Justice Lost! The Failure of International Human Rights Law To Matter Where Needed Most

407

Introduction

By almost any measure, nearly half of theworld’s governments today are repressive,

systematically abusing human beings livingwithin their borders.1 Many scholars andpolicymakers have been working to bringthese governments to justice as they attemptto identify which human rights policies holdrepressors accountable for their actions andstop future abuses (Koh, 2002). Among theirtools are international human rights laws,designed in the aftermath of severe atrocitiesduring World War II to prevent futurerepression. Human rights scholars and advo-cates have long been hopeful about the

© 2007 Journal of Peace Research,vol. 44, no. 4, 2007, pp. 407–425Sage Publications (Los Angeles, London, New Delhiand Singapore) http://jpr.sagepub.comDOI 10.1177/0022343307078942

Justice Lost! The Failure of International HumanRights Law To Matter Where Needed Most*

EMILIE M. HAFNER-BURTON

Woodrow Wilson School of Public and International Affairs and Departmentof Politics, Princeton University

KIYOTERU TSUTSUI

University of Michigan, Ann Arbor

International human rights treaties have been ratified by many nation-states, including those ruled byrepressive governments, raising hopes for better practices in many corners of the world. Evidence increas-ingly suggests, however, that human rights laws are most effective in stable or consolidating democraciesor in states with strong civil society activism. If so, treaties may be failing to make a difference in thosestates most in need of reform – the world’s worst abusers – even though they have been the targets of thehuman rights regime from the very beginning. The authors address this question of compliance by focus-ing on the behavior of repressive states in particular. Through a series of cross-national analyses on theimpact of two key human rights treaties, the article demonstrates that (1) governments, including repres-sive ones, frequently make legal commitments to human rights treaties, subscribing to recognized normsof protection and creating opportunities for socialization and capacity-building necessary for lastingreforms; (2) these commitments mostly have no effects on the world’s most terrible repressors even longinto the future; (3) recent findings that treaty effectiveness is conditional on democracy and civil societydo not explain the behavior of the world’s most abusive governments; and (4) realistic institutionalreforms will probably not help to solve this problem.

* Direct correspondence to [email protected] [email protected]. We would like to thank Gary Bass,James Ron, and Kathryn Sikkink for their helpful commentson earlier drafts of this article, as well as participants at theAmerican Political Science Association meeting (September2005, Washington, DC) and the International StudiesAssociation meeting (March 2006, San Diego, CA) wherethis research was presented, and six anonymous reviewers.We would also like to thank Steven Poe and his team at theUniversity of North Texas and Mark Gibney and his team atthe University of North Carolina Asheville for sharing theirdata. Replication data for the analysis and an appendixare available at http://www.princeton.edu/~ehafner/andhttp://www.prio.no/jpr/datasets.

1 In 2003, 81 of 168 states in the Amnesty Internationalreport, 88 of 176 states in the US State Department report,and 81 of 188 states in the Freedom House report wereobserved to commit significant acts of repression.

prospects for reform. While a few quantita-tive scholars have culled new evidence sug-gesting that laws often do not work very well(Hathaway, 2002; Hafner-Burton, 2005),others show hope for reforms on the margins,in democratizing (Simmons, 2006) oralready democratic states (Neumayer, 2005).Our concern in this article is with the effi-cacy of these laws to reform those actors mostin need of improvement: repressive stateswhose governments violate or allow violationof human rights within their borders. Arehuman rights laws really hopeless to bringabout reforms, even marginal, among theworst abusers?

Despite recent skepticism, scholars ofinternational relations, law, and sociologyhave long argued that laws can make a differ-ence, and hope for improvement is common(Landman, 2005; see Hafner-Burton & Ron,2006). Many politicians and nongovern-mental activists also believe that human rightslaws initiate processes and dialogues thatinvolve learning over time and, throughlearning, the eventual change in belief aboutrational or appropriate actions (Abbott &Snidal, 2000). They provide rules andorganizational structures that constrainnational sovereignty, serving as justificationand a forum for action that can shape govern-ments’ political interests and belief aboutappropriate actions (Chayes & Chayes, 1998;Franck, 1988; Lutz & Sikkink, 2000). Andpersuasive accounts argue that governmentsratify human rights treaties, not always assymbolic acts, but also as expressions of pref-erence for reform (Simmons, 2006). Byalmost all such accounts, if human rights lawsmatter for political reform, they will take timeto be of importance, as belief change andcapacity-building for implementation areunlikely to be easy or immediate and maywell happen in fits and starts (Chayes &Chayes, 1993).

Theories of compliance, however, are tosome extent divorced from research. Current

findings largely emphasize that treaties workin some cases – democracies. But these studieslargely ignore the dynamics of compliance.This is troubling because the human rightsregime was created precisely to stop outbreaksof extreme violence among the world’s worstabusers, and its founders knew this processwould take time. Perhaps researchers arefinding that treaties matter most on themargins because studies are not taking thedynamics of compliance seriously. Mayberepressive autocrats simply need more time tocome under the sway of international lawsand build capacity than other, more democ-ratic, states.

Consider first what we know about effec-tiveness. In the face of widespread confidencethat laws matter, Hathaway’s (2002) path-breaking article shook scholarly faith inhuman rights treaties, arguing that they dolittle to ensure better behaviors. Since thisprovocative study, other scholars have beennotably more optimistic. Simmons (2006)argues that international legal commitmentsdo matter; they have their most importantconsequences for states that have experienceddemocratic accountability and refuse toallow their governments to turn back.Hafner-Burton & Tsutsui (2005) demon-strate that linkage to international civilsociety often encourages reform in caseswhere international law alone is unsuccess-ful. Neumayer (2005) extends both argu-ments to show that commitment tointernational law often does improve respectfor human rights, primarily for states withdemocratically accountable governments orstrong civil society. The optimism, however,is narrow in scope, as current scholarshipimplies that human rights laws matter leastamong governments that were the primarytargets of the legal regime – terribly repres-sive, autocratic states without internal advo-cates for reform.

Consider next what we know about thedynamics of treaty compliance. Conformity

journal o f PE AC E RE S E A RC H volume 44 / number 4 / july 2007408

with international law is a domestic politicalprocess. Implementing human rights lawsrequires not only the political will at home, butalso the political capacity. Both probably will behardest to build in repressive non-democracies,and conformity with international humanrights laws will almost certainly take longer tostick in these cases. The burgeoning empiricalliterature on human rights compliance has yetto effectively consider whether treaty effective-ness fluctuates over time. For instance,Neumayer (2005) and Keith (1999) considerwhether several global and regional humanrights treaties make a difference in humanrights behaviors the very same year as ratifica-tion. Not surprisingly, they find no directempirical relationship. Hafner-Burton (2005)examines whether any of the core UN humanrights laws encourage protection of peoplefrom political terror one year after ratificationand also finds no significant association. Allunreservedly overlook basic theoretical argu-ments suggesting that soft laws generally taketime to be successfully implemented, and thathuman rights laws in particular are likely to beeffective only after substantial learning andcapacity-building have taken place – features ofinternational human rights law that ‘may beseen as an extreme case of the time lag betweenundertaking and performance’ (Chayes &Chayes, 1998: 16).

Other empirical research acknowledgesthat ‘human rights treaties, if they have effectson country practices, do so relatively slowly’(Hathaway, 2002: 1990). To consider thesedynamics, both Hathaway (2002) and Hafner-Burton & Tsutsui (2005) analyze the relation-ship between the duration in years sinceratification of the core UN human rights lawsand compliance behavior. In so doing, theytest the proposition that, as the years go on,human rights laws should be more and moreeffective in producing results. They findno evidence. Yet neither study is a good testof dynamic theories of international law.Treaties may certainly take time to influence

behaviors, but, in the realm of human rights,it is unlikely that learning or capacity-buildingtakes place at a steady or uniform pace overtime. Compliance with international humanrights laws, if it takes place at all, may wellhappen sporadically and in fits and starts. Ifso, these duration variables are a weak test ofimportant theories on the matter.

Does this methodological problem explainthe discouraging results about compliance?Maybe human rights laws do protect the peoplemost in danger of violations, but only in fitsand starts and only long after ratification, whenleaders’ minds can be swayed and nationalcapacities for reform built. Perhaps democracyis not the only answer.

In the following pages, we advance fourpropositions about repressive governments’compliance with international human rightslaw. First, advocates are correct: an impressivecascade of norms has taken place in the realmof international justice. Governments, includ-ing repressive ones, easily and frequently makelegal commitments to international humanrights treaties, subscribing to recognized normsof protection and creating opportunities forsocialization, learning, and capacity-buildingprocesses necessary for lasting reforms. Second,the problem is not only methodological; treatycommitments to the pursuit of justice have noclear or independent effects on most veryrepressive states’ behaviors, either immediatelyor, more importantly, long into the future.Either most repressive governments have failedto learn that the protection of human rights isessential or they lack the capacity necessary toimplement policies of protection. As a result,recent statistical confidence about the treatyregime implies a broader problem – that theregime is actually failing in countries wherereform is most urgently needed and that moretime for learning and capacity-building isunlikely alone to solve the problem. Third,recent findings that treaty effectiveness is condi-tional on democracy and civil society do notexplain the behavior of the world’s serious

Emil i e M. Hafner-Bur ton & Kiyoteru Tsut sui HU M A N RI G H TS LAW 409


repressors. Fourth, most realistic institutionalreforms are unlikely to help much; so far,deeper delegation of legal authority to the inter-national regime does not make reforms muchmore likely, even over time.

The Good News About TreatyCommitment

Significant thought has been given to explain-ing why governments commonly belong to theinternational human rights legal regime (Cole,2005; Goodliffe & Hawkins, 2005; Vreeland,2007; Wotipka & Tsutsui, 2001). We build onthese insights with a focused attention to thebehaviors of those states arguably most in needof reform, where repression is severe and indi-viduals experience considerable brutality.

Human rights laws are abundant. Amongthem, the International Covenant on Civiland Political Rights (CCPR) and theInternational Convention Against Tortureand Other Cruel, Inhuman, or DegradingTreatment or Punishment (CAT) are com-monly distinguished among treaties as the

most successful and important, outlawing themost severe kinds of violations (Hathaway,2002). They are certainly among the moststudied. Each treaty articulates an impressivearray of obligations for governments to follow,including precise details as to what behaviorsare and are not appropriate. Authority tomonitor implementation is modestly dele-gated to two independent committees: theHuman Rights Committee for the CCPR andthe Committee against Torture for the CAT;governments can choose to recognize or rejectthe jurisdiction of both committees and toimplement or ignore their recommendations.2

In the following section, we aim to demon-strate a remarkable fact. No matter how wemeasure repression of personal integrity rights,repressive states that allow murder, torture,kidnapping, and other cruel, inhuman, ordegrading treatment or punishment of peoplejust as commonly belong to the CCPR and theCAT outlawing these behaviors as govern-2 Complete details of the treaties’ provisions are availablefrom http://www.unhchr.ch/html/menu3/b/a_ccpr.htmand http://www.unhchr.ch/html/menu3/b/h_cat39.htm(accessed 10 December 2005).

Figure 1. The Predicted Probability of Treaty Commitment, 1976–2003: Repressors Versus Protectors

0

0.2

0.4

0.6

0.8

1

Year20

0219

7619

7819

8019

8219

8419

8619

8819

9019

9219

9419

9619

9820

00

protector repressor

ments that protect human rights reasonablywell. Figure 1 presents evidence that has beensystematically culled from our data from 1976,the year the CCPR came into force, to the lastyear of observation, 2003.

We begin by defining what we mean by arepressive state, repressori, and how we identifythem. Because we are concerned with two par-ticular treaties, the CCPR and the CAT, weconsider repression of the human rightsenshrined in both: the rights to personalintegrity. We examine annual data published byAmnesty International (AI), although we haverun all analyses in this article on an alternativemeasure collected by the Bureau of Democracy,Human Rights, and Labor at the US StateDepartment (SD) as a robustness check; anydiscrepancies are reported in the footnotes. Wecombine data collected across three samples byPoe & Tate (1994), Gibney & Dalton (1996),and Hafner-Burton (2005) to create a singlemeasure with observations on 182 states from1976 to 2003.3 Following Gastil (1980), wedefine repression, repressionit, as an ordinal vari-able ranging across five levels of behavior:

(1) countries are under secure rule of law,political imprisonment and torture are rare,and political murders are extremely rare;

(2) imprisonment for nonviolent politicalactivities is limited, torture and beating areexceptional, and political murder is rare;

(3) political imprisonment is extensive, executionand political murders may be common, anddetention for political views is acceptable;

(4) the practices of level 3 are expanded to alarger segment of the population, murdersand disappearances are common, butterror affects primarily those who interestthemselves in political practice or ideas;

(5) levels of terror are population-wide anddecisionmakers do not limit the meansby which they pursue private or ideo-logical goals.

In order to identify repressor states, weconsider any state that achieves a score of 3or higher at the time of treaty ratification to beunquestionably repressive, employing politi-cal torture and terror. We call this state arepressor, repressori, and code them dichoto-mously for efficiency.4 We consider statesthat achieve a score of 1 or 2 to be reasonablyprotective by contrast, because acts of repres-sion, if they are observed, are extremely rare.5

Using this common definition for repressorito identify our population of violating states,we consider the extent to which these statesbelong to the human rights legal regime, rati-fying the CCPR or the CAT into national law.We accordingly estimate Model 1:

treaty commitmentit = a + !1 repressori+ !2polityit + !3 regime durabilityit + !4 GDPpcit + !5tradeit + !6 civil warit + !7 warit + !8 populationit+ !9 North Americait + !10 Europeit + !11 Africait+ !12 Middle Eastit + !13 Asiait + "i + µit (1)

Our dependent variable, treaty commit-mentit, is a binary variable coded 0 if a state iin year t has made no formal commitment toeither the CCPR or the CAT by ratifying,acceding, or succeeding to the treaties, and 1if that state has committed to either or bothtreaties. Our aim in Model 1 is to determinewhether states that are observably repressiveat the time of ratification are any more or lesslikely to belong to human rights laws thanthose that are reasonably protective, havingalready institutionalized the norms to whichthey make commitments. To pursue this aimsystematically, we control for a variety ofother motivations considered relevant in statedecisions on treaty membership (Cole, 2005;


3 For details of the data, see Appendix (http://www.prio.no/jpr/datasets).

4 We have run sensitivity analyses to ensure that ourdichotomous coding is appropriate. In these analyses, wecoded states that received a score of 3 in a given year as amoderate repressori, 4 a severe repressori, and 5 an extremerepressori. Analyses using moderate, severe, or extreme repres-sori in the place of repressori, produce consistent results.5 A complete list of repressor states is available in the repli-cation file.


Goodliffe & Hawkins, 2005; Hathaway,2002; Moravcsik, 1995).6

Polityit measures domestic regime charac-teristics. The well-known variable, whichtakes on values ranging from 10 (most demo-cratic) to –10 (most autocratic), is measuredas an index of five primary institutional fea-tures: the competitiveness of chief executiveselection, the openness of that process to socialgroups, the level of institutional constraintsplaced on the chief executive’s decisionmakingauthority, the competitiveness of political par-ticipation, and the degree to which bindingrules govern political participation.7 Regimedurabilityit counts the number of years since astate has undergone a structural regime tran-sition. This variable is commonplace in theliterature, and we control for it accordingly. Atransition is defined as a movement on thePolity IV scale of three points or more.

To control for economic factors that schol-ars believe may influence treaty commitment,we employ several standard variables. Grossdomestic product per capita in constant USdollars, pcGDPit, controls for the effect of eco-nomic development. Tradeit controls for thesum of a state’s total exports and imports ofgoods and services measured as a share of grossdomestic product. Both measures are collectedby the World Bank and logged to reduce theskew of their distributions.

To control for political conditions whereviolence is openly sanctioned by the state, wecontrol for civil warit and warit.. Both aredichotomous variables equaling 1 if a countryis at war and 0 otherwise.8 Finally, we controlfor regional effects by including dummy vari-ables indicating whether a state is a part ofAmerica, Europe, Africa, the Middle East,or Asia, and for a state’s total populationit,logged. "i are fixed effects for time that

de-trend correlations within states acrosstime (although not within time across states,which is equal to zero by assumption), andµit is a stochastic term.

Figure 1 plots predicted probabilities thatrepressors (repressori = 1) and protectors (repres-sori = 0) will make commitments to eithertreaty (Long, 1997).9 The y-axis reports prob-abilities calculated at the mean of all variablesin the model. In order to explore ratificationpropensity over time, the x-axis reports the yearin which the prediction was calculated – forexample, in 1976, a repressive state had a 0.2probability of joining the regime, and soforth.10 The figure indicates two systematicfeatures of the commitment process. First,there is nothing about using torture or other-wise repressing rights to personal integrity thatprevents a government from making commit-ments to international laws to abolish repres-sion. Repressors are just as likely to commit tothe human rights legal regime as protectors;this information includes consideration of allpolitical behaviors controlled for in Equation(1). Second, ratification has become nearlyubiquitous over time; the probability thateither a protector or repressor had made acommitment to the human rights regime in1976 – the year the CCPR went into force –was about 0.02; by 2003, this probability wasalmost 1.11

Evidence clearly shows that governmentsthat torture and terrorize their people com-monly pledge commitment to human rightslaws, obligating themselves and their futureleaders to implement norms protecting all

6 For detailed justifications of each control variable, see thesources cited. 7 See online Appendix for details on the data.8 For a detailed explanation of the data, see onlineAppendix.

9 All other results, including the coefficients and standarderrors from which these probabilities are calculated, areavailable in the replication file.10 Confidence intervals are not reported around the pre-dictions for visual clarity. However, they are available in thedata replication file and indicate that there is no statisticaldifference between repressors and protectors in their likeli-hood of treaty ratification, regardless of the year in whichwe compute the prediction.11 Although there are some slight variations in samples andreports between AI and SD, results are generally consistentacross both human rights-reporting sources.

human beings.12 For pundits expecting thathuman rights regimes make a difference, thisobservation should be good news. Suchcommitments may initiate learning, social-ization, and persuasive compliance dynam-ics that, over time, can encourage reform.But do these acts of commitment actuallyhelp to protect the people most in need ofprotection? We consider this question in thefollowing section.

The Bad News About Compliance

Many scholars are hopeful that, when statesmake commitments to international legalnorms, governments will act accordingly(Chayes & Chayes, 1998; Finnemore, 1996;Goodman & Jinks, 2003; Henkin, 1979;Koh, 1996–97; Mitchell, 1993). Some believethat repressive leaders commonly adopt humanrights laws instrumentally as a means to gainrelated benefits, but that instrumental adapta-tion can, over time, lead to processes of moralconsciousness-raising, argumentation, per-suasion, institutionalization, and even habitu-ation (Risse, Ropp & Sikkink, 1999). Theseprocesses can, in turn, create the conditionsfor behavioral change, as governments stopcommitting or supporting acts of repressionand build capacity both to protect humanrights and to punish those who violate them.

Others argue that governments join humanrights laws when they are committed to theirfundamental goals in the first place, evenwhen belonging is costly (Hathaway, 2003,2005; Simmons, 2006). Indeed, states are reg-ularly observed to comply with internationallaw without any enforcement, a likely artifact

of state selection into legal regimes thatrequire only modest changes in behavior(Downs, Rocke & Barsoom, 1996).13 Stillothers suggest that international human rightslaws reform repressive states through accultur-ation – a general process by which actorsadopt the beliefs and behavioral patterns ofsurrounding cultures through mimicry orassimilation (Goodman & Jinks, 2003; Powell& DiMaggio, 1991; Scott, Meyer & Boli,1994). All of these theories are plausible;however, they suggest that any observablecompliance behavior will likely take time, asbeliefs do not change overnight, and buildingphysical and legal infrastructures to supporthuman rights can be costly and time consum-ing. Maybe human rights treaties do matterfor the world’s repressive autocrats, but theysimply require more time to have an effect.

We are skeptical that international humanrights laws in general, and the CCPR and theCAT in particular, directly or regularly encouragemost repressive states to substantially reform, tovalue or protect human beings’ fundamentalrights to life, liberty, and justice, even over time.To a handful of skeptics who see internationallaws generally as cheap talk, this is not news; tomany scholars of international law, internationalrelations, and sociology, as well as to the foundersof the human rights regime and the nongovern-mental organizations who support their cause,this claim is worrisome. Our concerns are mani-fold, but we emphasize four in particular(Hafner-Burton, 2005).

First, we share the view that governments arestrategic actors that make commitments tohuman rights treaties for deliberate and self-interested reasons. Understanding compliancedynamics accordingly obliges us to know some-thing about the process of commitment –whether and why repressive states join treaties inthe first place which almost certainly helps to


13 This finding is nicely articulated in the case of humanrights by Moravcsik (1995), who shows that Europeanhuman rights regimes are likely to have little effect on thosestates that are not already disposed toward transformation,namely newly developing states.

12 The data replication file provides further informationabout spatial and institutional variation in ratificationbehaviors. We consider six major world regions and dis-cover (using predictions) that repressive states that makelegal commitments to the human rights treaty regime aredistributed around the world fairly equally, coming fromevery major region except Asia. Surprisingly, repressorsruled by democratic and autocratic governments areequally likely to have ratified one treaty, although democ-racies are slightly more likely to have ratified both.


explain whether treaties themselves make adifference. Violators, we believe, by and large donot join human rights laws in such great numbersbecause they are committed to reforms or becausetreaties require only modest changes in behavior.On the contrary, the CCPR and the CAT requiremajor reforms and oblige governments with nointerest or capacity to comply with these laws toregularly commit anyway, knowing that neithertreaty can successfully enforce the norms they aredesigned to protect. It is our belief that repressivestates commonly belong to the human rightsregime because they gain certain political advan-tages from membership but all the while can getaway with murder. Most governments joining theinternational human rights legal regime accord-ingly are not open to new ways of behaving, thusmaking learning, socialization, and persuasionimprobable.

Second, we are skeptical that repressive states,once they join the treaty regime, will come tointernalize the legal norms to which they sub-scribe over time, through active processes ofsocialization or learning. We do not believe thatsocialization and learning are impossible orundesirable. Rather, we believe there are strongreasons to be skeptical that either method ofbelief change is likely to take place or providestrong incentives to change most repressiveactors’ behaviors most of the time, no matterhow much time passes after ratification.Socialization and learning require changingactors’ preferences for repression, and these pref-erences are likely to be highly valued by repres-sive states, whose leaders accumulate power andwealth through terror. Socialization and learningare also likely to be slow-acting forms of influ-ence, taking place over a very long time horizon,but confronting resistance to beliefs that areoften sticky and resistant to change (Anderson,1989; Slusher & Anderson, 1996). In those rareinstances when socialization and learning do takeplace, newly persuaded leaders may not be con-sistent across time, as new rulers may come topower and new opponent groups may form with

preferences for repression. Moreover, socializa-tion and learning require repeated access to targetrepressors and many of these actors are likely tobe marginalized from participation in humanrights institutions, remaining isolated fromactive processes of norm inculcation (Hafner-Burton & Tsutsui, 2005).14

Third, we are most skeptical that repressivestates, facing degrees of cognitive and socialpressures to conform, adopt human rightsreforms through processes of acculturationalone.15 We do not argue that acculturation pro-cesses, like socialization and learning, are irrele-vant to all political choices. They may certainlyinform states’ identities, preferences, and inter-ests in important ways (Meyer et al., 1997).However, we emphasize that repressive states arecharacterized by leaders that employ or condonerepression purposively and strategically; acts ofterror are accordingly seldom accidental orrandom, and they often bring high rewards forthose that use them. Reforms are usually delib-erate and costly, often requiring leaders in powerto give up certain authorities and privileges theyhave become accustomed to enjoying. Improve-ments in protection of human rights do not, asa result, happen tacitly or through simple pro-cesses of mimicry without some convincingmotivation. Repressive leaders can certainlyreform, but they are unlikely by any stretch ofthe imagination to give up repression simplybecause their neighbors have.

Finally, even when leaders decide to reformtheir human rights practices, they may notsucceed in changing the government’s or non-state actors’ actual practices quickly. Building the

14 Owing to the lack of substantive penalties in the CCPRand the CAT, repressive governments can choose to ignoredeadlines for their country reports or submit falsifiedreports, hence avoiding good-faith participation in thetreaty mechanisms, which could trigger some form ofsocialization.15 We distinguish between socialization and acculturationin terms of the degree of agency at work. Socialization takesplace with purposive actors making decisions to changepolicies, while acculturation is guided more by imitation inthe face of uncertainty.

legal and physical infrastructure to protecthuman rights and to punish violations is a costlyprocess that requires not only conviction onbehalf of a government’s leaders, but also sub-stantial resources and expertise that are oftenlacking in violating states suffering from legaciesof repression. Leaders seeking reform commonlyface serious resistance from other elites who havevested interests in continuing repression, or fromlower-level officials who have grown accustomedto the organizational culture of repression, whichcan delay or paralyze leaders’ efforts to improvethe practices (Ron, 2000).

Do terribly repressive states reform whenthey belong to human rights treaties over along period of time? How much time does ittake to internalize the human rights norms towhich they make commitments? We areencouraged that many common and import-ant responses suggest that legal commitmentsmatter; that repressive states can and doreform; and that reform, if it does not happenright away, can take place over time, as normsbecome more and more internalized and ascivil society actors use international lawsas a recourse for lobbying. Our arguments,however, have led us to mainly expect thecontrary. We advance three hypotheses: (1)that repressive states’ legal commitments tothe human rights regime do not typicallypromote reforms; (2) that this gap betweencommitment and practice will often persistover time, as norms of justice rarely becomeinstitutionalized through processes of inter-national law alone; and (3) that most feasiblereforms to the legal regime will probably notsolve this problem. The following sectionexplores the merits of our conjectures.

Evidence

We begin with a general replication of previ-ous studies by estimating Model 2 predictingrepressionit behavior (from 1 to 5 on ourordinal scale) to evaluate the impact of treaty

ratification after one year. We use Hafner-Burton (2005) as our base model.16

repressionit = a + !1 CATit–1 + !2 CCPRit–1+ !3 GDPpcit–1 + !4 tradeit–1 + !5 popula-tionit–1 + !6 polityit–1 + !7 regime durabilityit–1+ !8 civil warit–1 + !9 warit–1 + "i + µit (2)

We first consider all states in our sample17 inorder to replicate existing studies and controlfor standard economic and political factorsthought to influence repression. Column 1 ofTable I summarizes our ordered logit estimatesappropriate to the structure of the dependentvariable and reports Huber–White standarderrors appropriate to the nature of our data. Weinclude "i in order to de-trend correlationswithin states across time, as the data by naturesuffer from autocorrelation.18 Our findingsconfirm that state commitment to either treatydoes not increase the likelihood of reform.Governments more often than not commit toprotect norms of human rights but do notfollow through on those commitments.

Columns 2 and 3 of Table I summarizeour replication of previous findings thattreaty effects are conditional on democracyand civil society; we accordingly introduceinteraction terms between treaties and poli-tyit–1 and civil societyit–1, which we measure as


16 The model and control variables are very similar to thoseemployed by other major studies of repression, includingPoe & Tate (1994), Keith (1999), Hathaway (2002),Hafner-Burton & Tsutsui (2005), and Neumayer (2005).See these studies for detailed discussions and justificationsof these variables. 17 Please see online Appendix for a discussion of the impactof missing data on our analyses.18 An alternative procedure for addressing autocorrelationis to include a lagged dependent variable. However, thereis substantial debate as to whether this practice is appro-priate or necessary, as the method risks bias in the estima-tions. Fixed time effects are a suitable alternative, and weinclude them here on this basis. However, we estimate amodel with a lagged dependent variable as a robust checkand find no substantive difference in the estimates; the lagis positive and significant, but all other variables remainconsistent with the results reported in Table I. Results arereported in the replication file. We therefore do not includethe lagged variable in the remainder of our analyses.


Tabl

e I.

Ord

ered

Log

it Es

timat

es o

f Sta

te R

epre

ssio

n, 1

976–

2003

All s

tate

sSt

ates

repr

essiv

e at t

he ti

me o

f rat

ifica

tion

Vari

ables

(1) B

ase

mod

el(2

) Dem

ocra

cy(3

) Civ

il so

ciet

y(1

) Bas

e m

odel

(2) D

emoc

racy

(3) C

ivil

soci

ety

CAT

it–1

–0.0

080.

064

0.10

0–0

.019

0.00

70.

508

(0.2

11)

(0.2

07)

(0.2

84)

(0.2

38)

(0.2

32)*

(0.4

49)

CC

PRit–

10.

019

–0.0

480.

909*

*0.

181

0.02

30.

565

(0.2

27)

(0.2

21)

(0.3

02)

(0.2

50)

(0.2

68)

(0.3

96)

CAT

it–1

x po

lity i

t–1

–0.0

02–0

.005

(0.0

23)

(0.0

31)

CC

PRit–

1 x

polit

y it–

1–0

.073

*–0

.046

(0.0

28)

(0.0

29)

CAT

it–1

x ci

vil s

ocie

tyit–

12.

00E-

04–7

.37E

-04

(3.0

0E-0

4)(6

.47E

-04)

CC

PRit–

1 x

civi

l soc

iety

it–1

–0.0

02**

*–1

.16E

-03+

(4.5

5E-0

4)(6

.11E

-04)

Civ

il so

ciet

y it–

18.

83E-

040.

002*

(6.4

6E-0

4)(8

.82E

-04)

GD

Ppc i

t–1

–0.2

17*

–0.2

02*

–0.1

150.

119

0.12

20.

091

(0.0

92)

(0.0

92)

(0.1

19)

(0.1

18)

(0.1

18)

(0.1

37)

Trad

e it–1

–0.6

32**

–0.6

70**

–0.7

31**

–0.9

41**

*–0

.941

***

–1.0

38**

*(0

.208

)(0

.206

)(0

.213

)(0

.242

)(0

.245

)(0

.262

)Po

pula

tion i

t–1

0.40

3***

0.38

6***

0.38

4***

0.27

8**

0.26

60.

115

(0.0

69)

(0.0

71)

(0.1

02)

(0.0

85)

(0.0

86)*

*(0

.130

)Po

lity i

t–1

–0.0

89**

*–0

.047

*–0

.091

***

–0.0

280.

008

–0.0

30(0

.017

)(0

.023

)(0

.020

)(0

.023

)(0

.025

)(0

.027

)R

egim

e du

rabi

lity i

t–1

–0.0

21**

*–0

.021

***

–0.0

22**

*0.

002

0.00

30.

004

(0.0

05)

(0.0

05)

(0.0

05)

(0.0

11)

(0.0

10)

(0.0

11)

Civ

il w

arit–

11.

717*

**1.

699*

**1.

714*

**1.

636*

**1.

649*

**1.

726*

**(0

.348

)(0

.334

)(0

.324

)(0

.312

)(0

.320

)(0

.326

)W

arit–

11.

185*

1.10

8*1.

322*

**0.

169

0.00

3–0

.231

(0.5

28)

(0.4

36)

(0.3

30)

(0.5

73)

(0.6

69)

(0.8

55)

(con

tinue

d)


Tabl

e I.

(con

tinue

d)

All s

tate

sSt

ates

repr

essiv

e at t

he ti

me o

f rat

ifica

tion

Vari

ables

(1) B

ase

mod

el(2

) Dem

ocra

cy(3

) Civ

il so

ciet

y(1

) Bas

e m

odel

(2) D

emoc

racy

(3) C

ivil

soci

ety

N3,

345

3,34

52,

764

1,99

81,

998

1,64

2C

hi2

249.

22**

*26

6.87

***

310.

55**

*18

7.38

***

218.

6***

165.

14**

*Lo

g lik

elih

ood

–4,0

33.1

7–4

,006

.65

–3,2

59.9

2–2

,561

.64

–2,5

56.7

5–2

,095

.47

+ p

#0.

1; *

p#

0.05

; **

p#

0.01

; ***

p#

0.00

1. N

umbe

rs in

par

enth

eses

are

Hub

er–W

hite

sta

ndar

d er

rors

.


the number of nongovernmental organiz-ations registered in each state according tothe Union of International Associations.19

Consistent with Hafner-Burton & Tsutsui(2005), Neumayer (2005), and Simmons(2006), we find that human rights treaties,specifically the CCPR, are most effectivewhen ratified by states with democraticsystems of government or strong civil societyadvocates. We now use these replications asour base for hypothesis testing.

In order to test our first hypothesis, we con-sider whether states that were repressive at thetime they ratified the CAT or the CCPR havereformed one year after commitment (repres-sori), improving their human rights practices.20

We accordingly re-estimate Equation (2),observing only states that were repressive at thetime of treaty ratification. Column 4 of Table Ireports our findings. Repressors that acceptlegal norms to protect human beings, ratifyingeither the CAT or the CCPR, are not likely,with any degree of confidence, to reform theirpractices after their commitments (Hafner-Burton & Tsutsui, 2005). Moreover, when wetest whether this hypothesis is conditional ondemocracy and civil society, as previousresearch assumes, we find that neither treaty iseffective even when they are ratified by repres-sive states on the more democratic end of thespectrum or by repressive states with strong civilsociety; previous research, it seems, has locateda set of conditional effects that matter least forthose states that need reform the most.21

Moreover, we see some evidence to suggest that,among repressors, the more democratic statesthat ratify the CAT are actually likely to haveworse practices. Estimates also suggest to usthat states that have higher per capita incomes

and trade, are free from civil war, and aregoverned by democratic institutions are morelikely to respect human rights – findings con-sistent with many studies before ours(Henderson, 1991; Poe, Tate & Keith, 1999;Richards, Gelleny & Sacko, 2001).22 Repressorstates, it seems, are keen to join the treatyregime but not equally enthusiastic to imple-ment those commitments, at least not rightaway, although the regime could be working tokeep them from worse brutality. Yet, as we havementioned, core theories of compliance suggestthat implementation is likely to happen onlyover some unknown but potentially considerableperiod of time, and the path to reform may be abumpy (rather than an upward trending) one.

Our next concern, correspondingly, is totest our second hypothesis to determinewhether repressive states that have ratifiedeither treaty put reforms into practice overtime. Socialization toward internalization ofthese norms, as well as the capacity to imple-ment reforms, may simply take time formany governments not accustomed to pro-tecting human rights or without the properresources. Moreover, this dynamic processmay not be steadily increasing. In order totest our second hypothesis, we run the samebase model (Column 4 of Table I) 14 moretimes, now replacing CATit–1 and CCPRit–1with lags from 2 to 15 years.23 For the pur-

19 We do not include civil societyit–1 into the base model,because several hundred observations are missing; we laterinclude the variable as a robustness check on our results.20 Consistent with our earlier analyses, we again define thispopulation as any state for which repressioni = 1 at the timeof ratification.21 Of the states observed in this sample, 20% are repressorswith democratic systems of government; 47% are repres-sors with above average links to civil society.

22 Although our primary concern is not to explain variationin effect of human rights treaties on protectors versus repres-sors, it is worth noting that comparing results for repressorsand protectors shows interesting variations. For instance,protectors with higher GDP per capita and greater regimestability are more likely to improve their human rights prac-tices, while these effects disappear among repressors. Also,ratification of the CCPR has a modest positive impact onpractice, as does Polity. Conversely, trade can improve prac-tice among repressors, but the effect does not stand amongprotectors. Equally interesting is the finding that populationpressure and civil war have worsening effects on practice inboth repressors and protectors. 23 We have chosen a period of 15 years in order to coverthe period of time observed in our sample between theentry into force of the CAT (1987) and the last year of ourdata observation (2003). Lags allow us to observe newinformation about compliance to assess the ‘bumpiness’ ofthe path, although they have costs and are not an ideal formof dynamic modeling.


poses of efficiency, we have chosen to illus-trate our tests graphically by plotting ourcoefficients of substantive interest and theirconfidence intervals at 95%; all results areavailable in the replication file.

The solid black lines in Figures 2 and 3 plotthe coefficients on CATit–1 and CCPRit–1

across each lag model, controlling for all thevariables in Equation (2) and considering onlystates repressive at the time of treatyratification. The dotted lines represent theirconfidence intervals at 95%. Both figuresunquestionably show that point estimatesmay fall anywhere within the confidence

Figure 2. Point Estimates of the Effects over Time of CAT Ratification on Human Rights Reform: ForGovernments Repressive at the Time of Ratification, 15-Year Duration

−0.6

−0.4

−0.2

0

0.2

0.4

0.6

0.8

1

1.2

Lagged years

point estimates 95% confidence interval

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Figure 3. Point Estimates of the Effects over Time of CCPR Ratification on Human Rights Reform: ForGovernments Repressive at the Time of Ratification, 15-Year Duration

−1

−0.8

−0.6

−0.4

−0.2

0

0.2

0.4

0.6

0.8

Lagged years

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

point estimates 95% confidence interval


intervals which contain zero. Simply put,neither treaty has a systematic effect on repres-sive states’ behaviors even a decade and a halfafter commitment.

We investigate our findings further inFigure 4 by offering predicted probabilitiesto better clarify what these coefficients meansubstantively. In particular, we are con-cerned with understanding whether a stateguilty of the most radical forms of violations(repressionit = 5) at the time it ratified theCAT or the CCPR is more likely to reformas the years go by. This figure providesinsight on two questions: what is the proba-bility that an average violent repressor24 willreform human rights after joining varioushuman rights treaties and after various yearsof commitment?25

In our previous graphs, the black lines rep-resented point estimates. In Figure 4, the linesrepresent predicted probabilities calculatedfrom the point estimates (Column 4 in TableI) that an extreme repressor has undertakenany notable human rights reforms at the timeof our observation. We consider any move-ment toward categories 4, 3, 2, or 1 on repres-sionit to indicate reform, a very liberalinterpretation of improvement. Our predic-tions indicate that, without any global legalcommitments to protect human rights norms,the world’s most violent repressors are likely toundertake reforms about 50% of the time; thisis about the same chance that they will reformafter they have ratified either human rightstreaty.26 International law, it seems, does notincrease the chance of reform at all in most

Figure 4. Predicted Probabilities that the World’s Most Repressive States Will Reform Human Rightsover 15 Years

0.3

0.35

0.4

0.45

0.5

0.55

0.6

0.65

0.7

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

Lagged years

CAT CCPR none

24 We define average by custom as the mean of all variablesin Equation (2), excluding the dependent variable, repres-sionit–1 and our variables of substantive interest, CATit–1..Nand CCPRit–1..N-.25 Figures 4B and 4C, available in the data replication file,plot the same findings for less extreme forms of violations,repressionit = 4 and repressionit = 3. Findings are relativelyconsistent and so not reported here.

26 Although the probabilities for reform after 15 years ofCAT and CCPR ratification are slightly lower and higher,respectively, than for no commitment to treaties, these esti-mates fall within 95% confidence intervals and are accord-ingly indistinguishable from each other. Confidenceintervals for all predications are not shown here for sake ofvisual clarity but are available in our replication data.

terribly repressive states, although it mayprevent slippage into worse violations.

All told, this information suggests thathuman rights treaties are least effective inmaking improvements in precisely those statesthat need them the most and for which theywere originally designed – the world’s worstabusers. It is not clear whether the results aredue to a selection effect or the weak achieve-ments of the treaties themselves, and it is pos-sible that treaties are actually working a littleby preventing represents from abusing evenmore. Yet, if repressive governments are beingsocialized by or learning from the humanrights legal regime to believe something newabout human rights norms, either they aremostly learning the wrong message, thatrepression in the face of commitment to inter-national law is acceptable behavior, or social-ization to new beliefs is weak in the face ofleaders’ political incapacity to reform. Andthese roadblocks to better human rights prac-tices appear persistent over time.

Delegation of Authority

If commitments to the CAT and the CCPRprovide little in the way of motivations for mostrepressive governments to reform their humanrights behaviors in accordance with lawfulnorms, many of our critics, especially in the legaland policy communities, believe laws providingfor greater delegation can help. Indeed, theProtocol was explicitly designed to improve onand add to existing enforcement mechanisms.One such law is already available, the firstOptional Protocol to the InternationalCovenant on Civil and Political Rights (CCPROP). By itself, the CCPR’s monitoring capacityis very limited. Governments that belong to the

CCPR pledge only to submit reports to theHuman Rights Committee (commonly referredto as the Committee) evaluating the measuresthey have adopted in support of treaty norms.27

Without further consent from governments,however, the Committee only has the authorityto study the reports, submitted by governmentsunder scrutiny for violations, ask questions ofgovernment representatives, and respond withgeneral comments. A law with moderatelygreater delegation, however, is available.

When a state adopts the CCPR OP, thatgovernment makes an additional commitmentto recognize the competence of the Committee.Authority has been delegated to the Committeeto entertain complaints made against a state byindividual victims of a violation of any of therights protected by the treaty.28 The OptionalProtocol is certainly not hard law, but it doesdelegate greater external competences for over-sight and intervention and, therefore, placesmore substantial constraints on governments’authority to violate the norms to which theycommit. Are repressive states that commit tothe CCPR OP, a harder form of internationallaw, any more likely to reform?

We again seek to economize and present ourresults graphically rather than in table format.Figure 5 plots coefficients to interpret our findings.We calculate these coefficients from our coremodel, Equation (2), on the sample of repressorstates with the addition of a new variable,CCPROPit–1..N, where N equals 1–15 laggedyears.29 As before, we run 15 separate models, onefor each lag year, and we plot the results only of ourvariable of substantive interest.30 We find that thelikelihood of reform after deeper commitment to


27 Participating governments must submit these reportswithin one year of the entry into force of the CCPR or when-ever the Committee makes a request. The Secretary-Generalof the United Nations, after consultation with theCommittee, may circulate the reports to specialized UNagencies. Article 40 of the CCPR and subsequent Committeeresolutions describe this procedure in greater detail.

28 Article 2 of the Optional Protocol stipulates that indi-viduals who claim that any of their rights enumerated inthe Covenant have been violated and who have exhaustedall available domestic remedies may submit a written com-munication to the Committee for consideration.29 Over 50% of states that have ratified the CCPROptional Protocol were repressors.30 For example, in our first model, we rerun Equation (2)results, presented in Column 2 of Table I; however, we addone new variable, CCPROPit–1. In our second model, we rerunthe same equation, now adding CCPROPit–2, and so on.

human rights treaty law, ratifying the CCPROptional Protocol, is indistinguishable from zero.Simply put, delegating more authority for moni-toring the implementation of human rightsby providing victims with an official forum toarticulate the injustices they have suffered does notseem to matter for implementation – a result thatmay also be explained by the selection process.Repressive states that join simply are not listening,or, if they are, they lack the resources or politicalwill to take action.

Conclusion

Much human rights research is dedicated toshowing the ways in which internationalinstitutions, including human rights treaties,can make a difference, even inside horriblyrepressive regimes. Recent statistical studiesqualify this belief, showing the ways thattreaties matter but only under some circum-stances. They raise an important question:can human rights laws help in the most

severe cases of abuse? Is the problem that theworld’s worst repressors might need moretime to come under the sway of inter-national law, to build capacity to implement eforms, and to change the minds of theirperpetrators?

Our findings add to recent statistical find-ings, which discover some hope for effectamong a small group of states. We identify sys-tematic limitations to the human rights legalregime for precisely those states that need themost oversight – there in severe crisis. Is itenough that the human rights regime can helponly a small number of states and a small per-centage of the world’s population, leavingbehind those who are worst abused? How canwe build and support a better internationalregime that, after decades of binding commit-ments and various institutional reforms toboost compliance, is still not making a notice-able diffference in so many cases?

Evidence shows that international laws areworking in some democratic states with an


Figure 5. Point Estimates of the Effects over Time of CCPR OP Ratification on Human Rights Reformfor Governments Repressive at the Time of Ratification, 15-Year Duration

−1

−0.5

0

0.5

1

1.5

2

Lagged years

95% confidence intervalpoint estimates

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

active civil society. But we are confronted withoverwhelming evidence that these cases ofinfluence are not applicable to the world’s mostrepressive states; that, more often than not,repressive governments that formally committo international treaties protecting our mostbasic human rights never come close to reform;that socialization, persuasion, and learning, ifthey are taking place over time, are not teach-ing the right messages to the governments thatneed the most help or are weak in the face ofincapacity to implement; and that, in the fewinstances where new ways of acting are indeedbeing learned, leaders are not effectively imple-menting those ideas to better protect people.Moreover, the democratic scope conditions areeven narrower than previously thought.

These findings raise important questions. Ifcommitment to international law appears tohave no direct effect on reform for the world’sworst abusers, either because the treaties aretoo weak to do much good or the abusive statesmost open to reform are selecting not to jointhem, why are so many governments, organiz-ations, and human rights advocates concernedwith ratification? Why do the United Statesand the European Union make commitmentsto human rights laws a core principle of foreignpolicy, trade, and aid? Why do human rightsorganizations spend so many resources mobil-izing campaigns for membership? Why doesthe UN do the same? What is driving the selec-tion process? If treaties do not matter wherethey need to, why do states and advocates pushthem so fiercely? What other purposes andinterests are treaties serving? And what othertools can help the worst abusers reform?

References

Abbott, Kenneth W. & Duncan Snidal, 2000.‘Hard and Soft Law in InternationalGovernance’, International Organization 54(3):421–456.

Anderson, Craig A., 1989. ‘Causal Reasoning andBelief Perseverance’, in D. W. Schumann, ed.,Proceedings of the Society for Consumer Psychology.

Knoxville, TN: University of Tennessee(115–120).

Chayes, Abram & Antonia Handler Chayes, 1993.‘On Compliance’, International Organization47(2): 175–205.

Chayes, Abram & Antonia Handler Chayes, 1998.The New Sovereignty: Compliance with Inter-national Regulatory Agreements. Cambridge, MA:Harvard University Press.

Cleveland, Sarah H., 2001. ‘Norm Internalizationand U.S. Economic Sanctions’, Yale Journal ofInternational Law 26(1): 1–102.

Cole, Wade M., 2005. ‘Sovereignty Relinquished?Explaining Commitment to the InternationalHuman Rights Covenants, 1966–1999’,American Sociological Review 70(3): 472–495.

Downs, George W.; David M. Rocke & Peter N.Barsoom, 1996. ‘Is the Good News AboutCompliance Good News About Cooperation?’,International Organization 50(3): 379–406.

Finnemore, Martha, 1996. ‘Norms, Culture, andWorld Politics: Insights from Sociology’sInstitutionalism’, International Organization50(2): 325–347.

Finnemore, Martha & Kathryn Sikkink, 1998.‘International Norm Dynamics and PoliticalChange’, International Organization 52(4):887–917.

Franck, Thomas M., 1988. ‘Legitimacy in theInternational System’, American Journal ofInternational Law 82(4): 705–758.

Gastil, Raymond, 1980. Freedom in the World:Political Rights and Civil Liberties, 1980. NewBrunswick, NJ: Transaction.

Gibney, Mark & Matthew Dalton, 1996. ‘ThePolitical Terror Scale’, in Stuart S. Nagel& David Louis Cingranelli, eds, Human Rightsand Developing Country. Greenwich, CT: JAI.

Gibney, Mark; Vanessa Dalton & Marc Vockell,1992. ‘USA Refugee Policy: A Human RightsAnalysis Update’, Journal of Refugee Studies5(1): 33–46.

Goodliffe, Jay & Darren Hawkins, 2005.‘Explaining Commitment: States and theConvention Against Torture’, Journal of Politics68(2): 358–371.

Goodman, Ryan & Derek Jinks, 2003.‘Measuring the Effects of Human RightsTreaties’, European Journal of InternationalLaw 14(1): 171–183.


Hafner-Burton, Emilie M., 2005. ‘Trading HumanRights: How Preferential Trade ArrangementsInfluence Government Repression’, InternationalOrganization 59(3): 593–629.

Hafner-Burton, Emilie & James Ron, 2006.‘Seeing Double: Human Rights ThroughQualitative and Quantitative Eyes’, unpub-lished manuscript (http://www. princeton.edu/~ehafner/).

Hafner-Burton, Emilie & Kiyoteru Tsutsui,2005. ‘Human Rights in a Globalizing World:The Paradox of Empty Promises’, AmericanJournal of Sociology 110(5): 1373–1411.

Hathaway, Oona A., 2002. ‘Do Human RightsTreaties Make a Difference?’, Yale Law Journal111(8): 1935–2042.

Hathaway, Oona A., 2003. ‘The Cost ofCommitment’, Stanford Law Review 55 (May):1821–1862.

Hathaway, Oona A., 2005. ‘The Promise andLimits of the International Law of Torture’, inOona A. Hathaway & H. H. Koh, eds,Foundations of International Law and Politics.New York: Foundation (228–237).

Henderson, Conway W., 1991. ‘ConditionsAffecting the Use of Political Repression’,Journal of Conflict Resolution 35(1): 120–142.

Henkin, Louis, 1979. How Nations Behave: Lawand Foreign Policy. New York: ColumbiaUniversity Press.

Jaggers, Keith & Ted Robert Gurr, 1995.‘Tracking Democracy’s Third Wave with thePolity III Data’, Journal of Peace Research32(4): 469–482.

Keith, Linda Camp, 1999. ‘The United NationsInternational Covenant on Civil and PoliticalRights: Does It Make a Difference in HumanRights Behavior?’, Journal of Peace Research36(1): 95–118.

Keohane, Robert O.; Andrew Moravcsik & Anne-Marie Slaughter, 2001. ‘Legalized DisputeResolution: Interstate and Transnational’,International Organization 54(3): 457–488.

Koh, Harold Hongju, 1996–97. ‘Why DoNations Obey International Law’, Yale LawJournal 106(8): 2599–2659.

Koh, Harold Hongju, 2002. ‘A United StatesHuman Rights Policy for the 21st Century’,Saint Louis University Law Journal 46(2):293–344.

Landman, Todd, 2005. Protecting Human Rights:A Comparative Study. Washington, DC:Georgetown University Press.

Long, J. Scott, 1997. Regression Models forCategorical and Limited Dependent Variables.Thousand Oaks, CA: Sage.

Lutz, Ellen L. & Kathryn Sikkink, 2000.‘International Human Rights Law andPractice in Latin America’, InternationalOrganization 54(3): 633–659.

Mearsheimer, John, 1994/1995. ‘The False Promiseof International Institutions’, InternationalSecurity 19(3): 5–49.

Meyer, John W.; John Boli, George M. Thomas &Francisco O. Ramirez, 1997. ‘World Societyand the Nation-State’, American Journal ofSociology 103(1): 144–181.

Mitchell, Ronald B., 1993. ‘Compliance Theory:A Synthesis’, Review of European Communityand International Environmental Law 2(4):327–334.

Moravcsik, Andrew, 1995. ‘Explaining InternationalHuman Rights Regimes: Liberal Theory andWestern Europe’, European Journal of Inter-national Relations 1(2): 157–189.

Moravcsik, Andrew, 2000. ‘The Origins of HumanRights Regimes: Democratic Delegation inPostwar Europe’, International Organization54(2): 217–252.

Neumayer, Eric, 2005. ‘Do International HumanRights Treaties Improve Respect for HumanRights?’, Journal of Conflict Resolution 49(6):925–953.

Poe, Steven C. & C. Neal Tate, 1994. ‘Repressionof Human Rights to Personal Integrity in the1980s: A Global Analysis’, American PoliticalScience Review 88(4): 853–872.

Poe, Steven C.; C. Neal Tate & Linda CampKeith, 1999. ‘Repression of the Human Rightto Personal Integrity Revisited: A GlobalCross-National Study Covering the Years1976–1993’, International Studies Quarterly43(2): 291–313.

Powell, Walter W. & Paul J. DiMaggio, eds, 1991.The New Institutionalism in OrganizationalAnalysis. Chicago, IL: University of ChicagoPress.

Richards, David L.; Ronald D. Gelleny & David H.Sacko, 2001. ‘Money With a Mean Streak?Foreign Economic Penetration and Government


Respect for Human Rights in DevelopingCountries’, International Studies Quarterly 45(2):219–239.

Risse, Thomas; Stephen C. Ropp & KathrynSikkink, eds, 1999. The Power of Human Rights:International Norms and Domestic Change. NewYork: Cambridge University Press.

Ron, James, 2000. ‘Savage Restraint: Israel,Palestine and the Dialectics of LegalRepression’, Social Problems 47(4): 445–472.

Scott, W. Richard; John W. Meyer & John Boli,1994. Institutional Environments and Organi-zations: Structural Complexity and Individualism.Thousand Oaks, CA: Sage.

Simmons, Beth, 2006. ‘Complying with the Law:The Case of International Human RightsTreaties’, unpublished manuscript, HarvardUniversity.

Slusher, Morgan P. & Craig A. Anderson, 1996.‘Using Causal Persuasive Arguments ToChange Beliefs and Teach New Information:The Mediating Role of ExplanationAvailability and Evaluation Bias in theAcceptance of Knowledge’, Journal ofEducational Psychology 88(1): 110–122.

Strange, Susan, 1982. ‘Cave Hic Dragones! ACritique of Regime Analysis’, InternationalOrganization 36(2): 479–496.

Vreeland, James Raymond, 2007. ‘Political Institutionsand Human Rights: Why Dictatorships Enter Intothe United Nations Convention Against Torture’,International Organization.

Waltz, Kenneth N., 1979. Theory of InternationalPolitics. New York: McGraw-Hill.

Wotipka, Christine Min & Kiyoteru Tsutsui,2001. ‘Global Human Rights and StateSovereignty: Nation-States’ Ratification ofInternational Human Rights Treaties,1965–1999’, presented at the annual meetingsof the American Sociological Association,Anaheim, CA, 18–21 August.

EMILIE M. HAFNER-BURTON, b. 1973,PhD in Political Science (University ofWisconsin, Madison, 2003); Assistant Professor,Woodrow Wilson School of Public andInternational Affairs and Department of Politics,Princeton University (2006– ).

KIYOTERU TSUTSUI, b. 1971, PhD inSociology (Stanford University, 2002); AssistantProfessor, SUNY at Stony Brook (2002– ); SSRCAbe Fellow (2004–05); Visiting Assistant Professor,Stanford University (2005–06); current main inter-est: global human rights and social movements.


Justice Lost! The Failure of International Human Rights Law To Matter Where Needed Most

Documents

Transcript of Justice Lost! The Failure of International Human Rights Law To Matter Where Needed Most