=50                 73
2001 n=59                61%
2002 n=34                 53
2003 n=71                87%
2004 n=121               48%
2005 n=130               37%
2006 n=121               31%
2007 n=33                39%
2008 n=17                65%

Note: Table made from bar graph.

Figure 3b: IACHR Resistance by judgment date.

Year reparation
judgment issued    Resistance level

1989                  0.00
1990                     0
1991                     0
1992                     0
1993                     0
1994                  0.50
1995                  3.33
1996                  3.11
1997                  3.44
1998                  1.73
1999                     0
2000                  1.35
2001                  2.07
2002                  1.09
2003                  1.00
2004                  1.75
2005                  1.01
2006                  1.05
2007                  0.94
2008                  0.35

Note: Table made from bar graph.

Figure 4, which examines compliance by the year of the compliance report--rather than by the year of the original judgment--corrects for some of these issues, and suggests that beginning in 2003, compliance has improved dramatically. Prior to 2003, states had complied with three of the 98 actions required by the Court as examined in compliance reports, an average of 3% per year. Since 2003, the lowest annual compliance rate has been 20% and the highest has been 39%, with no discernible pattern over time in these years. This change coincides with the Court's 2003 determination that it could monitor compliance, and with its renewed effort to do so. The Court's more focused and persistent efforts appear to have increased state compliance.

Figure 4: IACHR Compliance by date of compliance report.

Year compliance
report issued     Compliance rate

1996 n=1                100%
1997 n=1                  0%
1998 n=4                  25
1999 n=11                 0%
2000 n=6                  0%
2001 n=15                 0%
2002 n=60                 2%
2003 n=132               31%
2004 n=120               39%
2005 n=125               35%
2006 n=142               20%
2007 n=175               31%
2008 n=208               28%
2009 n=121               24%
2010 n=317               24%

Note: Table made from bar graph.

Figure 5 addresses the question of whether persistent Court efforts within a single case can make much of a difference. For any given case, the Court can follow up with repeated compliance reports over time. We number these reports for each case. For example, if the Court follows up with a compliance report one year after a given judgment, that is the first compliance report for that judgment. If it then follows up the next year, it is the second compliance report, and so on. At each stage, instances of compliance are removed from the data set and thus each subsequent report examines only outstanding orders with which states have not yet complied.

Figure 5: IACHR Compliance by Compliance Report.

Compliance report number   Compliance rate

1 n=703                        30%
2 n=346                        24%
3 n=168                        24%
4 n=57                         23%
5 n=24                         25%
6 n=5                           0%
7 n=5                           0%

Note: Table made from bar graph.

The data suggest a possible diminishing return for each compliance report but also suggest that the Court does indeed make some progress each time it follows up. The highest compliance rate is at the first report, with 30%, providing more evidence that compliance is highest when it is easiest because one would expect the easiest tasks to be done first. If the Court does a second compliance report, states comply with an additional 24% of the Court orders they did not comply with in the previous report. In the third, fourth and fifth compliance reports, states comply with 23-25% of the outstanding Court orders. There have been too few instances of orders beyond five to analyze fruitfully, though the minimal evidence (from the aforementioned Loayza-Tomayo case) suggests the Court may eventually run out of influence, as compliance reports mount. (77) In the Loayza-Tomayo case, the Court has issued a sixth and seventh report, without Peru responding at all. Cavallaro and Brewer (78) have argued that the Court is most effective when it is able to connect with interested domestic actors through publicity or through legal procedures. Multiple compliance reports give the Court a mechanism to maintain contact with such actors and to maintain some public awareness of the case. Additionally, the Court has devised mechanisms such as ordering states to publish Court decisions in national newspapers that create public awareness and perhaps increase domestic pressure on the state to comply with other orders in the judgment at subsequent points of time. This suggests that the court's persistence may be rewarded: even the minimal results observed after multiple compliance reports may have long term effects on state behaviour.

Figure 6a examines compliance by state. The top five compliers are Bolivia (87%), Honduras (85%), Brazil (83%), Costa Rica (80%) and Chile (76%). States with very low compliance rates are Trinidad and Tobago (eight%), Paraguay (17%), Colombia (40%), Peru (40%) and Venezuela (44%). It is difficult to identify any patterns here that correspond to differences in domestic politics.

Figure 6a: IACHR Compliance by State.

                              Compliance rate

Argentina n=26                        69%
Bolivia n=15                          87%
Brazil n=6                            83%
Chile n=17                            76%
Colombia n=75                         40%
Costa Rica n=5                        80%
Dominican Republic N=5                60%
Ecuador n=44                          61%
El Salvador n=14                      71%
Guatemala n=141                       63%
Honduras n=40                         85%
Nicaragua n=13                        54%
Panama n=12                           58%
Paraguay n=63                         17%
Peru n=167                            40%
Suriname n=11                         55%
Trinidad and Tobago n=13               8%
Venezuela n=41                        44%

Note: Table made from bar graph.

It is possible that the low number of cases for some of these states throws off the analysis. If we discard any state with fewer than 40 compliance orders, we are left with seven states that can be split into two groups. Venezuela, Peru, Colombia and Paraguay fall in the lower category with compliance rates of 44, 40, 40 and 17% respectively. No obvious similarities exist among these states. Paraguay is a clear outlier and may merit further study. Colombia has faced an ongoing civil war, which may account for its low compliance rate. Peru faces significantly more cases than any other state and may be overwhelmed by the sheer number of compliance orders. Venezuela has struck a separate ideological and political path in recent history, which may affect its desire to comply with the Court. Ecuador, Guatemala and Honduras have higher compliance rates of 61, 63, and 85%, respectively. Guatemela and Honduras are small Central American states still struggling to overcome many years of violence and authoritarian rule. It is possible that compliance rates are higher because most of their cases come from old, discredited regimes and their governments have a strong desire to distinguish themselves from those regimes, and secure their own authority. (79) Figure 6b shows resistance by state. Among the low compliers with a reasonably high number of orders (Venezuela, Peru, Colombia, Paraguay), resistance varies significantly, from 1.05 for Venezuela to 2.0 for Paraguay. While Peru's compliance rate is 40% and Paraguay's is 17%, they have virtually identical resistance rates. This provides some further evidence that Peru is overwhelmed by the number of cases because it resists them as much as Paraguay, but then complies at a higher rate--suggesting some difficulties in processing everything. Among the high compliers with a significant number of cases (Ecuador, Guatemala, Honduras), Honduras's resistance is quite a bit higher, but so is its compliance. This is an interesting pattern that could also suggest a lack of capacity, despite the lower number of orders.

Figure 6b: IACHR Resistance by State.

                         Resistance level

Argentina                        1.31
Bolivia                          0.87
Brazil                           0.83
Chile                            0.53
Colombia                         1.24
Costa Rica                       2.40
Dominican Republic               0.80
Ecuador                          0.86
El Salvador                      1.14
Guatemala                        1.04
Honduras                         1.38
Nicaragua                        1.54
Panama                           1.58
Paraguay                         2.00
Peru                             1.90
Suriname                         0.73
Trinidad and Tobago              1.08
Venezuela                        1.05

Note: Table made from bar graph.

Figure 7a examines compliance by the type of right violated. Most of the compliance rates fall in the range of 40-55%. Some rights with lower compliance rates tend to cluster around political participation issues: assembly, association, and participation in government, though these also have small numbers of orders. Resistance levels reported in figure 7b also do not vary greatly, but again high resistance levels correspond to political participation issues such as free conscience, assembly, association and participation in government--again with the caveat of low numbers of orders in these categories.

Figure 7a: IACHR Compliance by Issue Type

Article Violated from the American               Compliance rate
Convention on Human Rights

3: Right to juridical personality n=36                 47%
4: Right to life n=462                                 45%
5: Right to humane treatment n=578                     48%
6: Freedom from slavery n=28                           57%
7: Right to personal liberty n=427                     53%
8: Right to a fair trial n=650                         50%
9: Freedom from Ex Post Facto laws n=64                55%
11: Right to privacy n=30                              37%
12: Freedom of conscience and religion n=12            42%
13: Freedom of thought and expression n=57             70%
15: Right of assembly n=4                              25%
16: Freedom of association n=32                        34%
17: Rights of the family n=10                          60%
18: Right to a name n=5                                40%
19: Rights of the child n=104                          43%
20: Right to a nationality n=9                         44%
21: Right to property n=93                             44%
22: Freedom of movement and residence n=36             39%
23: Right to participate in government n=16            38%
24: Right to equal protection n=35                     46%
25: Right to judicial protection n=614                 49%

Note: Table made from bar graph.

Figure 7b: IACHR Resistance by Issue Type.

Article Violated from the American               Resistance level
Convention on Human Rights

3: Right to juridical personality                     1.42
4: Right to life                                      1.35
5: Right to humane treatment                          1.31
6: Freedom from slavery                               1.75
7: Right to personal liberty                          1.28
8: Right to a fair trial                              1.37
9: Freedom from Ex Post Facto laws                    1.45
11: Right to privacy                                  1.27
12: Freedom of conscience and religion                2.17
13: Freedom of thought and expression                 1.39
15: Right of assembly                                 3.75
16: Freedom of association                            1.97
17: Rights of the family                              0.90
18: Right to a name                                   0.60
19: Rights of the child                               1.24
20: Right to a nationality                            1.44
21: Right to property                                 1.31
22: Freedom of movement and residence                 0.86
23: Right to participate in government                1.81
24: Right to equal protection                         0.57
25: Right to judicial protection                      1.37

Note: Table made from bar graph.

2. The European Court

This section introduces five pieces of evidence for partial compliance: trends in pending cases, length of pending cases, high profile pending cases, execution of judgment data, and interim resolutions from the Committee. Together these sources paint a picture of a court that regularly continues to achieve full compliance with its judgments, but which also faces a substantial minority of cases in which compliance is partial for quite extended periods, and some other cases in which partial compliance may be the long term outcome.

First, section III noted that cases on which the ECtHR has already rendered judgment can either be closed (which equates to a Committee judgment of full compliance), or pending (which means the Committee has not yet closed the case). (80) Through its Execution of Judgements of the European Court of Human Rights portal, the Committee makes available "snapshots" of currently pending cases, and these can be used to discover cases of partial compliance. (81) All such cases remain pending before the Committee, which examines them at its quarterly Human Rights meetings, until the adoption of a final resolution acknowledging that the measures chosen by the respondent state have achieved the result required by the Convention. Again, this means that the state has remedied, where possible, the consequences of the violation for the applicant (by adopting individual measures and the payment of just satisfaction) and sought to prevent new similar violations from occurring (by adopting general measures).

Figure 8 shows the development of pending cases in recent years, from fewer than 800 in 1996 to well over 8000 by 2009. Figure 9 further distinguishes pending cases from the three most recent years by dividing leading cases from clone cases. (82) Even as the total number of pending cases has grown, leading cases have consistently constituted fewer than 1 in 8 of all pending cases. As a result of the high proportion of clone or isolated cases, states that can manage the general measures to resolve leading cases can simultaneously resolve many others as well (provided that appropriate individual measures are also taken). These patterns tell us only that the scope for full compliance, partial compliance, and non-compliance is large and growing, though it tells us little about the distribution of such compliance.

Figure 8: Development in the Number of Cases Pending,
1996-2009.

1996     709
1997     997
1998    1435
1999    1732
2000    2298
2001    2624
2002    3227
2003    3540
2004    3970
2005    4323
2006    5523
2007    6248
2008    7328
2009    8661

Source: COE, Committee of Ministers, Third Annual Report
2009, supra note 13 at 32.

Note: Table made from bar graph.

The second piece of evidence is the length of pending cases. Long-pending cases are more likely to be instances of partial compliance than are recent judgments, which are either so new as to be complete non-compliance or are rapidly complied with. To get a first clue about this distribution, we turn to figure 10, which summarizes the length of pending cases that are leading cases and shows that just over half have been pending for less than two years (data as of December 2009), while just over a third have been pending for two to five years and a further 11% pending for more than five. This is highly suggestive evidence insofar as while one could expect that general measures--including complex legal and administrative changes--might take a year or two under the best of circumstances, compliance delays approaching and indeed exceeding five years might be good indicators of more stable partial compliance or non-compliance. This is particularly true when five years is more than the usual lifespan of most European governments. A significant minority of pending cases are potential examples of stable partial compliance outcomes.

Figure 9: Pending cases on 31 December, 2007-09.

                          2007  2008  2009

Clone and Isolated Cases  5417  6331  7533
Leading Cases              831   997  1128
                          6248  7328  8661

Source: COE, Committee of Ministers, Third Annual Report 2009, supra
note 13 at 33.

Note: Table made from bar graph.

Figure 10: Length of leading cases pending before the Committee of
Ministers, December 31, 2009.

Leading cases pending for more than 5 years;  11%
Leading cases pending for 2 to 5 years;       35%
Leading cases pending for 2 years or less;    54%

Note: Table made from pie chart.

Figure 11 then shows the geographical distribution of those pending leading cases that had in 2007 been outstanding for at least two years. Turkey, Italy, and Bulgaria have the most cases. Together with Romania, these four states contribute half of the long term pending leading cases before the Committee. When one expands the list to include all cases still pending at the end of 2008 (e.g. including clone cases), the pattern changes a bit, though Italy and, to a lesser extent, Turkey still stand out among a range of other ECtHR members. Countries with more than 30 pending cases were:

Figure 11: Leading cases pending for more than two years.

Other States    16%
Ukraine          4%
Russia           5%
Greece           6%
United Kingdom   6%
Poland           6%
France           7%
Romania          9%
Bulgaria        13%
Italy           13%
Turkey          15%

Note: Table made from pie chart.

Figure 12: Distribution of Cases Pending.

Country         # of pending cases

Italy           2428

Turkey           977

Russia           467

Poland           463

Ukraine          420

Romania          289

Greece           247

Slovenia         206

Bulgaria         174

France           119

Hungary          111

Czech Republic    91

Moldova          104

Slovakia          58

Croatia           54

Portugal          45

Belgium           41

UK                34

Finland           32

Italy            38%
Turkey           15%
Russia            7%
Poland            7%
Ukraine           7%
Romania           5%
Greece            4%
Slovenia          3%
Bulgaria          3%
France            2%
Hungary           2%
Other Countries   7%

Note: Table made from pie chart.

This data has important ramifications, as it shows that many of the countries with the most long-pending cases are precisely those countries that we know from non-ECHR literature to be least responsive to the implementation of, for example, EU directives. (83) Meanwhile, non-EU members such as Turkey, Russia, Ukraine, and Moldova clearly have a number of outstanding human rights issues. Together, these two patterns lend more credence to the idea that many of these long-pending cases are unlikely to turn rapidly into full compliance. (84) It also is clear that countries have not compiled larger numbers of pending cases just because they have been CoE members longest. To the contrary, the relative "lateness" of Italy (1973) and Turkey (1990) to full ECtHR membership means they have required less time to compile a worse record, to say nothing of the many members who joined after 1990.

The third source of evidence is the Committee's assessments in its Annual Reports of trends in high profile leading cases. These reports suggest that apparent delays in full compliance actually reflect many cases of partial compliance. States often have taken many measures to comply, but not enough to see the case closed. For example, in many of these cases, just satisfaction has been paid, but individual and/or general measures are stalled. This is partial compliance. To illustrate, Italian law has not allowed the courts to re-open closed cases, a remedy available in almost all other European states. Of the pending Italian cases, nearly 2,200 are connected to one broad issue--the excessive length of judicial proceedings. (85) Italian authorities long ago began developing legislation to address this issue, a process which itself has gotten so troublesome that the Council of Europe's Parliamentary Assembly began issuing reports on the problems as well. (86) But this problem has been many years in the making, and may yet be many years in unwinding.

The Annual Reports, which began in 2008 (for 2007 data), summarize the leading cases of most concern to the Committee and thus give us a clue as to where the most acute problems lie. Table 1 arrays the data discussed by the Committee in calendar years 2007-09. It contains those cases that the Committee declared required "further general measures." (87) Following section 3, cases closed (or slated for closure) during that year are counted as full compliance, while those remaining open on December 31 of each year count either as non-compliance or partial compliance.

Table 1: Cases Reviewed and Cases Closed by Committee of Ministers,
2007-2009.

                                   2007             2008

Areas of                       Cases    Cases   Cases   Cases
Judgment                      Reviewed  Closed  Closed  Closed
                                2007     2007    2007    2008

Access to efficient              111      46      65      30
justice

Protection of private             37       8      20       9
and family life

Protection of rights              36      12      18       5
in detention

Right to life and                 35       5      16       1
protection against torture

Percentage                            32%             38%
of Cases Closed

                                     2009

Areas of                        Cases   Cases
Judgment                      Reviewed  Closed
                                 2009    2009

Access to efficient               34      12
justice

Protection of private             17       4
and family life

Protection of rights               9       3
in detention

Right to life and                 12       1
protection against torture

Percentage                            28%
of Cases Closed

Source: COE, Committee of Ministers, First Annual Report 2007, supra
note 25; Second Annual Report 2008, supra note 68; Third Annual Report
2009, supra note 13.

Of 18 total issue types categorized by the Committee in each report, the data show four issue types with the largest numbers of total cases overseen by the Committee: access to efficient justice, protection of private and family life, protection of rights in detention, and right to life and protection against torture. This rank order is consistent across all three years of available data. Moreover, all four issue types reveal instances where compliance that was still pending outweighed instances of full compliance. The same pattern holds when one considers issue areas with fewer pending cases. For example, in 2007 data, cases closed outnumbered cases still pending in only three out of 18 total issue types (see figure 13). (88) The lengthy qualitative case descriptions are replete with cases in which just satisfaction and individual measures have been completed but in which some or all general measures remain unconvincing or unreported to the Court.

Figure 13: ECHR Compliance by Issue Type in 2007.

                                        Compliance rate (%)

Access to and efficient functioning of                   42
justice n=111

Protection of private and                                22
family life n=37

Protection of rights in detention n=37                   32

Right to life and protection                             14
against torture and ill-treatment n=35

Property rights n=20                                     20

Issues related to aliens n=17                            41

Freedom of assembly and                                  59
association n=17

Freedom of expression and                                60
information n=10

Co-operation with the ECHR and respect                   20
of right of individual petition n=5

Discrimination n=5                                       40

Cases concerning environmental                           20
protection n=5

Effective remedies n=4                                   25

Freedom of movement n=3                                   0

Freedom of religion n=3                                   0

Electoral rights n=3                                     33

Right to education n=2                                    0

Prohibition of slavery                                    0
and forced labor n=1

Right to marry n=1                                      100

Note: Table made from bar graph.

Our fourth piece of evidence is the comprehensive set of reports available in the Committee's Execution of Judgments Database. (89) This data is critical because, unlike the selected cases in the Annual Reports, it provides information on all currently pending cases (though each quarter, as cases are closed, they are dropped from the database of pending cases). Because this includes over 8,000 cases, many with very long case narratives, it was not practical to code each one. Instead, we chose two countries with between 12 and 30 pending leading cases (Finland (14) and Belgium (18)) and two countries with high numbers (France (31) and Poland (60)). (90)

Of Finland's 14 leading cases (including one with an additional 28 clone cases), there was adequate information available on nine cases, all of which clearly were cases of partial compliance. In most cases, for example, both just satisfaction (if required) and all other individual measures had been satisfied, while the Committee had not yet been satisfied on one or more general measures. In Belgium, there were 18 pending cases in which there was sufficient information available, of which four were associated with a further 81 clone cases. Of the 18 cases, only one was a case of non-compliance and one further now seems a case of full compliance (likely to be judged so at the next opportunity). The other 16 were cases of partial compliance. As noted, some cases of partial compliance might be driven by the relatively recent date of judgment. We indeed see that eight of the individual cases are from 2007-2009, possibly suggesting that many may soon be cases of full compliance. On the other hand, five of the individual cases go back at least five years, and when considering the 81 clone cases, we find another 22 that go back that far.

Turning to states with higher numbers of cases, the pattern remains largely unchanged: partial compliance remains the dominant outcome. Take France, with 31 leading cases in the execution dataset. Again, there are a number of cases decided in 2009 for which there have yet been no Committee hearings and where it is thus impossible to judge the current state of compliance from the information provided. (91) Of the remaining 20 French leading cases in the dataset, however, 17 show clear evidence of partial compliance, while two show likely full compliance and one shows complete non-compliance. Finally, an examination of the 60 leading cases from Poland show 17 cases in which information is insufficient to judge (all from 2009 judgments), while all 43 other leading cases show clear evidence of partial compliance.

In short, while most ECtHR cases are still complied with in the first two years, a very substantial number of older pending cases show clear evidence of partial compliance. When states do not comply quickly, it is still rare that they do nothing. Having taken some steps, they quite often are unable or unwilling to comply fully to the satisfaction of the Committee, often for many years. Indeed in the four countries just considered, we see clear evidence of partial compliance in 85 of the 90 leading cases that contain adequate information.

To see this picture more fully, we turn to the fifth piece of evidence: data on the Committee's interim resolutions (IR). (92) According to the Council of Europe, IRs are a way for the Committee to facilitate discussion on the compliance process:

  The Committee may take various measures to facilitate execution
  of the judgment. It may adopt interim resolutions, which usually
  contain information concerning the interim measures already taken
  and set a provisional calendar for the reforms to be undertaken or
  encourage the respondent state to pursue certain reforms or insist
  that it take the measures needed to comply with the judgment. (93)

The typical inclusion in IRs of "interim measures already taken" clearly implies that IRs often concern instances of partial compliance by states. For example, in one of the many IRs directed at the aforementioned problem of the length of judicial proceedings in Italy, the Committee followed the Council of Europe's Parliamentary Assembly recommendations in recommending that, "although recognizing the measures, legislative and others, taken in the meanwhile, urged the Italian authorities at the highest levels to maintain their political commitment to resolving the problem." (94) Italian authorities responded by setting up a special commission (Mirabelli Commission) to outline further steps and engaged in a series of meetings with the Council Secretariat in which they provided "an exhaustive presentation of the legislative measures already taken and those on the way to adoption by Parliament." (95) The Committee asked for further legal and administrative changes and a timetable for implementation, but the 2009 Annual Report showed that insufficient progress was made, and another IR was then released (2009/42). (96)

IRs have become quite common, especially as a way of efficiently addressing large numbers of clone cases. Cumulatively, the 7815 cases pending as of April 2009 have, to date, generated 7560 IRs. (97) As in the IACHR, some cases have required multiple follow-ups. We note each IR separately since this is an important measure of partial compliance, roughly akin to the "institutional nagging" that we noted in the IACHR data. As cases are closed by the Committee, IRs associated with them later drop from the database, such that the numbers are in constant flux. Nevertheless, this snapshot reveals several interesting pieces of information. First, a substantial number of pending cases have generated IRs (just over 2900 (38%) have generated at least one IR). Second, most IRs are relatively new, although this is a feature of Committee practice to make more use of IRs and does not mean that only recently have cases become hard to resolve. Indeed, 2005 was the first year in which the bulk of the (many very old) Italian cases received IRs, and of the then-7815 pending cases, 3046 went back to 2004 or earlier and 1430 went back to the 1990s.

Third, many cases have generated more than one IR. The Italian cases generated new IRs in 2007 and 2009. Of the 2186 cases that have received three IRs, only 11 are from non-Italian cases. Fifty-six further cases have received two IRs, and here many different countries are represented. There are three cases (two of them Turkish and one from Moldova) where four IRs have been required (so far) and one Turkish case still pending despite five IRs. (98) Fourth, a certain rhythm has developed around the most intractable sets of cases. As noted, the Italian cases are generally being reviewed every other year, while the Turkish cases seem to be on a three-year cycle (the Committee only did IRs on Turkey and Ukraine in 2008). (99) Because there are not multiple compliance orders, it is not possible to generate a resistance measure or ratio as in the IACHR case. Again, delegated compliance is not a checklist. But it is clear that for very many cases, partial compliance is not a brief transitory way station on the road to full compliance.

This section developed five indicators of partial compliance in the ECtHR. The first indicator, pending cases, simply outlined the contours of the compliance challenge and established, uncontroversially, that the ECtHR generates many more judgments each year than the IACHR. The second, length of pending cases, showed that a substantial number of cases remained pending for between two and five years, and a smaller number remained pending even after five years. The third, based on detailed case studies from the Committee's Annual Reports, showed that many cases in both of these pending categories are indeed ones of partial compliance and that the largest problems occur in the same kinds of cases over time. The fourth, data from the execution database, allows us to separate partial compliance in pending cases from cases of full (or soon to be full) and complete non-compliance. Both are unusual among long-pending cases. The fifth, interim resolutions, takes a closer look at the magnitude of the long-term partial compliance pattern and establishes that some partial compliance is a relatively stable outcome. This cumulative picture is in tension with the older claims of the ECtHR itself--echoed by many scholars--that state compliance with the ECtHR judgments was nearly universal (100) and tracks more much closely with a growing alarm in the Court and Committee that compliance is a major and growing issue. (101)

3. Types of Partial Compliance: The IACHR and ECtHR Compared

We have examined general patterns and levels of compliance with court rulings, but what does partial compliance look like in practice? Our final section identifies four types of partial compliance at the state level: split decisions, state substitution, slow motion compliance, and ambiguous compliance amid complexity. We illustrate each with examples from Latin America and Europe. We do not intend these four forms to be exhaustive or mutually exclusive. In any given case, partial compliance could take on characteristics of one or more of these forms or perhaps fall outside all four of these categories. We generate these categories inductively from the patterns we observe.

Split Decision

In a "split decision," the offending state complies with part of the overall checklist in a given case (e.g. monetary compensation) but not with other parts (e.g. legal changes). In delegated compliance, the state most often takes adequate individual measures but, in the Committee's judgment, inadequate general measures.

In Maritza Urrutia v. Guatemala, (102) the IACHR ruled that the state should investigate, publish, and punish those who committed human rights violations against the victim (including torture) and pay compensation for material and moral damage. (103) In 2005, the Court declared that Guatemala had paid the compensation in full and requested information about the state's investigation. Guatemala subsequently submitted information, but in a 2007 report the Court found that the information submitted by Guatemala concerned measures adopted from 1992-99 and that the Court had already reviewed that information. Here, it is not difficult to see how Guatemala's government might more easily pay a fine than investigate a difficult human rights case that might implicate powerful people. Partial compliance results.

Regarding the ECtHR, a similar pattern has already been indicated in many pending Italian cases, where just satisfaction has been paid but Italian law has not allowed the reopening of cases that might satisfy the need for individual measures (e.g. Scordino (1) v. Italy). (104) In H.L. v. the United Kingdom, (105) the Court found the state's procedures for obtaining consent for commitment to a psychiatric hospital were inadequate. In the interim, the patient in question, who had been diagnosed with autism, has long been released, and the operative laws and procedures have been changed in England, Wales, and Scotland. But until they are also changed in Northern Ireland, the Committee has indicated its refusal to close the case.

State Substitution

In "state substitution," the state sidesteps a specific court order from the checklist and offers a different response than the one the Court demanded. (106) In such cases, the state can reasonably argue it is obeying the general spirit of the compliance order, but not the letter of the order. It is reasonable to call such behaviour partial compliance because the state is not completely fulfilling the Court's compliance order, but it also seems unfair to say the state is completely disregarding that order. Partial compliance offers a helpful way to characterize such behaviour.

In Villagran-Morales et al. v. Guatemala, (107) the Court ordered the government to rebury a victim in a place to be chosen by the victim's next of kin. (108) The government then performed a symbolic reburial rather than actually exhuming the remains and reburying them. Authorities claimed that the victim's mother authorized the symbolic reburial, but the Court was not satisfied because the government never presented any documents from the victim's mother to that effect, nor did the government make a case that the victim's mother was unable to provide written documents. In this case, the government can make a reasonable case that it acted in a way that respected the intent behind the compliance order--perhaps even in cooperation with a victim's relative. Still, because the government did not do exactly as the Court requested and did not provide a compelling reason for its failure to comply that might have convinced the Court to alter its order, the case seems properly coded as partial compliance.

ECtHR cases are harder to code this way since it is rare for the Court to make specific orders. Yet states often try to circumscribe their reactions to the Court in ways that arguably offer analogous kinds of "substitutes." Sometimes this is done in good faith and in the expectation that the Committee will accept the state response and close the case. In other cases, however, it seems that states have sought to offer the Committee a response that it is unlikely to accept or to continue offering a response already rejected in the past as inadequate. For example, in Skibinscy v. Poland, (109) the Court ruled that state land use laws had violated individual rights, and in a series of cases, the state responded by paying just compensation and by changing the law governing subsequent planning disputes. The Committee, however, has refused to close the case on the grounds that additional complainants may come forward, and the state has not yet made general provisions for those cases. (110) Gorgulu v. Germany (111) resulted in a judgment for custodial rights for a man whose girlfriend had given their biological child up for adoption without his knowledge or consent. But because the man only enjoys visitation rights rather than custody, the Committee has held that the state is not yet in full compliance and has kept the case open. (112)

And in Baucher v. France, (113) the Court held that an individual's rights were violated when, upon conviction by a French court, he was not informed of the specifics of his conviction in such a way as to make possible an appeal within the operative 10-day window. Individual measures taken included just satisfaction (for non-pecuniary damages) and the Committee was satisfied. However, on general measures, France argued that heavy court workloads sometimes meant that the appropriate documents (upon which foundation an appeal could be lodged) could only be produced after an appeal was lodged. The Committee has rejected this position, arguing that further "measures appear necessary to ensure that defendants may always obtain the reasons for their conviction early enough to be in a position to lodge an appeal." (114)

Slow Motion

In "slow motion" compliance, the state takes steps towards remedial action with respect to a specific action required by the Court or that follows from the Court's decision but does not fulfil that demand completely. This category differs from split decisions because it applies to a specific action rather than to a set of actions within a case. It differs from state substitution because the state suggests (implicitly or explicitly) that it will do more later. Cases with high levels of resistance (meaning, by definition, some absence of compliance across time) and full compliance most clearly fall in this category. But slow-motion compliance can also occur in cases of high resistance and no full compliance if the state seems to be making progress.

In the case of Carpio-Nicolle et al. v. Guatemala, (115) the state argued that it could not pay compensation to the victims all at once because it was in a fiscal crisis due to Hurricane Stan. (116) It proposed paying the victims 33% of their money each year, but the victims rejected the offer. The state proceeded with its plan despite the objections. The state eventually achieved full compliance on this particular matter, according to the Court. One of the victims receiving compensation passed away before the state made the final payment, which thus went to the victim's next of kin. This circumstance illustrates the way in which slow-motion compliance, even when it ultimately turns into full compliance, might still be considered ethically or morally incomplete and partial. In this case, the victim himself never did receive his due recompense because of the state's choice to comply in slow motion. Justice delayed is not necessarily justice denied, but it does seem fair to call it justice partial.

On the ECtHR side, the Committee's 2010 Annual Report stated that "the issue of slowness and negligence in execution has attracted special attention." (117) The Court has been informed by authorities in Northern Ireland that measures to resolve H.L. v. United Kingdom (described above) will likely be in place by 2013, although the judgment became final in 2004. (118) Analogous to the Honduran case just noted, in Stran Greek Refineries v. Greece, the Committee explicitly rejected the Greek government's proposed payment schedule. (119) However, state refusal to pay just compensation is rare. More common is state foot-dragging on general measures. For example, in Zwierzynski v. Poland, (120) Poland sought to defend its 1992 expropriation of private property. At the time the ECtHR held against the state for excessive length of civil proceedings, the case had dragged on in the Polish civil courts without resolution for over eight years. The case, decided by the ECtHR in 2001 but then subject to at least five further appeals by Poland, has still not been closed by the Committee. Poland has, however, taken a number of individual measures in that time, and the case now seems closer to a conclusion than in prior years. (121) Thus, while this case is also a split decision (satisfactory individual measures having long since been taken), the state has been markedly slow about pushing through general measures that would satisfy the Committee. More generally, while any given IR could be evidence of all four kinds of partial compliance, multiple IRs noting very slow progress on one individual matter often are a marker of slow motion compliance.

Ambiguous Compliance Amid Complexity

In "ambiguous compliance amid complexity," states face particularly daunting or demanding tasks. State compliance is problematic because the task is complicated enough that it is difficult to say with certainty whether state behaviour is completely congruent with Court opinions. In some cases, full compliance may be so complex as to require actions beyond the state's capabilities. This category differs from the others because the disparity between state behaviour and court preferences is much clearer in the other categories. In this category, compliance rests much more in the eye of the beholder.

In Garrido and Baigorria v. Argentina, (122) the Court ordered Argentina to pay compensation to the families of the victims, who had been "disappeared" by the government. (123) In 2007, the Court agreed that Argentina had paid compensation to all of the family members it could find. But the Court also insisted that Baigorria had out-of-wedlock children who should also be compensated. Argentina insisted that it contacted all known family members and that Baigorria had no such out-of-wedlock children. It even reported that Baigorria's brother had reported that Baigorria lied about having children as a way to get out of jail earlier. This seems to be a case where it is impossible to tell what the truth may be, yet the Court continues to insist that Argentina pay the alleged children and has classified this as a case of partial compliance as a result.

In Myrna-Mack Chang v. Guatemala, (124) the Court ordered the state to investigate the facts with the "aim" of prosecuting and punishing the perpetrators. (125) The Court decided Guatemala failed to do so, in part because it could not find the perpetrators. The task of punishing perpetrators is far more difficult than paying compensation. It may be beyond the resources of some states to prosecute some human rights violations successfully. Likewise, it is inherently difficult to tell whether a state is faithfully complying with its duty to "aim" for prosecution and punishment. The Court of course cannot order a conviction or a punishment and hence must decide whether the state was sincere in its efforts or not. In such situations, compliance will frequently be ambiguous and is reasonably categorized as partial.

In the ECtHR, it is tempting to argue that the bulk of pending cases with IRs are ones in which ambiguity is a central feature. As noted in several places earlier, states have clearly tried to take some measures to satisfy the Court, but have often failed to do so in a way that satisfied the Committee. At the same time, the ECtHR has found a number of ingenious solutions, including the ability to satisfy claims of people who are dead or who have disappeared (e.g. Muller v. Switzerland). (126) Thus, problems that have challenged Latin American states have been successfully managed in the European context.

Far more challenging are cases like those that have arisen from the Chechen wars in Russia, some of which relate to torture, extrajudicial killings and disappearances of Chechen activists. (127) Russia has set up a special Investigating Committee in its Prosecutor General's Office to investigate these cases and to oversee the implementation of individual measures. (128) But the daunting list of general measures required speaks to a huge gulf between the Court's expectations and even a partially motivated Russian state's willingness to meet those expectations. In this sense, the Court can signal the distaste of much of the Council of Europe for the way Russia has conducted itself in the two Chechen Wars. But it seems increasingly implausible that signalling displeasure--through Court judgments, Committee reports, and IRs--can be, itself, an indications that Russia might be prepared to be more compliant. Put differently, if adequate general measures are taken in these cases (over 100 are associated with Chechnya), they would seem likely to come after the resolution of the Chechen conflict and not prior to it. In that sense, the ECtHR faces a different kind of ambiguity problem than does the IACHR, but one that is no less daunting.

In the face of such complexity in many very difficult cases, the Court and Committee have, in recent years, developed several new tools that have been transforming the way the Court works. (129) Since 2004, the Court has issued so-called pilot judgments, which do in fact lay out some general measures in cases of "wide-scale, systemic human rights violations." (130) Broadly, states have so far complied with these very specific judgments, though they remain quite exceptional. The Court has also delivered judgments ordering specific individual measures in a few cases. Some appear to have worked well (e.g. Assanidze v. Georgia) while others again show the ambiguity references in this category.

For example, in Ilascu v. Moldova and Russia, (131) four political activists in Moldova's breakaway territory of Transdniestria won a judgment against both Moldova and Russia, the latter of which was held to exercise substantial control over the region. Despite the Court's specific judgment and several resolutions from the Committee, demanding two of the applicants' immediate releases, Russia long argued that the measure ordered was neither in its competence nor in its power, given Moldovan sovereignty. Such cases may also fit the category of ambiguous orders, at least if the Russian position is taken seriously. (132)

This section has shown that partial compliance varies, and that four patterns can be found across both very different regimes. Meanwhile, previous sections suggested a pattern in which, over time, the IACHR seems to be achieving more compliance than in the past, while the ECtHR may well be achieving less. In that context, the initial indications that the ECtHR and the Committee may be, in some cases, prepared to move towards more checklist compliance are certainly worthy of further scrutiny.

V. Conclusions

We have dealt above with two quite different regional human rights courts. The IACHR essentially tells state violators, "Complete this list of remedies, and tell us when it's finished. We will then check what you have done." By contrast, the ECtHR essentially tells states, "You've done wrong. Find a way to undo or compensate for the harm you've caused and to avoid future harm. When it's done, tell our designated third party, and they will check." Our central finding is that notwithstanding these very different approaches, states in both systems often find various forms of partial compliance to be a preferred response, which they often attempt to sustain in the face of monitoring and explicit warnings not to do so.

While scholars tend to discuss compliance as a dichotomous, all or nothing outcome, we suspect that partial compliance is likely to be very common and sometimes the most common outcome for many international rules. The presence of similar state behaviour despite the large differences between the European and American regimes suggests that regime type is not driving the most important observed outcomes. Partial compliance is a remarkably stable and common outcome even when examined from a variety of analytical perspectives. In the Americas, levels of partial compliance do not vary much by the year in which a judgment is issued, by the date of the compliance report, by the number of compliance report within a given case, or by the nature of the right violated. Compliance varies a bit more by state in the American system and by the type of right violated in the European system, but compliance levels are still generally in the middling range.

In Europe, conventional wisdom suggests that compliance with Court rulings is very high. In the Americas, there is not much conventional wisdom because of the Court's lower profile, but Posner and Yoo (133) suspect compliance is very low, around 5%. We have found some initial evidence that both assessments are wrong. That evidence is stronger in the Americas, where compliance is easier to monitor because of very specific judgments. The European picture is muddied by the fact that compliance is less objective and harder to judge, where the ECtHR does not order states to undertake any particular steps other than payments to victims, leaving to the states a substantial area of discretion around other aspects of compliance. (134)

In the Americas, compliance with payments is the most common form of compliance, hovering around 40-50%. In Europe, compliance with payments appears to be even higher. For example, in 2009, only 5% of just satisfaction payments were not made on time. (135) But most cases that have been pending for over two years are, in fact, ones in which states have partially complied, and data from interim resolutions from the Committee showed a similar pattern of relatively stable partial compliance.

What drives partial compliance? While that is beyond the scope of this paper, we note that the dependent variable we offer is amenable to further research from within all three major approaches to compliance that we introduced. At the most basic level, since each points to factors that push both for and against compliance, these factors, considered together, might well lead to partial compliance. Some scholars point to international enforcement in the form of diffuse penalties and rewards, both material and social. In the case of human rights (and perhaps in other international issues), those rewards and penalties can point in very different directions. States, multinational corporations and other actors can push others to either increase or decrease their respect for human rights. Strategic actors might easily respond with partial compliance and adjust compliance levels as necessary. Other scholars relate compliance to management capability, which is likely to vary by issue area and by state. Hence, states with middling capabilities may have partial compliance outcomes. Courts sometimes impose difficult demands on states that might desire to comply fully but find it difficult to do so in practice. Finally, still others argue that domestic actors most strongly influence compliance. Domestic actors are likely to vary substantially in their attitudes toward compliance. As a result, strategic political actors may prefer middling compliance levels that can be adjusted as domestic preferences shift over time.

These findings have implications for existing theories and suggest additional theoretical questions that deserve more exploration. First, political science scholars of all theoretical persuasions need to more carefully consider a range of outcomes when considering the effects of international institutions and international law. Scholars have missed the prevalence of partial compliance because they have failed to conceptualize it. The effects of international institutions are often portrayed as either deeply transformational (as in the socialization approach) or barely noticeable (as in realist-style approaches). The middle ground is worth exploring--and some scholars have begun exploring it (136) --partly because it seems quite common and partly because many partial compliance outcomes do not seem to be merely transient waypoints on the road to full compliance.

Second, the mechanisms by which international institutions influence state behaviour might include more subtle means than either incentives (rewards, threats) or persuasion. Here, strategic theories from political science might be fruitful ways to reflect on older international legal process theories. After all, courts facilitate repeated interaction between legal specialists and state policy-makers. Those interactions are potentially costly to both sides, though those costs are frequently intangible, involving goods such as bureaucratic, state or court reputations and precedents for future cases. The interactions also involve careful legal reasoning and repeated discussion. In these interactions, various actors may adjust their positions on issues to preserve their reputations, to advance shared understandings that they prefer, to set particular precedents for future cases, or to engage in reciprocity. These repeated small scale interactions that occur in the give and-take of the legal process may shape state behaviour in important ways. Partial compliance seems like an optimal outcome in such an environment. State officials and judges sometimes win and sometimes lose, but they both want to keep playing the game of bringing cases and implementing decisions. State behaviour changes slowly and gradually, not merely as a result of rewards and punishments on the one hand or deep changes in values and interests on the other, but also due to the legal processes of institutional negotiations. This at least is one paradoxical possibility suggested by our research: that sustained interaction between courts and politicians may both reflect the unsatisfactory nature of partial compliance and, by pitting them against one another in years of iterated exchange, generate plenty more of it.

(1) Louis Henkin, How Nations Behave: Law and Foreign Policy, 2d ed. (New York: Columbia University Press, 1979) at 47.

(2) Abram Chayes & Antonia Chayes, "On Compliance" (1993) 47 Int'l Org. 175.

(3) Jeffrey Checkel, "International Institutions and Socialization in Europe: Introduction and Framework" (2005) 59 Int'l Org. 801.

(4) George Downs, David M. Rocke & Peter Barsoom, "Is the Good News About Compliance Good News About Cooperation?" (1996) 50 Int'l Org. 379.

(5) Emilie Hafner-Burton & Kiyoteru Tsutsui, "Human Rights in a Globalizing World: The Paradox of Empty Promises" (2005) 110 Am. J. Soc. 1373.

(6) Andreas Schedler, "Electoral Authoritarianism" in Todd Landman & Neil Robinson, eds., The Sage Handbook of Comparative Politics (Newbury Park, CA: Sage Publications, 2009) 381; Larry Diamond, "Thinking about Hybrid Regimes" (2002) 13 J. Democracy 21.

(7) Thomas Risse & Kathryn Sikkink, "The Socialization of International Human Rights Norms into Domestic Practices: Introduction" in Thomas Risse, Stephen Ropp & Kathryn Sikkink, eds., The Power of Human Rights: International Norms and Domestic Change (Cambridge: Cambridge University Press, 1999).

(8) David Bearce & Stacy Bondanella, "Intergovernmental Organizations, Socialization, and Member-State Interest Convergence" (2007) 61 Int'l Org. 703.

(9) Following standard practice, we reserve the acronym ECHR for the Treaty that established the Court: European Convention for the Protection of Human Rights and Fundamental Freedoms, Council of Europe, 4 November 1950, E.T.S. 5, 213 U.N.T.S. 222 (entered into force 3 September 1953).

(10) Arend Lijphart, "The Comparable-Cases Strategy in Comparative Research" (1975) 8 Comp. Pol. Stud. 158.

(11) We adopt Stephen Krasner's classic definition of the concept of a regime: "a set of implicit or explicit principles, norms, rules, and decision-making procedures around which actors' expectations converge in a given area of international relations." ("Structural Causes and Regime Consequences: Regimes as Intervening Variables" in Krasner, ed., International Regimes (Ithaca: Cornell University Press, 1983) 1 at 2). Specifically, we refer to the routine practices and procedures of the Courts on compliance issues and the shared understandings of related states and non-state actors about those practices and rules.

(12) This evidence is consistent with recent studies finding that international courts are most effective when they create ways to connect with interested domestic groups: James L. Cavallaro & Stephanie Erin Brewer, "Reevaluating Regional Human Rights Litigation in the Twenty-First Century: The Case of the Inter-American Court" (2008) 102 Am. J. Int'l L. 768.

(13) Council of Europe, Committee of Ministers, Supervision of the Execution of Judgments, Third Annual Report 2009 (Strasbourg: Directorate General of Human Rights and Legal Affairs, 2010) at 35, online: <http://www.coe.int/t/DGHL/MONITORING/EXECUTION/Source/Publications /CM_annreport2009_en.pdf> [COE, Committee of Ministers, Third Annual Report 2009]. The aggregate figures for 2007-2009 were 2067 cases closed and 4307 new final judgments against states.

(14) Mark Janis, "The Efficacy of Strasbourg Law" (2000) 15 Conn. J.Int'l L. at 39-46; Andreas von Staden, "Assessing the Impact of the Judgments of the European Court of Human Rights on Domestic Human Rights Policies" (Paper presented to the annual meeting of the American Political Science Association, Chicago, August 2007), online: <http://www.allacademic.com/meta/p212106_index.html>.

(15) Kal Raustiala, "Compliance and Effectiveness in International Regulatory Cooperation" (2000) 32 Case W.Res.J.Int'l L. 387.

(16) Ibid. at 388.

(17) "Introduction: The Reception of the ECHR in National Legal Orders" in Keller & Stone Sweet, eds., A Europe of Rights: The Impact of the ECHR on National Legal Systems (Oxford: Oxford University Press, 2008) 3.

(18) Ibid. at 24-26.

(19) Beth Simmons, "International Law and State Behavior: Commitment and Compliance in International Monetary Affairs" (2000) 94 Am. Pol. Sci. Rev. 819; David Cortright & George A. Lopez, eds., Smart Sanctions: Targeting Economic Statecraft (New York: Rowman and Littlefield, 2002); Judith Kelley, "International Actors on the Domestic Scene: Membership Conditionality and Socialization by International Institutions" (2004) 58 Int'l Org. 425; Milada Anna Vachudova, Europe Undivided: Democracy, Leverage and Integration After Communism (Oxford: Oxford University Press, 2005).

(20) Simmons, ibid.

(21) Emilie Hafner-Burton, "Trading Human Rights: How Preferential Trade Agreements Influence Government Repression" (2005) 59 Int'l Org. 593.

(22) Jonas Tallberg, "Paths to Compliance: Enforcement, Management, and the European Union" (2002) 56 Int'l Org. 609.

(23) Checkel, supra note 3; Kelley, supra note 19.

(24) Chayes & Chayes, supra note 2.

(25) "The Committee of Ministers has so far always been able to conclude that respondent states have fully executed the judgments rendered against them": Council of Europe, Committee of Ministers, Supervision of the Execution of Judgments, First Annual Report 2007 (Strasbourg: Directorate General of Human Rights and Legal Affairs, 2008) at 9-10, online: <http://www.coe.int/t/DGHL/Monitoring/Execution/Source/Publications/CM_annreport2007_en.pdf> [COE, Committee of Ministers, First Annual Report 2007].

(26) Eric Posner & John Yoo, "Judicial Independence in International Tribunals" (2005) 93:1 Cal. L.Rev. 3; Laurence Helfer & Anne-Marie Slaughter, "Why States Create International Tribunals: A Response to Professors Posner and Yoo" (2005) 93:2 Cal. L. Rev. 3 .

(27) Downs, Rocke & Barsoom, supra note 5.

(28) Jana Von Stein, "Do Treaties Constrain or Screen? Selection Bias and Treaty Compliance" (2005) 99 Am. Pol. Sci. Rev. 611.

(29) Gerda Falkner & Oliver Treib, "Three Worlds of Compliance or Four? The EU-15 Compared to New Member States" (2008) 46 J. Common Market Stud. 293.

(30) Wade Jacoby, "Inspiration, Coalition, and Substitution: External Influences on Postcommunist Transformations" (2006) 58 World Pol. 623; Frank Schimmelfennig, "Strategic Calculation and International Socialization: Membership Incentives, Party Constellations, and Sustained Compliance in Central and Eastern Europe" (2005) 59 Int'l Org. 827; Kelley, supra note 19; Mitchell Orenstein, Stephen Bloom & Nicole Lindstrom, Transnational Actors in Central and Eastern European Transition (Pittsburgh: University of Pittsburgh Press, 2008).

(31) Todd Landman, Protecting Human Rights: A Comparative Study (Washington: Georgetown University Press, 2005); Eric Neumayer, "Do International Human Rights Treaties Improve Respect for Human Rights?" (2005) 49 J. Confl. Resol. 925; Robert Keohane, Andrew Moravcsik & Anne-Marie Slaughter, "Legalized Dispute Resolution: Interstate and Transnational" (2000) 54 Int'l Org. 457.

(32) We deal with some recent exceptions in the discussion of "pilot judgments", below in Section IV.3.

(33) Article 46 ECHR simply provides that "The High Contracting Parties undertake to abide by the final judgment of the Court in any case to which they are parties. The final judgment of the Court shall be transmitted to the Committee of Ministers, which shall supervise its execution."

(34) Rolv Ryssdal, "The Enforcement System Set Up Under the European Convention on Human Rights" in Mielle Bulteman & Martin Kuijer, eds., Compliance with Judgments of International Courts (The Hague: Nijhoff, 1996) 49.

(35) Some have argued that the Court ought to be involved in monitoring compliance with its judgments, as the IACHR is. See S.K. Martens, "Commentary" in Bulterman & Kuijer, eds., ibid. at 71.

(36) Posner & Yoo, supra note 26.

(37) Druscilla L. Scribner & Tracy H. Slagter, "Domestic Institutions and Supranational Human Rights Adjudication: The ECtHR and the IACtHR Compared" (Paper presented to the annual meeting of the American Political Science Association, Toronto, September 2009), online: <http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1449192>.

(38) Posner & Yoo, supra note 26 at 43-44. The authors do not clearly detail their methodology or data sources in these two pages. They refer in the key paragraph on page 43 to "our survey" of the cases, but all citations in that paragraph are to secondary sources or to the Inter-American Commission. In the subsequent paragraph, Posner and Yoo refer to their review of the cases in the Court's annual reports, but provide no other information on coding decisions.

(39) Ibid. at 65.

(40) An excellent overview of the Court is found in Jo Pasqualucci, The Practice and Procedure of the Inter-American Court of Human Rights (Cambridge: Cambridge University Press, 2003).

(41) Velasquez Rodriguez Case (Honduras) (1987), Inter-Am. Ct. H.R. (Ser. C) No. 1, Annual Report of the Inter-American Court of Human Rights: 1987, OEA/Ser.L/V/III.17/Doc.13 (1988).

(42) Cavallaro & Brewer, supra note 12 at 781-783.

(43) OAS, Inter-American Court of Human Rights, Annual Report of the Inter-American Court of Human Rights: 2009 (San Jose: Inter-American Court of Human Rights, 2010) at 77, online: IACHR, <http://www.corteidh.or.cr/docs/informes/eng_2009.pdf>.

(44) Cavallaro & Brewer, supra note 12 (a helpful summary of the most important changes).

(45) Ximenes Lopes Case (Brazil) (2006), Inter-Am. Ct. H.R. (Ser. C) No. 149, at 92-93, Annual Report of the Inter-American Court of Human Rights: 2006 (San Jose: Inter-American Court of Human Rights, 2007), online: IACHR, <http://www.corteidh.or.cr/docs/informes/20063.pdf>.

(46) This impression of the Court's typical behaviour is based on reading all of the Court's judgments. Summaries of the Court's judgments that sometimes include the designated timelines may be found in the Court's annual reports. See, for example, OAS, Inter-American Court of Human Rights, Annual Report of the Inter-American Court of Human Rights: 2008 (San Jose: Inter-American Court of Human Rights, 2009) at 28-31, online: IACHR, <http://www.corteidh.or.cr/docs/informes/eng2008.pdf> [IACHR, Annual Report 2008].

(47) The Court reasoned--apparently through the principle of effectiveness--that although the practice is not explicitly authorized by the Convention, "the effectiveness of the judgments depends on compliance with them." Baena Ricardo et al. Case (Panama) (2003), Inter-Am. Ct. H.R. (Ser. C) No. 104, at para. 129, Annual Report of the Inter-American Court of Human Rights: 2003, OEA/Ser.L/V/III.61/Doc. 1 (2004).

(48) OAS, Office of the Secretary-General, Annual Report of the Secretary General 2005-2006, OR OEA/Ser.D/III.56 (2006) at 122.

(49) IACHR, Annual Report 2008, supra note 46 at 15-17. See also an example of the president of the Court ordering such a hearing, Saramaka People v. Suriname, Monitoring Compliance with Judgments, "Order of the President of the Inter-American Court of Human Rights" (20 April 2010), online: IACHR, <http://www.corteidh.or.cr/supervision.cfm>.

(50) The compliance reports issued after 2001 are available on the Court's website. Before 2001, however, there were several reports issued. We have not been able to locate seven of those reports despite repeated requests to the Court. See Inter-American Court of Human Rights, "Jurisprudence, Monitoring Compliance with Judgments", online: <http://www.corteidh.or.cr/supervision.cfm?&CFID=578183&CFTOKEN=81389875> [IACHR, "Monitoring Compliance"].

(51) Loayza Tamayo Case (Peru) (1997), Inter-Am. Ct. H.R. (Ser. C) No. 33, Annual Report of the Inter-American Court of Human Rights: 1997, OEA/Ser.L/V/III.39/Doc. 5 (1998) [Loayza Tamayo Case].

(52) We exclude presidential reports from our count. The Court's president sometimes issues presidential reports, which appear to be efforts to move states toward compliance, but do not officially report on compliance.

(53) We did not code compliance on procedural judgments, such as orders to submit reports, comply with provisional measures, and comply within a certain deadlines. We also did not code presidential reports, which are reports issued by the Court to call the state, victim's representatives, and Commission together for a private hearing to determine compliance.

(54) Pasqualucci, supra note 40 at 283.

(55) We will see a similar distinction between individual and general measures with the European Court.

(56) Ryssdal, supra note 34 at 50; COE, Committee of Ministers, First Annual Report 2007, supra note 25 at 19.

(57) John Cary Sims, "Compliance Without Remands: The Experience Under the European Convention on Human Rights" (2004) 36 Ariz. St. L. J. 639.

(58) Karen Alter, Establishing the Supremacy of European Law: The Making of an International Rule of Law in Europe (Oxford: Oxford University Press, 2000); Lisa Conant, Justice Contained: Law and Politics in the European Union (Ithaca: Cornell University Press, 2002).

(59) Helen Keller & Alec Stone Sweet, "Assessing the Impact of the ECHR on National Legal Systems," in Keller & Stone Sweet, eds., supra note 17 at 682-89.

(60) Georg Ress, "The Effect of Decisions and Judgments of the European Court of Human Rights in the Domestic Legal Order" (2004-2005) 40 Texas Int'l L.J. 359 at 371-373.

(61) COE, Committee of Ministers, First Annual Report 2007, supra note 25 at 11.

(62) Ibid. at 17. For more detail on the Committee's compliance tools, see ibid. at 17-21; Peter Leuprecht, "The Execution of Judgments and Decisions" in Ronald St. J. Macdonald, Herbert Petzold & Franz Matscher, eds., The European System for the Protection of Human Rights (Dordrecht: Nijhoff, 1993) 791.

(63) There have been 11 reports and recommendations issued since 2000. See Council of Europe, P.A., Committee on Legal Affairs and Human Rights, Implementation of Judgments of the European Court of Human Rights, Documents, AS/JUR (2009) 36 at 2.

(64) COE, Committee of Ministers, Third Annual Report 2009, supra note 13 at 18-19; Ryssdal, supra note 34 at 50.

(65) Sims, supra note 57 at 643-645.

(66) COE, Committee of Ministers, First Annual Report 2007, supra note 25 at 16.

(67) Sims, supra note 57 at 655.

(68) Council of Europe, Committee of Ministers, Supervision of the Execution of Judgments, Second Annual Report 2008 (Strasbourg: Directorate General of Human Rights and Legal Affairs, 2009) at 7 [COE, Committee of Ministers, Second Annual Report 2008], online: <http://www.coe.int/t/DGHL/Monitoring/Execution/Source/Publications/CM_annreport2008_en.pdf>.

(69) Darren Hawkins & Wade Jacoby, "How Agents Matter" in Darren Hawkins et al., eds., Delegation and Agency in International Organizations (Cambridge: Cambridge University Press, 2006) 199 at 217.

(70) COE, Committee of Ministers, First Annual Report 2007, supra note 25 at 16; COE, Committee of Ministers, Second Annual Report 2008, supra note 68 at 10; COE, Committee of Ministers, Third Annual Report 2009, supra note 13 at 10.

(71) Including just satisfaction payments, technically a form of individual measure. Following convention, we exclude just satisfaction when reporting on individual measures, unless specifically noted.

(72) COE, Committee of Ministers, First Annual Report 2007, supra note 25 at 219; COE, Committee of Ministers, Third Annual Report 2009, supra note 13 at 51. Note this is a very conservative estimate of partial compliance, as on time payments average only around a third of total cases. The balance is cases where just satisfaction may have been paid late. We don't count these cases for two reasons. First, they may have been paid too close to the deadline to be picked up in the data as "on time." This is consistent with the idea of states delaying payment as long as possible. Second, these data do not allow us to distinguish late payment from non-payment (e.g non-compliance).

(73) A major difference with the IACHR is that the latter relies heavily on victims to report on state compliance. While the ECtHR does this for just satisfaction payments, victims play almost no role in the much more subjective aspects of individual measures and, as far as we can see, no role at all in monitoring general measures. There is some recent evidence that the Committee is communicating somewhat more with victims on compliance matters, but much more research would be needed to substantiate the claim (COE, Committee of Ministers, First Annual Report 2007, supra note 25 at 16.)

(74) Posner and Yoo, supra note 26 at 43-44.

(75) Ibid. at 44.

(76) Trinidad and Tobago officially withdrew from the IACHR because it could not accept the Court's rulings against the use of the death penalty.

(77) Loayza Tamayo Case, supra note 51.

(78) Cavallaro & Brewer, supra note 12.

(79) This possible explanation could be supported by Andrew Moravcsik's work in the European context, arguing that newly emerging democracies join human rights regimes to secure external support for nascent institutions: "The Origins of Human Rights Regimes: Democratic Delegation in Postwar Europe" (2000) 54 Int'l Org. 217.

(80) In some cases, the state may already have satisfied Court requirements but not adequately reported these actions. There is a long dispute about whether the Court could express disapproval of the Committee closing a case by agreeing with a plaintiff that state actions were, in fact, not consistent with the Court's ruling. See Ryssdal, supra note 34 at 49; Martens, supra note 35.

(81) Council of Europe, "Supervision of execution: Implementation of judgments of the European Court of Human Rights", online: <http://www.coe.int/t/dghl/monitoring/execution/Reports/Current_en.asp> [COE, "Supervision of Execution"].

(82) As noted earlier, the ECHR distinguishes "leading" cases from "clone/repetitive" and "isolated" cases. Data through 2009 shows essentially the same picture: COE, Committee of Ministers, Third Annual Report 2009, supra note 13 at 62.

(83) Falkner & Treib, supra note 29.

(84) That said, the point raised in the previous section, concerning the long term societal effects of pressure from international human rights institutions, would also apply here.

(85) COE, Committee of Ministers, First Annual Report 2007, supra note 25 at 209.

(86) Council of Europe, P.A., Implementation of the Judgments of the European Court of Human Rights, R 1516, Assembly Debate, 24th Sitting, Doc. 11020 (2006) online: <http://assembly.coe.int/Main.asp?link=/Documents/AdoptedText/ta06/ERES1516.htm>. This report urges states to develop new domestic mechanisms to speed compliance and to make responsibility for compliance much more transparent. Traditionally, the Parliamentary Assembly has not been an important factor, but the Committee seems to be outsourcing to the Assembly reporting on some of the most persistent cases. In future research, we plan to use decay models to look into a random sample of the huge pool of pending cases to distinguish among cases of stable noncompliance, stable partial compliance, and trajectories that seem likely to lead to full compliance.

(87) COE, Committee of Ministers, First Annual Report 2007, supra note 25 at 27. In subsequent Annual Reports, the language shifted slightly to indicate that these were cases in which the general and/or individual measures required were "particularly interesting" (e.g. COE, Committee of Ministers, Second Annual Report 2008, supra note 68 at 27; COE, Committee of Ministers, Third Annual Report 2009, supra note 13 at 99).

(88) Because 2007 was the first Annual Report, it contained the largest number of case descriptions because subsequent reports only repeated case descriptions when substantial new developments occurred. Thus, 2007 data represent the broadest single picture of the state of partial compliance, with 2008 and 2009 data reported as supplements.

(89) COE, "Supervision of Execution", supra note 81.

(90) 31 countries have 12 or fewer pending cases in the execution database.

(91) Of 11 French leading cases deficient in information, 9 are from 2009 judgments.

(92) Although they very occasionally also occur in cases of complete non-compliance.

(93) Council of Europe, "About Execution: A Unique and Effective Mechanism", online: <http://www.coe.int/t/dghl/monitoring/execution/Presentation/About_en.asp>.

(94) Committee of Ministers, Second Annual Report 2008, supra note 68 at 129.

(95) Ibid.

(96) Committee of Ministers, Third Annual Report 2009, supra note 13 at 124.

(97) All data in this paragraph are compiled from the Council of Europe's Execution of Judgments database located online: COE, Human Rights and Legal Affairs, <http://www.coe.int/t/e/human_rights/execution/02_documents/PPIndex.asp>.

(98) A case now closed went to 6 IRs.

(99) In a small handful of cases, the Committee has delivered two IRs on the same case in the same year.

(100) Martens, supra note 35.

(101) COE, Committee of Ministers, First Annual Report 2007, supra note 25; COE, Committee of Ministers, Second Annual Report 2008, supra note 68; COE, Committee of Ministers, Third Annual Report 2009, supra note 13.

(102) Maritza Urrutia Case (Guatemala) (2003), Inter-Am. Ct. H.R., (Ser. C) No. 103, Annual Report of the Inter-American Court of Human Rights: 2003, OEA/Ser.L/V/III.61/Doc.1 (2004), online: Human Rights Library-University of Minnesota <http://www1.umn.edu/humanrts/iachr/Annuals/annual-03.pdf>.

(103) See compliance reports for this case from 2005 (Spanish only) and 2007 at IACHR, "Monitoring Compliance", supra note 50.

(104) Scordino v. Italy No. 1 (2006), [GC] No. 36813/97, [2006] V E.C.H.R., 45 E.H.R.R. 7.

(105) H.L. v. The United Kingdom (2004), No. 45508/99, [2004] IX E.C.H.R., 40 E.H.R.R. 761 [H.L v. U.K.].

(106) This category is necessarily different across our two cases. In the IACHR, a state substitutes a response to a specific order from the Court. In the ECtHR, the Committee judges that what a state has undertaken as a general measure is not, in fact, adequate to solve the problem going forward. What they clearly share is an unresolved tension between state and court over the adequacy of a state response.

(107) Villagran Morales et al. Case (the "Street Children" Case) (Guatemala) (1999), Inter-Am. Ct. H.R. (Ser. C) No. 63, Annual Report of the Inter-American Court of Human Rights: 1999, OEA/SerL/V/ III.47/doc. 6 (2000) 665, online: Human Rights Library-University of Minnesota <http://www1.umn.edu/humanrts/iachr/Annuals/appendix-1999.html>.

(108) See IACHR, "Monitoring Compliance", supra note 50 (compliance report of 27 November 2003, online: <http://www.corteidh.or.cr/docs/supervisiones/villagran_27_11_03_ing.pdf>).

(109) Case of Skibinscy v. Poland (14 November 2006), No. 52589/99, E.C.H.R., online: European Court of Human Rights <http://www.echr.coe.int/echr/en/hudoc/>.

(110) COE, "Supervision of Execution", supra note 81, "Poland".

(111) Gorgulu v. Germany (26 February 2004), No. 74969/01, E.C.H.R., online: European Court of Human Rights <http://www.echr.coe.int/echr/en/hudoc/>.

(112) von Staden, supra note 14 at 18.

(113) Baucher v. France (24 July 2007), No. 53640/00, E.C.H.R., online: European Court of Human Rights <http://www.echr.coe.int/echr/en/hudoc/> (French only).

(114) COE, "Supervision of Execution", supra note 81 at "France".

(115) Carpio Nicolle Case (Guatemala) (2004), Inter-Am Ct. H.R. (Ser. C) No. 117, Annual Report of the Inter-American Court of Human Rights: 2004, OEA/Ser.L/V/III.65/Doc.1 (2005), online: Human Rights Library-University of Minnesota <http://www1.umn.edu/humanrts/iachr/Annuals/annual-04.pdf> (Spanish only).

(116) See compliance reports for Carpio Nicolle Case case: IACHR, "Monitoring Compliance", supra note 50.

(117) COE, Committee of Ministers, Third Annual Report 2009, supra note 13 at 25.

(118) H.L v. U.K, supra note 105; COE, "Supervision of Execution", supra note 81 at "United Kingdom".

(119) See Interim Resolution 96/251, in Council of Europe, Collection of Interim Resolutions 1988-2008, H/Exec(2008)1 (13 October 2008) at 42-43, online: <http://www.coe.int/t/dghl/monitoring/execution/Documents/InterimResolutions2008_en.pdf>; Stran Greek Refineries and Stratis Andreadis v. Greece (1994), 301B E.C.H.R. (Ser. A), 19 E.H.R.R. 293.

(120) Zwierzynski v. Poland (2001), No. 34049/96, [2001] VI E.C.H.R. 73, online: European Court of Human Rights <http://www.echr.coe.int/echr/en/hudoc/>.

(121) COE, "Supervision of Execution", supra note 83 at "Poland".

(122) Garrido and Baigorria Case (Argentina) (1996), Inter-Am. Ct. H.R. (Ser. C) No. 26, Annual Report of the Inter-American Court of Human Rights: 1996, OEA/Ser.L/V/III.35/Doc.4 (1997) 75.

(123) See IACHR, "Monitoring Compliance", supra note 50 (compliance report of 27 November 2007, online: <http://www.corteidh.or.cr/docs/supervisiones/garrido_27_11_07_ing.pdf>.)

(124) Myrna Mack Chang Case (Guatemala) (2003), Inter-Am. Ct. H.R., (Ser. C) No. 101, Annual Report of the Inter-American Court of Human Rights: 2003, OEA/Ser.L/V/III.61/Doc.1 (2004).

(125) See IACHR, "Monitoring Compliance", supra note 50 (compliance report of 26 November 2007, online: <http://www.corteidh.or.cr/docs/supervisiones/mack_26_11_07_ing.pdf>.)

(126) Muller v. Switzerland (5 November 2002), No. 41202/98, E.C.H.R., online: <http://cmiskp.echr.coe.int/tkp197/view.asp?item=1&portal=hbkm&action=html&highlight=41202/98&sessionid=59023880&skin=hudoc-en>.

(127) See e.g. Khashiyev v. Russia and its associated clone cases, Khashiyev and Akayeva v. Russia (24 February 2005), Nos. 57942/00 & 57945/00, E.C.H.R., online: European Court of Human Rights <http://www.echr.coe.int/echr/en/hudoc/>.

(128) COE, Committee of Ministers, Third Annual Report 2009, supra note 13 at 103-04.

(129) We thank Mikhail Lobov for his suggestions on this matter.

(130) Human Rights & Social Justice Research Institute, London Metropolitan University, "Pilot Judgments: Warsaw Seminar Programme and Papers", online: <www.londonmet.ac.uk/research-units/hrsj/research-projects/pilot-judgments.cfm> (A good recent symposium on pilot judgments).

(131) Ilascu v. Moldova and Russia (2004), [GC] No. 48787/99, VII E.C.H.R., online: European Court of Human Rights <http://www.echr.coe.int/echr/en/hudoc/>.

(132) In addition, the Committee also may try to mandate very specific steps, as it has in high-profile Turkish and Russian cases but also, e.g., in McKerr v. United Kingdom (2001), No. 28883/95, [2001] III E.C.H.R., online: European Court of Human Rights <http://www.echr.coe.int/echr/en/hudoc/>; Kress v. France (2001), No. 39594/98 [2001] VI E.C.H.R..

(133) Posner & Yoo, supra note 26.

(134) Even if the Committee has been increasingly sharp in its warnings about partial compliance. See COE, Committee of Ministers, Third Annual Report 2009, supra note 13 at 8-9, 12-14.

(135) Ibid. at 51.

(136) Cavallaro & Brewer, supra note 12; Keller & Stone Sweet, eds., supra note 17.

DARREN HAWKINS AND WADE JACOBY *

* Professors of Political Science at Brigham Young University. The authors thank Karen Alter, Rachel Cichowski, Samantha Besson, Laurence Helfer, Mikhail Lobov, Will Moore, Beth Simmons, Jeffrey Staton, and participants in various roundtables and workshops for helpful comments on earlier drafts. Heidi Gasparrini, Moises Costa, Britney Draney, Annie Fuller, and Nicholas Jones provided valuable research assistance.