reaching the individual: eu accession, ngos, and human rights · human rights monitors look harder...

American Political Science Review Vol 110 No 3 August 2016

doi101017S000305541600023X c⃝ American Political Science Association 2016

Reaching the Individual EU Accession NGOs and Human RightsANA BRACIC University of Oklahoma

C an human rights institutions influence individual behavior This article tests the ground leveleffectiveness of two strategies that aim to eliminate discrimination a powerful top-down combi-nation of incentives and norm promotion and a bottom-up NGO-based effort The study uses a

hard case that of discrimination against the Roma (commonly known by the disfavored term ldquoGypsiesrdquo)spans three towns Murska Sobota and Novo mesto in Slovenia and Cakovec in Croatia and includesaltogether 606 subjects Levels of discrimination are estimated via trust games played with money whichare particularly appropriate because the Roma are widely stereotyped as cheaters and thieves The findingssuggest that the EU accession process widely regarded as a strong incentive-based and norm promotingrights change mechanism may not substantially reduce discrimination on the ground Instead theysuggest that ground level organizing aimed at improving relations between Roma and non-Roma helpsreduce discrimination

Individual behavior is increasingly salient in thecontext of rights abuses Human rights monitorslook harder for abuse find abuse in more places

and classify different types of abuse as human rightsviolations (Fariss 2014) As those violations are morebroadly monitored and classified holding individualsaccountable is becoming a norm (Sikkink 2011) TheUS BlackLivesMatter movement reacting to policebrutality against black citizens both exemplifies thegrowing movement towards accountability and reflectsthe individual-level link between discrimination andother rights abuses Nevertheless the global rights com-munity has far to go in battling racial injustice and otherrights abuses at the level of the individual whetherdirected at Dalits in India individuals of Haitian de-scent in the Dominican Republic or members of anyother marginalized group Human rights scholars rarelysystematically gather individual-level data about hu-man rights behaviors to link them to broader humanrights institutions By examining efforts to eliminateindividual-level discrimination which itself often re-sults in further rights abuses this article aims to provideinsights into macro- and microlevel processes that po-tentially lead to positive human rights outcomes

Using the context of discrimination against theRoma in Slovenia and Croatia I test the ground-leveleffectiveness of two strategies The first strategy is atop-down combination of incentives and norm promo-tion I ask if it affects individual behavior I then test thesame ground-level effectiveness of the second strategya local NGO-led effort to improve contact betweenthose who discriminate and those who are targets of

Ana Bracic is Assistant Professor in the Department of PoliticalScience at the University of Oklahoma (bracicouedu) web wwwanabraciccom 301 821 5887 The appendix is available online atwwwanabraciccom

I would like to thank George Downs Michael Gilligan BernardaBracic and W Nicholson Price II for invaluable help in developingthis project I am also very grateful to Stefan Bajic Vanessa Bo-betic Dragan Bogdan Zeljko Balog Mojca Becaj Jozek HorvatMuc Manja Munda Ziva Kleindienst Vera Klopcic Jaka KukavicaRebecca Morton Tena Posel Sabina Ranogajec Monika SandreliDavid Stasavage Filip Skiljan Eva Trapecar Joshua Tucker MilenaTudija Lucija Vihar the APSR editorial team and four anonymousreviewers and an anonymous Croatian government official

discrimination In testing the first strategy I conducta border study of two closely matched towns in Croa-tia and Slovenia The EU accession process providesincentives and norm promotion in Croatia but not inSlovenia In examining the second I perform a within-country comparison of two closely matched Slovenetowns One has an NGO effort to improve contactand the other does not I measure discrimination atthe individual level in an experimental setup using thetrust game (Berg et al 1995)1

The top-down rights improvement strategy does notappear to reduce discrimination on the ground butthe bottom-up NGO effort does The findings suggestthat incentive-based and traditional norm-promotingstrategies while powerful at the state level may nothave effects that reach individual behavior The resultsfurther suggest that in targeting individuals a humanrights strategy that is experiential and not merely per-suasive might be particularly effective

Human rights scholarship tends to examine the twocomponents of the first strategy separately Incentive-based mechanisms are generally aimed at state actorsand affect rights practices through rights condition-ality in exchange for benefits like preferential tradeprivileges (Hafner-Burton 2005 Simmons et al 2008)Norm promotion addresses states and individuals alikeand affects rights practices and public opinion chieflyby naming and shaming (Brysk 1993 Davis et al2012 Hafner-Burton 2008 Hendrix and Wong 2012Keck and Sikkink 1998 Khagram et al 2002 Sikkink1993) While these two types of strategies are quitedistinct they are sometimes linked in practice Whena state commits to rights improvements in exchangefor externally granted benefits or when such benefitsare suspended in light of rights violations norm en-trepreneurs use the opportunity to mobilize In factsuspension of benefits may in part be due to namingand shaming Human Rights Watch for example re-ports extensively on inadequate uses of the AfricanGrowth and Opportunity Act (AGOA) conditional-ity in improving rights practices in various African

1 This study was approved by the New York University IRB (HS11-8405)

530))amp)())amp-13$))amp$+()$$)( ))))((+)

American Political Science Review Vol 110 No 3

countries (Kasambala 2014) When the United Statesrevoked Swazilandrsquos AGOA status due to insufficientprotections of workersrsquo and human rights the decisionaccordingly received much attention (Hughes 2014)The two strategies while chiefly operating at state andinternational levels at least partly rely on domesticaudiences to mobilize in favor of rights change (Keckand Sikkink 1998 Simmons 2009) An individual actorobserving or participating in public discourse howevermight not distinguish between incentives and normpromotion but simply perceive that rights should bechanged In studying individuals then the effects oftwo strategies that often appear together ought to beconsidered together as well

An alternative strategy to change discriminatory be-havior much smaller in scope than the approachesjust mentioned stems from NGO efforts that encour-age contact between the majority population and themarginalized group in order to foster inclusion Whilethe merits of intergroup contact remain contestedlaboratory-based scholarship in psychology stronglysuggests that intergroup contact reduces prejudice anddiscriminatory intent (Allport 1954 Dovidio et al2004 Miller 2002 Pettigrew 1998 Pettigrew and Tropp2006 Wagner et al 2008) Recent evidence from fieldexperiments offers additional support when whitechildren (Green and Wong 2009) or college students(Boisjoly et al 2006 Laar et al 2005) are randomlyassigned to diverse groups for short-term or extendedinteractions they show lower levels of prejudice Ef-forts on the part of NGOs to increase contact betweenthose who transgress and their targets at the individuallevel may then decrease prejudice and discriminationat that level as well Most research on intergroup con-tact is conducted in the American context predomi-nantly examining prejudice of whites towards blacksby studying discrimination in Europe this article offersa departure from this tradition

I conduct a test of the intergroup contact strategyand the combination of incentives and norm promotionin the context of discrimination against the Roma inSlovenia and Croatia The Roma commonly referredto by the disfavored term ldquoGypsiesrdquo (Hancock 2002)are the largest ethnic minority in Europe Historicallythe Roma have been variously enslaved deportedforcibly assimilated and subjected to genocide (Barany2002) Today many remain segregated and face dis-crimination in schools on the job market in hospitalsin police stations and on the street (European RomaRights Centre (ERRC) 2004) This study examines theRoma because they are a clear case of a minority thathas suffered discrimination abuse and exclusion forcenturies In such a seemingly intractable case identi-fying interventions that promote inclusion would showpromise for studying and curtailing marginalization invarious other communities that historically have beendenied or have themselves resisted inclusion

My empirical strategy involves two approaches aborder study and a within-country design I use a bor-der study to examine the combination of incentivesand norms The EU accession process is arguably thestrongest mechanism of human rights conditionality

and is deemed particularly powerful in the contextof Roma rights (Schimmelfennig et al 2005) To gainmembership EU candidate states must enact legisla-tive and policy changes that protect Roma rights atstate and local levels Conditionality demands are ac-companied by norm advocacy and naming and sham-ing via NGOs and official EU channels I leverage thisprocess and measure discrimination against the Romain two towns one in Croatia and one in Slovenia thatmatch closely on 43 Roma and human rights relatedfactors at state regional and local levels At the timeof investigation Croatia was at the height of the EUaccession process and thus faced enormous pressures toimprove its treatment of the Roma In contrast Slove-nia had been an EU member for eight years and facedno pressure to improve Roma rights

I test the intergroup contact strategy in the context ofdiscrimination against the Roma in two towns in Slove-nia that match exceptionally well on 12 localregionalRoma and human rights-related factors but differin type of NGO action Action in the first town isinclusivemdashtargeting both Roma and non-Romamdashandpromotes intergroup contact NGO action in the sec-ond town is not inclusive and focuses on service provi-sion to Romani settlements Accordingly roughly halfof randomly sampled non-Roma from the first townare familiar with the NGO in the other town the sameholds for only two percent NGO leadersrsquo personal id-iosyncrasies completely define the organizationsrsquo fociThis mitigates the endogeneity concern that the twotypes of NGOs developed because of beliefs that theirtypes would be particularly effective in their respectiveenvironments

To test the two hypotheses I use simple gamesintended to capture other-regarding behavior whichmap remarkably well onto the stereotype that moti-vates anti-Roma sentiment Discrimination against theRoma is largely motivated by the belief that the Romaare cheaters and thieves The trust game (Berg et al1995) played with money elicits behavior by non-Romathat demonstrates distrust towards the Roma given thestrong link between distrust and discrimination in thisparticular case the method likely captures discrimina-tory behavior

I find no support for the hypothesis that a high-level combined strategy of incentives and norm pro-motion affects individual behavior At the height of theEU accession process Croats discriminated against theRoma while Slovenes just across the border under nopressure to improve their treatment of the Roma didnot I find support however for the intergroup contacthypothesis Slovenes from a town with inclusive NGOaction treated Roma no differently than non-Romawhereas Slovenes from a town with noninclusive NGOaction discriminated against the Roma

My findings suggest that (1) while powerful at thestate level incentive-based and traditional norm pro-moting strategies inherent in the EU accession processmay not necessarily have effects that reach the indi-vidual and (2) NGOs promoting intergroup contactbetween Roma and non-Roma can help reduce dis-crimination In the context of changes at the individual

531))amp)())amp-13$))amp$+()$$)( ))))((+)

Reaching the Individual August 2016

level then microlevel strategies that engage individu-als as participants in normative behavior may be prefer-able to (otherwise powerful) macrolevel strategies thatmerely advocate for it

Quantitative human rights scholarship tends to viewrights violations in a top-down manner with the stateas an abuser and the citizen a victim Some cases ofabuse are not as clear-cut State actors may be theprimary perpetrators but their behavior may reflectthe general tenor of the environment For examplecitizens may implicitly or explicitly support abuse InIndia government food security programs are oftensituated in locales that exclude Dalits if not Dalit chil-dren are often chased away from schools when they tryto take advantage of the midday meal scheme to whichthey are entitled (Thorat and Lee 2005) Actions ofparents teachers and local administrators are inter-twined parents mobilize in favor of Dalit exclusion(Thorat and Lee 2005) while the officials curb the chil-drenrsquos right to food Further citizens can themselvesbecome state actors Consider American juries In mur-der cases of white victims black defendants are signifi-cantly more likely than white defendants to receive thedeath sentence (Baldus et al 1998) particularly so ifthey look more stereotypically black (Eberhardt et al2006) Some cases of human rights violations then de-mand that we examine individual actors both state andnonstate if we aim to understand and improve humanbehavior This article offers a small contribution to thateffort

In what follows I first introduce the Roma NextI present the two strategies of human rights changealong with expectations about their efficacy at the levelof the individual The empirical strategy results anda broader discussion of implications follow The lastsection concludes

WHO ARE THE ROMA AND WHY USE THEIRCASE HERE

The Roma are the largest ethnic minority in EuropePopulation counts are unreliable conservative esti-mates report that as many as 9 million Roma currentlyreside in Europe (Polzer-Srienz 2003) The Roma firstarrived in Europe at the end of the 13th century havingemigrated from north-western India centuries before(Courthiade 2003) Contrary to romanticized popu-lar perceptions they are generally no longer itinerant(Matras 2000)mdashsome groups in fact were never peri-patetic2 Until very recently many engaged in tradi-tional economic activities which primarily defined theirtribal identities3 While modernization rendered manyof those crafts obsolete tribal diversity survives and toa large extent characterizes Romani individuals Thelargest concentration of Roma in Eastern Europe is inRomania followed by Hungary Historically the Roma

2 All facts attributed to Barany (2002) unless otherwise noted3 Trades ranged from blacksmiths and umbrella makers to horsetraders (Strukelj 1980)

have been variously enslaved4 deported5 forcibly as-similated6 and subjected to genocide (Djuric 2007)

While today many Roma have integrated into theirrespective majority populations most remain segre-gated The segregated populations are generally so-cially disadvantaged and bear the brunt of discrimi-nation In many states Roma children are schooled inremedial special schools for mentally disabled whetheror not they have actual disabilities (Cahn 2002) Romaface significant barriers in accessing employment fre-quently live without electricity or sewerage gener-ally receive substandard health care and are oftendeemed undeserving of social welfare Many do nothave personal documents and are effectively statelessthey are more likely than non-Roma to be abused bythe police in general while in detention and while inprison (ERRC 1997) These marginalized populationsare therefore predominantly poor unemployed andundereducated To survive numerous Roma are forcedto engage in illicit activities that range from small-scaletheft of scrap metal to usury and transnational traffick-ing in weapons drugs and humans (Anonymous 2012)Criminality fuels the already existing intolerance anddiscrimination and in a downward spiral strengthensthe barriers that drive the Roma to crime in the firstplace

HYPOTHESIS I THE EU ACCESSIONPROCESS

The EU accession process is a prime example of aprocess that aims to influence rights through a com-bination of incentives and norm promotion (Kelley2004) EU conditionality is a powerful incentive-basedmechanism (Hafner-Burton 2005 Vachudova 2005)The requirements are generally non-negotiable thebenefits substantial and the costs of exclusion consid-erable (Schimmelfennig et al 2005) Although citizensupport for accession varies considerably eligible statestypically choose to accede (Tucker et al 2002) Statesreceive financial support from the EU to implement thechanges required (Phare 2005) face yearly evaluationsby the European Commission and do not receive mem-bership until they have sufficiently complied with theconditions The process itself is not without challengesBulgaria and Romania for example gained member-ship with outstanding commitments and face postac-cession monitoring in the areas of judicial reform cor-ruption and organized crime (Vachudova 2009) ForRoma rights however there is arguably no strongertop-down mechanism of change

Accession requirements strongly reflect EUrsquos in-creasing concern for the Roma In addition to adopt-ing comprehensive antidiscrimination legislation can-didate states must also make changes to policyand practice When evaluating compliance the EU

4 In Moldavia and Walachia Romani slavery persisted until 18645 Britain and Portugal deported Roma to the colonies in the 18thcentury6 The Austro-Hungarian Empire and European socialist regimesforcibly assimilated Roma

532))amp)())amp-13$))amp$+()$$)( ))))((+)


Commission visits Romani settlements and interviewslocal Roma as well as Roma rights experts regardingprogress (Anonymous 2014) The visits and the result-ing reports are crucial to both strategies of rights im-provement In the context of incentives the reportsevaluate compliance with the conditions detailing thechanges required before membership can be grantedIn the context of norm promotion they amplify reputa-tional concerns by offering official praise or criticism bythe Commission (Finnemore and Sikkink 1998 Kelley2004) The information they provide is then availableto third parties to further apply normative pressure

Norm promotion of course extends beyond shamingstrategies The EU organizes numerous Roma rightsinitiatives (Baluh 2012 European Commission 2012b)actively engages with Roma rights advocacy networks(European Commission 2012b) and puts Roma issueson national regional and local agendas of candidatestates Activists describe the accession period as anldquoopening of spacerdquo in which Roma rights become apart of a much larger international agenda and thusgain much needed legitimacy (Balazek 2012 Tudija2012) Incentive-based and norm promoting strategiesare therefore intertwined and generally scholars findthe improved treatment of minorities in Eastern En-largement states a success (Gelazis 2004 Kelley 2004Pridham 2008 Tesser 2003 Vachudova 2005)

Why EU Accession may InfluenceIndividuals

Accession requirements include efforts to reduceindividual-level discrimination against the Roma Of47 EU Commission reports on Bulgaria Romania Slo-vakia Czech Republic Hungary Slovenia and Croa-tia 36 explicitly discuss discriminatory attitudes amongcitizens (European Commission 2014) Pressured toreduce ground-level discrimination states and munic-ipalities generally with EU financing engage in pro-gramming on antidiscrimination awareness and advo-cacy (Anonymous 2014) The ldquoPolice and Citizensrdquoproject successfully carried out in Osijek-Baranja andVukovar-Sirmium in Croatia and evaluated by the EUCommission is one example of such programming(Ministarstvo unutarnjih poslova 2006) countless oth-ers resulting in posters ads and brochures make EUrsquosefforts quite visible on the ground7

In addition to responding to programming citizensmay directly respond to EUrsquos expectations regard-ing minority protections Accession to the EU is anextremely salient political issue Immediately beforeCroatia closed accession negotiations for example 75percent of 100 randomly sampled Croats knew thathuman rights were among the conditions for entry and44 percent pointed out minority rights specifically8 Infact respect for human rights was identified more timesthan any other negotiation chapter justice freedomand security came as a close second with 69 percent of

7 See Online Appendix at wwwanabraciccom for a campaign sam-ple8 Findings are part of this study

participants identifying it as a requirement Whetherin response to programming or public discourse moregenerally Croats knew that rights play a role in EUaccession

Postaccession Backsliding

Roma rights experts claim that the accession processonly temporarily reduces discrimination which returnsonce the process is complete and the state nameda member of the EU (Daniel 2010 Jovanovic 2010Oravec 2010 Tichy 2010) While findings regardingbacksliding on accession reforms diverge (Dimitrovaand Toshkov 2009 Hollyer 2010 Levitz and Pop-Eleches 2010 Meyer-Sahling 2008 Pridham 2008) theEU neither systematically monitors its member stateswith respect to Roma rights protection nor enforces thestandards it sets for candidate states (De Witte 2003Vachudova 2005) At the level of the individual sup-port for reforms and the resulting lack of backsliding islinked to the exposure of citizens to the desired policiesin place in Western European countries (Levitz andPop-Eleches 2010) Such positive exposure is much lesslikely in the context of Roma rights however instead ofobserving exemplary rights protection citizens witnessrights violations that range from refusing state entryto Roma visitors (United Kingdom) (ERRC 2001) torazing of settlements (Italy) (ERRC et al 2008) andforced deportations (France) (Erlanger 2010)

With the absence of previously strong conditional-ity and evidence of discrimination in old EU memberstates new members likely backslide in their respectfor rights The first hypothesis is therefore as follows

H1 A state actively undergoing the EU accession processhas lower levels of ground-level discrimination than a statenot undergoing the process

HYPOTHESIS II INCLUSIVE NGO ACTION

NGO action that aims to eliminate discrimination bypromoting intergroup contact (Allport 1954) is a mech-anism that is neither as expansive nor as frequentlyused as incentives and norm promotion Scholarshipfrom psychology and political science however sug-gests that Romanon-Roma interaction could help re-duce prejudice and discriminatory behavior by non-Roma Further bottom-up efforts to eliminate deeplyentrenched abusive practices can be successful wheretop-down measures fail (Mackie 1996)

Work on intergroup contact originally intended tocapture the effects of contact on racial and ethnicprejudice and focused predominantly on the interac-tions between white and black Americans (Deutschand Collins 1951) A seminal study from the AmericanSouth for example paired racially prejudiced whiteyoung adults to work with a black and a white co-worker on a railroad management project for a month(Cook 1971) After the study the treated participantsrated their black co-workers highly in competencelikeability and attractiveness several months later the

533))amp)())amp-13$))amp$+()$$)( ))))((+)


treated participants also expressed less racial prejudicethan control subjects did

Contemporary studies in psychology tend to belaboratory-based and generally suggest that intergroupcontact reduces prejudice often by reducing feelings ofintergroup anxiety (Blair et al 2003 Blascovich et al2001 Brown and Hewstone 2005 Mendes et al 2002)Psychologists also show that prejudice leads to discrim-inatory behavior (Dovidio et al 2004 Schutz and Six1996) and that an increase in intergroup contact resultsin a decrease in both prejudice and discriminatory in-tent (Wagner et al 2008) Moreover intergroup contacteffects tend to generalize beyond participants in theimmediate contact situation people who experiencecontact may change their attitudes towards the entireoutgroup outgroup members in other situations andeven outgroups not involved in the contact situation atall (Pettigrew and Tropp 2006)

Findings on intergroup contact are not conclusivehowever For example whitesrsquo affect towards blackschanges more easily than their beliefs about them(Jackman and Crane 1986) In addition socioeconomicstatus appears to offset race-based status differen-tial rendering whitesrsquo attitudes less negative whentheir black friends have a higher than equal socioeco-nomic status Elsewhere higher levels of contact withblack and Asian peers over four years led white stu-dents at a liberal arts university to solidify stereotypedperceptions of those minorities (Rothbart and John1993)

Recent experimental evidence challenges these find-ings First-year college students who were randomly(and nonrandomly) assigned minority roommatesshow lower levels of prejudice both short- and long-term The effects persist across various combinationsof ethnic groups including those of only minorities(Boisjoly et al 2006 Laar et al 2005) Another studyrandomly assigned white teenagers to racially homoge-nous and heterogeneous camping expedition groups Amonth after the 2ndash3 week trip the participants were sur-veyed over the telephone the white teenagers assignedto the heterogeneous group described themselves asless prejudiced than participants from the homoge-neous group did and reported significantly lower levelsof antiblack and antigay sentiment (Green and Wong2009)

Although evidence in favor of intergroup contactis not consistent and much of it is laboratory-basedmdashand thus at risk of being too far removed from every-day interactions (Paluck and Green 2009)mdashthe find-ings strongly suggest that Romanon-Roma contactcould lead to reducing prejudice at the level of theindividual

NGOs present one possible environment for activi-ties that facilitate Romanon-Roma interaction Whilemany Roma rights NGOs came into existence duringand after transition in Central and Eastern Europe fewactively promote intergroup contact Chiefly NGOs fo-cus on service provision their second most likely focusis Roma rights advocacy As Romani communities arein want of both services and rights advocacy NGOsaiming to promote intergroup contact face challenges

in obtaining funding (Anonymous 2013)9 Lab-basedevidence and extant experimental work however sug-gest that it is precisely those efforts that can lead to adecrease in prejudice and discrimination at the level ofthe individual The second hypothesis is therefore asfollows

H2 Effective ground level organizing aimed at improvingRomanon-Roma relations reduces discrimination againstthe Roma

It was impossible to test the two hypotheses with-out collecting original data To avoid ethnic stereotyp-ing most Eastern European states no longer collectdata based on ethnicity if they do the data are notavailable to the general public (Daniel 2010 Hojsik2010 Ripka 2010 Tichy 2010) Even the most basicpopulation counts of Roma are grossly inconsistentwhile rights activists tend to overestimate populationcounts official estimates are often too low resulting ingaps as wide as 25 million people (Romania Barany2002) Since systematic and reliable cross-time quan-titative data on individual-level discrimination againstthe Roma are unavailable I constructed a measureof discrimination described after the section on caseselection

CASE SELECTION

I measured discrimination levels in two overlappingpairs of towns one pair per hypothesis While differingon the crucial covariatemdashEU accession process or in-clusive Roma NGO activitymdashthe two town pairs matchclosely on a set of factors that may influence human andRoma rights at state regional and town levels To se-lect the towns I performed nearest neighbor matching(Ho et al 2007a 2007b Nielsen 2014)10 using sevencovariates on all towns in Slovenia and Croatia thatdid not experience war violence in the Yugoslav warsin the early 1990s (The Central Intelligence Agency2002) and have a Roma population of at least 5011 I ex-cluded locations that experienced war violence in orderto control for the aftermath of ethnic contention thatmay have been more though not uniformly present incertain Croatian locations12

9 In addition to the NGO included in this study I have been ableto identify only two others that focus on intergroup contact OneVzajemne Souzitı supports an artificially integrated Romanon-Roma community in Ostrava Czech Republic The other PolicyCenter for Roma and Minorities from Ferentari Romania promotescontact among Roma and non-Roma children through alternativeeducation and soccer10 As I was looking to find only a few closely matched pairs buthad cases that varied too much to merit exact matching I used theldquogreedyrdquo matching method of nearest-neighbor matching in MatchIt(Ho et al 2011) This method finds the closest control match for eachtreated unit one at a time and was therefore more appropriate thanmethods that seek to minimize average absolute distance across allpairs11 I excluded locations with fewer than 50 Roma because I needed50 Roma participants from each location12 Anti-Roma sentiment was not central to wartime ethnic strifebut current prejudice may be heightened by residual nationalistsentiment

534))amp)())amp-13$))amp$+()$$)( ))))((+)


Of the seven covariates used in matching town pop-ulation Roma inhabitants as proportion of total popu-lation and ethnic majority as proportion of total pop-ulation were used to achieve balance in town sizes andtheir ethnic compositions Proximity to SloveneCroatborder and a triborder region dummy (HypothesisI) were included as an additional control for varia-tion in proximity to outgroups People from border-lands are significantly more likely than inlanders tohave contact with people across the border (Mirwaldt2010) which may increase the saliency of ingroup sen-timent (Branton et al 2007 Mirwaldt 2010) Ingroupfavoritism can in turn give rise to intergroup discrim-ination irrespective of attitudes toward specific out-groups (Brewer 2007) Controlling for increased prox-imity to a border therefore accounts for a potentialsource of variation in intergroup discrimination moregenerally

A regional capital dummy was included as a proxyfor resources and institutional capabilities at the mu-nicipal level both general and those pertaining to theRoma Of the 24 Slovene municipalities in which Romalive for example only four have a municipal strategyfor addressing the needs of their respective Roma com-munities (Vlada RS 2014) Both regional capitals inthe sample have such a strategy Finally dummies forthe EU accession process (Hypothesis I) and inclusiveRoma NGO action (Hypothesis II) selected for the twomain categories tested

The pair of towns used to test Hypothesis I was se-lected first the pair used to test Hypothesis II wasselected second To test the two hypotheses with thelowest possible number of subjects participants wererecruited from three towns data from one town istherefore used to test both hypotheses Matching sin-gled out Cakovec in Croatia and Murska Sobota inSlovenia as the best pair to test Hypothesis I MurskaSobota and Novo mesto also in Slovenia were selectedas the best pair to test Hypothesis II (see Figure 1 fora map) Beyond the factors used in matching the lo-cations match on 38 additional relevant characteristicsThe remainder of this section first discusses how thetowns differ on the crucial covariates and then lists theshared characteristics

Hypothesis I The EU Accession Process

A meaningful study of discrimination in the context ofEU accession and membership would necessarily spanseveral years if limited to one country To test Hypothe-sis I within a short time span I looked at cross-sectionalsnapshots of discrimination in towns of which one wasundergoing the accession process and the other hadalready completed it As Roma rights experts stronglyemphasized that any beneficial effects of the accessionprocess were short lived I constructed Hypothesis Iand chose the towns with the aim of evaluating thatclaim Conclusions drawn from this test will thereforespeak to levels of discrimination in the context of EUaccession and membership but will not speak to dis-crimination in the preaccession stage

At the end of June 2011 the EU closed negotiationsfor membership with Croatia (European Commission2011) Chapter 23 the negotiation chapter most rele-vant to Roma rights was among the last three to beclosed (European Commission 2010) in early summer2011 when Croatia was still reminded of the challengesfaced by the Roma minority in the context of accessionnegotiations this project was already in place I wastherefore able to capture peoplersquos attitudes towardsthe Roma during a particularly critical period whenCroatiarsquos treatment of Roma rights was among the lastfew things keeping Croatia from the EU13 For reasonslisted below the cleanest and most compelling cross-national comparison to a town in Croatia is a town in itsnorthern neighbor Slovenia Slovenia acceded to theEU in 2004 and since accession has experienced littlepressure if any to improve its treatment of the Roma(Vachudova 2005)

Hypothesis II Inclusive NGO Action

Testing Hypothesis II on the other hand utilizes awithin-country design Novo mesto and Murska Sobotain Slovenia see different types of Romani NGO action

Roma NGO activity in Murska Sobota aims to im-prove Romanon-Roma relations and is inclusive iteffectively engages the non-Roma just as much as itengages the Roma For example fairly early in itstenure Romani UnionndashZveza Romov in Murska Sob-ota began organizing events that would improve lo-cal awareness and relations between Roma and non-Roma Initially it was challenging to convince localnon-Roma that the events are intended for all butthe organization succeeded in 1992 with Ciganska noc(ldquoGypsy nightrdquo) an annual concert of Romani mu-sic and dances that traditionally evolves into a livelyparty (Horvat-Muc 2010 Sandreli 2012) Once thatbarrier was breached non-Roma began attending booklaunches plays workshops and the Romani summercamps as well A sister organization runs a Romaniradio Radio Romic and there too non-Roma lis-teners are invited to tune in Connecting Roma andnon-Roma through culture and awareness is the orga-nizationsrsquo chief objective which they have apparentlyattained14 Not only do Roma and non-Roma attendthe events together but the number of non-Romaamong event participants and radio listeners recentlysurpassed the number of Roma (Sandreli 2012) As itconnects Roma and non-Roma in a friendly nonthreat-ening context Romani Union likely lowers Romanon-Roma anxietymdashdoing precisely what recent literatureon contact intergroup anxiety and prejudice finds par-ticularly effective (Davies et al 2011 Page-Gould et al2008)

13 For a discussion on how significant progress tends to happen to-wards the end of the accession process see Pridham (2008)14 An additional and possibly vital characteristic of organizing inMurska Sobota is that Roma and non-Roma are always presented asequal Instead of generally proclaiming that the Roma need help theorganization demonstrates the ways in which while equal they aredifferent and interesting Entering the contact situation with equalstatus may additionally reduce bias (Moody 2001)

535))amp)())amp-13$))amp$+()$$)( ))))((+)


FIGURE 1 A Map of the Three Towns

Romano Veseli in Novo mesto in contrast is a ser-vice provision NGO and generally does not addressrelations between Roma and non-Roma The organiza-tion has a very strong presence in the Romani commu-nities and focuses on socioeconomic aid provision andefforts related to education of both Romani childrenand adults (Tudija 2012) Aside from an occasionalvolunteer non-Roma are not involved in the organiza-tionrsquos activities The leading activist is well known andrespected among the Roma in Novo mesto howeverevery randomly sampled Romani subject who partici-pated in the study was able to identify her by name andmany profusely praised her efforts The organization inNovo mesto is of somewhat lower capacity than that inMurska Sobota but is growing

Identification

As neither was randomly assigned expansive inclusiveorganizing in Murska Sobota and somewhat less expan-sive non-inclusive organizing in Novo mesto suggest anomitted variable bias namely the possibility that any

difference in discrimination today and the scope andtype of organizing in each town could be related toa difference in respect for Roma rights prior to orga-nizing The possibility of bias is weakened however byseveral decades of Yugoslav policies on minorities andmore importantly of equality in employment imposedupon all citizens (Baluh 2012 Siftar 1989) Before Ro-mani organizing began in Murska Sobota in early 1991(Horvat-Muc 2010) the Romani experiencemdashrangingfrom kindergarten (Balazek 2012 Horvat-Muc 2011aTancer 1997) and primary school attendance (Siftar1989 Tancer 1997) to a severe drop in employment(Balazek 2012 Klopcic 2012 Siftar 1989)mdashin the twotowns was as similar as ever

Further the nonrandom assignment of the type oforganizing in Murska Sobota and Novo mesto alsosuggests a possible endogeneity problem namely thatthe activists from both towns might have chosen theirfoci because they believed that those would be effectivewhile any others would be ineffective in their respectivetowns Extensive interviews with the leading activistshowever suggest that this is not the case Activists in

536))amp)())amp-13$))amp$+()$$)( ))))((+)


Murska Sobota focus strongly on Romanon-Roma re-lations because they are themselves passionate aboutthe issue and have been so from the very start (Horvat-Muc 2011a Sandreli 2012) Likewise activists in Novomesto focus on providing socioeconomic and educa-tional aid because they believe it to be by far the mostimportant cause (Tudija 2012) The choice of focusappears entirely driven by the idiosyncracies of theNGO leadersrsquo personalities and preferences render-ing the assignment of inclusive organizing almost ldquoasifrdquo random This substantially mitigates endogeneityconcerns

To establish the extent to which the non-Roma pop-ulation of each town is familiar with local Romani orga-nizing I asked 100 randomly sampled non-Roma fromeach town to identify it Forty-six percent of randomlysurveyed individuals in Murska Sobota were familiarwith local Romani NGO action15 In Novo mesto onlytwo percent were familiar with the same This findingconfirms the expectation from comparing types of or-ganizing in Murska Sobota and Novo mesto activismin Murska Sobota reaches almost every other individ-ual from the non-Roma population whereas activismin Novo mestomdashunderstandablymdashreaches only a fewIn testing the effectiveness of inclusive organizing atthe ground level Novo mesto is therefore a suitablecounterpart to Murska Sobota

Factors Common to the Test Locations

While the towns differ on the crucial covariates theymatch very closely on the state regional and local char-acteristics that most strongly influence human rightsgenerally and Roma rights specifically ranging fromaverage income to bigotry in the media National andEU laws regarding Roma rights are self-evidently thesame in Novo mesto as in Murska Sobota as are statelevel factors that may have historically shaped the rela-tionship between Roma and non-Roma The match ofMurska Sobota and Cakovec at the state level in turnis very close Most importantly the locations matchon the level of Roma rights protections both beforeaccession and while I collected data for this project(20112012) Before accession police brutality againstthe Roma was observed a few years prior to acces-sion in both states (Slovenia 2000 Croatia 2010)16

Discrimination in education employment difficultiesin housing and societal maltreatment were consis-tently present in both statesmdashbefore accession (Slove-nia 2000-2003 Croatia 2009-2012) right after (Slovenia2005 Croatia 2014) and during data collection (Slove-nia 20112012 Croatia 2011)14 Table 1 lists the factorscommon to the locations

15 This number only captures individuals sufficiently in contact withthe NGO to know its name and purpose It does not include (1)individuals who attended NGO events like the cultural festival with-out knowing that the NGO organized them or (2) individuals whomay have adopted attitudes or behaviors of those in their proximatesocial network who were themselves among the 46 percent directlyreached by the NGO (see Sinclair (2012))16 All facts attributed to the respective US Department of StateHuman Rights Reports

MEASUREMENT

Games

Between the summers of 2011 and 2012 I collected datato construct a measure of discrimination at the groundlevel dimension of everyday relationships betweenRoma and non-Roma in Cakovec Murska Sobota andNovo mesto Common transgressorsmdashpolice officersbureaucrats and teachersmdashare individuals whose anti-Roma sentiment may manifest itself differently basedon the circumstances of their interaction with RomaAs directly recording rights violations that Roma mayexperience in seeking employment or health care indetention or in prison is not possible the measure in-stead targets the root of discrimination It casts a widernet and aims to capture the basic visceral and quoti-dien discrimination upon which such various violationsare frequently based

Data on discrimination were collected through sim-ple games that have been shown to demonstrate riskpreferences and other-regarding behavior The gamesmeasured (1) risk preferences (lottery) (Holt andLaury 2002) (2) altruism (dictator game) (Hoffmanet al 1994) (3) trust in onersquos community members(trust game) (Berg et al 1995) and (4) trustworthi-ness with respect to onersquos community members (trustgame) The chief rationale for using the games wasthe difficulty of measuring individual levels of discrim-ination People may not admit to racist preferences ifasked directly in fact participants in both Croatia andSlovenia often agreed to participate on the groundsthat no personal questions would be asked but did notmind the game setup17

In the trust game subjects were assigned to play therole of a sender or a receiver and were then randomlyand anonymously paired with a partner Both senderand receiver began the game with an identical endow-ment The sender chose how much of the endowmentto share with the receiver knowing that the amountsent would be doubled and that the receiver wouldhave the chance to return to the sender a portion ofhis total amount The doubled amount was then givento the receiver The receiver decided how much of histotal amountmdashthat is his initial endowment plus thedoubled amountmdashto send back The amount sent wasused as a measure of trust and the amount returned asa measure of trustworthiness

Subjects were randomly paired to play the trust gamewith an anonymous randomly chosen Roma or non-Roma partner from their community The treatmentwas delivered in person subjects were told that theirpartner was randomly chosen and anonymous eitherRoma or non-Roma and that there would be no directinteraction with the partner Whether the participant

17 While the trust game involves an actual interaction between twopeople it is nonetheless a simulation of a real-life situation Whetherthe game captures discriminatory intent or merely prejudice may beup for debate but the issue is not dispositive Both experimental (Do-vidio 2004) and longitudinal (Wagner 2008) analyses demonstratethat the two are closely connected and that prejudice is causallylinked to discriminatory behavior

537))amp)())amp-13$))amp$+()$$)( ))))((+)


TABLE 1 Factors Common to the Test Locations

Factor Cakovec Croatia Murska Sobota Slovenia Novo mesto Slovenia

Regional capitalsradic radic radic

Largest Romani population inregion

radic radic radic

Roma as percent of townpopulationa

359 227 12-28

Distance from SloveneCroatianborder (miles)

10 10 10

Between 20000 and 30000inhabitants

radic radic radic

Roma in town over 200 yearsb radic radic radic

Vast majority of Roma in isolatedsettlementc

radic radic radic

Roma represented in localgovernmentsd

radic radic radic

Romani NGOs presentradic radic radic

Primary school curricula omitRomae

radic radic radic

Roma as percent of statepopulationf

018 019 019

Yugoslavian inclusive ethnictolerance policiesg

radic radic radic

Yugoslavian rights restrictionsradic radic radic

Parliamentary democracyradic radic radic

2011 average regional monthlyincome in US $h

1169 1927 2176

EU comparative price level indexscorei

73 85 85

Rise in intolerance duringtransitionj

radic radic radic

All core human rights treatiesratifiedk

radic radic radic

Membership in comparablenumber of INGOsl

radic radic radic

1 year before EU bid Romaintegrate poorly in schoolsm

radic radic radic

1 year before EU bid Romaprogram adoptedn

radic radic radic

Roma program focuses oneducationn

radic radic radic

Roma not recognized as a nationradic radic radic

Roma recognized as an ethnicminorityo

radic radic radic

Percent Roma employed 1 yearbefore EU bidp

171 17 17

Press bigoted when referring toRomaq

radic radic radic

No extrajudicial killings 4 yearsbefore accessionr

radic radic radic

No political imprisonments 4 yearsbefore accessionr

radic radic radic

No arbitrary arrests 4 years beforeaccessionr

radic radic radic

Police brutality against the Roma 3years before accessionr

radic radic radic

Preaccession incomplete Romapolitical representation at thestate levelr

radic radic radic

Preaccession Roma discriminationand segregation in educationr

radic radic radic

Preaccession Roma discriminationin employmentr

radic radic radic

538))amp)())amp-13$))amp$+()$$)( ))))((+)

Ana Bracic


TABLE 1 Continued


Preaccession Roma societaldiscriminationr

radic radic radic

no extrajudicial killings or politicalimprisonments in 20112012s

radic radic radic

No arbitrary arrests or policebrutality against the Roma in20112012s

radic radic radic

Incomplete Roma politicalrepresentation at the state levelin 20112012s

radic radic radic

Roma discrimination andsegregation in education in20112012s

radic radic radic

Employment and societal Romadiscrimination in 20112012s

radic radic radic

Comprehensive anti-discriminationlegislation adopted (year)t

2008 2007 2007

Regions border one anotherradic radic

Hungarian minority in regionu radic radic

Hungarian rule duringAustro-Hungarian empirev

radic radic

aMesojedec (2012) Republika Hrvatska - Drzavni Zavod za Statistiku (2001) Statisticni urad Republike Slovenije (2002)bHorvat-Muc (2011a) Klopcic (2012) Vugrincic and Siladi (2008) cAjdic (2008) Slezak (2009) dBajric (2012) Horvat-Muc(2011b) Tudija (2012) eKarba (2010) Vican and Litre (2006) f Republika Hrvatska - Drzavni Zavod za Statistiku (2001)Statisticni urad Republike Slovenije (2002) gKenrick (2001) hKolakovic (2013) Statisticni urad Republike Slovenije (2011)i Eurostat (2015) EU average is 100 j Barany (2002) kSee entries on Slovenia and Croatia at The University of MinnesotaHuman Rights Center (2011) lSee Hafner-Burton and Tsutsui 2007 mKlopcic (2007) Tancer (2003) nKlopcic (2007)oPolzer-Srienz (2003) pPokos (2005) Tancer (1994) qErjavec et al (2000) Kanizaj (2004) rSee US State DepartmentHuman Rights Reports Slovenia (2000-2003) and Croatia (2009-2012) sSee US State Department Human Rights ReportsSlovenia (2011-2012) and Croatia (2011) tEuropean Commission (2009) RS Urad za Narodnosti (2015) uVratusa (2008)vSiftar (1970)

received a treatment or a control condition was deter-mined beforehand with a coin toss Subjectsrsquo decisionswere confidential and made in private

The trust game played for a monetary sum is ex-ceptionally suitable as a measure of discrimination be-cause negative stereotypes and general dislike of theRoma are largely based on distrust regarding money18

Roma are universally stereotyped as cheaters andthieves (Scicluna 2007 Siftar 1989) Historically theyhave been wrongfully accused of stealing even children(Hancock 2002) today people still say ldquoI feel gyppedrdquoand frequently think nothing of it Exploiting this senti-ment the trust game elicits a gut response from a non-Roma to not send money to a Roma partner becausethat partner would not send anything (or enough) backAccordingly the primary quantity of interest was thedifference between the trust that non-Roma exhibitedin interactions with Roma and that which they exhib-ited in interactions with non-Roma

This method was well received in Romani communi-ties from Slovenia Croatia and Romania19 Numerous

18 The expected total payoff from the three games for one subjectwas approximately 60 of a daily wage in Slovenia approximatelythe equivalent of euro16 and in Croatia of 106 kuna19 Conversations with staff at Romani Union-Zveza Romov (MurskaSobota) and Romani CRISS (Bucharest) interview with Zeljko Ba-log conversation with interested Roma in the Sitnice settlement

activists were disinclined to trust answers given by non-Roma in response to survey questions on discrimina-tion but enthusiastically endorsed the trust game Theyfound that it appropriately captures the sentiment thatmotivates discrimination they experience in their dailyinteractions with non-Roma and appreciated its abil-ity to elicit a discriminatory response without directlyasking a non-Roma whether or not she discriminates

The lottery and the dictator game were included tocontrol for characteristics that likely influence subjectsrsquoresponses in the trust game and might confound theresults A highly altruistic person may for exampleoffer more money in the trust game but not necessarilybecause she trusts her potential partner (Cox 2004)Likewise a risk-loving person might offer a large sumbecause of the thrill not trust (Eckel and Wilson 2004Schechter 2007)

Non-Roma subjects participated individually intheir homes20 Roma subjects participated individuallyin several central locations including an NGO common

20 Participation protocols differed from the standard delivery of thetrust game in order to allow individual and private decision-makingI avoided playing the trust game in a group to protect subjects fromlikely contention The game was played sequentially first all sendersmade decisions the receivers followed Subjects played the gameswith cash They sealed the sum they sent to their partner in an enve-lope wrote their unique and confidential identifier on the envelopeand deposited the envelope in a closed box with a slot They were

539))amp)())amp-13$))amp$+()$$)( ))))((+)


room a kindergarten and several homes Participationgenerally lasted between 10 and 20 minutes

Participation began with reading and signing theconsent form continued with the three games andconcluded with a short exit survey The survey askedgeneral questions concerning the age gender educa-tion income bracket the nationality and ethnicity ofthe participant as well as two questions on the EUaccession process Participants were asked if they hadheard about this study or the games before if theyanswered yes they were asked to specify when and inwhat context If they heard about the games from aperson who had participated they were removed fromthe sample21

Stratified Random Sampling

The random population sample consisted of 202 sub-jects from each town22 Simple random sampling wasused to draw participants from the two strata thenon-Roma general population and the Roma generalpopulation The Roma community was oversampledbecause it is substantially smaller All streets in thetown or the Romani settlement were numbered andre-ordered based on a random number sample Par-ticipants were recruited from the chosen streetsmdashoneperson from each house with a systematic iterationbetween genders Any individual over the age of 18 waseligible to participate in the study The response ratewas approximately 60 percent23 In total 606 peoplefrom the general population participated in the games202 from each town24

RESULTS

The results can be summarized as follows First non-Roma from the EU candidate town discriminatedagainst the Roma while non-Roma from the EU mem-ber town just across the border did not Second non-Roma from the town with inclusive organizing did notdiscriminate against the Roma but non-Roma fromthe town with noninclusive organizing did

As this article discusses levels of discriminationagainst the Roma in Slovenia and Croatia I focus onthe behavior of the subjects who were senders in thetrust gamemdashin total 303 randomly chosen individuals

alone when making decisions and knew that the person handling thecontents of that box would have no way of identifying them See theOnline Appendix for protocol details21 Two subjects were excluded on this basis22 For a moderate effect size (around 025) and a power of 08 Irequired 50 subjects per treatment to find a statistically significantdifference at the 5 percent level This demanded 200 subjects pertown 100 non-Roma senders of which half were paired with 50 non-Roma receivers and half with 50 Roma receivers See Cohen (1988)23 Research teams attempted recruiting at a house on a selectedstreet three times The Statistical Office of the Republic of Sloveniareports response rates that range from 60 to 80 percent rates varywith the topic of the survey (Lah et al 2011 Remec 2005) I thankMatej Divjak from the Office for his consultation regarding responserates24 Two people decided to withdraw from the study shortly after par-ticipating one from Murska Sobota and one from Cakovec

The main quantity of interestmdashthe dependent variablein the modelmdashis the amount participants sent to theirrespective partners in the trust game As the currenciesin which participants were playing were not the samethe relevant variables are coded as proportions of totalendowment25

Figure 2 presents the average proportion of totaltrust game endowment that senders in each town sentto Roma and non-Roma partners The figure shows anegligible average treatment effect in the case of theEU member (Murska Sobota) where senders on aver-age sent 57 percent of their endowment to non-Romaand 58 percent to Roma partners a statistically insignif-icant difference-in-means The average treatment ef-fect in the EU candidate (Cakovec) on the other handis statistically significant (p lt 005) Senders there sent75 percent of endowment to non-Roma and 65 percentto Roma partners

Figure 2 also shows that senders from the non-inclusive NGO town (Novo mesto) on average sent 73percent of endowment to non-Roma and 59 percent toRoma partners (the average treatment effect is statisti-cally significant at p lt 005) In contrast senders fromthe inclusive NGO town (Murska Sobota) on averagesent almost the same proportion of their endowmentto Roma and non-Roma partners

Regression analysis provides further insight The ap-propriate specification for a model in which the de-pendent variable is a proportion is a generalized lin-ear model with the binomial variance and the logitlink function (McDowell and Cox 2004 Papke andWooldridge 1996)26 To estimate the treatment ef-fect the model includes interaction terms between thecontrol function and the treatment variable (Roma)(Morton and Williams 2010) Since partner ethnicityin the trust game was randomly assigned within eachtown the observations were accordingly weighted27

Table 2 presents results from the model that in-cludes the lottery chosen the proportion of the endow-ment sent to the family in need in the dictator gametown dummies and the main population controlsmdashagegroup gender education level and income bracketmdashas well as the interaction terms between these co-variates and the treatment (Roma) as independentvariables28

25 Table 4 in the Online Appendix summarizes key variables forsenders26 The results are substantively unchanged using an ordinary leastsquared or a tobit regression Results available upon request27 As each town was treated as a block each treated observation(sender partnered with a Roma receiver) was weighted by the inverseof the proportion of subjects in its block (town) who were assignedto the treatment condition and each control subject was weightedby the inverse of the proportion of subjects in its block who wereassigned to the control condition (Gerber and Green 2012)28 Numerous iterations of the model were run with various numbersand combinations of control variables The findings are robust toall additions Some iterations controlled for authorrsquos presence onthe research team This control variable is consistently insignificantMoreover the results presented here are from a pooled analysisFindings from analyses where the relevant towns are compared inpairs are substantively unchanged All results are reported in theOnline Appendix (Tables 5 and 6)

540))amp)())amp-13$))amp$+()$$)( ))))((+)


FIGURE 2 The Average Proportion of Total Endowment Sent to Partner in the Trust Game byTown and Partnerrsquos Ethnic Identity

Notes The difference-in-means or average treatment effect is statistically significant where marked The average treatment effect in thecase of the EU member (Murska Sobota) is negligible while the average treatment effect in the EU candidate (Cakovec) is statisticallysignificant (p lt 005) The average treatment effect is also statistically significant at p lt 005 in the noninclusive NGO town (Novomesto) but is not statistically significant in the inclusive NGO town (Murska Sobota)

TABLE 2 Effect of Game Partner Ethnicity(Roma or non-Roma) in Croatia and Sloveniaon the Amount Sent in the Trust GameGeneralized Linear Model with Weighted Data

Independent Variables Coefficients (SEs)

Roma dummy minus 0301(0993)

Cakovec 0603 lowastlowast

(0260)Cakovec x Roma minus 1122 lowastlowastlowast

(0413)Novo mesto 0592 lowastlowast

(0289)Novo mesto x Roma minus 1135 lowastlowastlowast

(0400)Number of observations 279

lowast p lt 01 lowastlowast p lt 005 lowastlowastlowast p lt 001 standard errors in paren-thesesTable listing the coefficients and standard errors of control co-variates is in the Online Appendix (Table 5)

Conditional marginal effects demonstrate the effecton the predicted mean proportion of the total endow-ment sent in the trust game as partner ethnicity changesfrom non-Roma to Roma29 The marginal effect for

29 The conditional marginal effects were calculated using the marginscommand in Stata with control covariates held at their means

Murska Sobota is not statistically significant essen-tially there is no evidence that a randomly chosensender from Murska Sobota would send any less or anymore to a Roma partner than she would to a non-Romapartner In Cakovec on the other hand the change inthe proportion of the endowment sent is statisticallysignificant (p lt 001) A randomly chosen sender fromCakovec (EU candidate) would send about 30 percentless to a Roma partner than to a non-Roma partnerThe marginal effect is also statistically significant in thecase of Novo mesto (non-inclusive NGO) There tooa randomly chosen sender would on average send 30percent less to a Roma partner (p lt 001)30

Curiously senders from Cakovec and Novo mesto onaverage sent more to their partners than did sendersfrom Murska Sobota31 They also sent substantiallymore to non-Roma receivers clearly exhibiting in-group favoritism (Hewstone et al 2002) In-group

30 While the match between Murska Sobota and Cakovec is supe-rior some information can nonetheless be gleaned from comparingCakovec and Novo mesto Comparing the two suggests that neitherthe accession process nor eight-year-long EU membership alone pro-duce a discrimination-free environment on the ground31 Similarly in the dictator game senders from Murska Sobota sentsignificantly less to the anonymous local family in need than didsenders from Cakovec and Novo mesto (on average donating 70 87(p lt 001) and 83 (p lt 005) percent of their endowment respec-tively) This discrepancy suggests a higher average level of socialcapital in Cakovec and Novo mesto

541))amp)())amp-13$))amp$+()$$)( ))))((+)


favoritism is a manifestation of discrimination gener-ally seen as a milder form of bias when not connectedto outright out-group derogation (Brewer 1999) Inthe case of study participants from Novo mesto andto a lesser extent from Cakovec however out-groupderogation was present as well

Unexpectedly participants offered unsolicited state-ments regarding the Roma In most cases the state-ments were given after or during the exit survey a fewparticipants commented on the Roma during the trustgame One participant tellingly mistook a researchteam for a pair of Roma going door to door asking formoney We received comments from senders pairedwith Roma and non-Roma alike whether or not theywere assigned the treatment had no effect on the sub-stance of the comments Thirty-six percent of sendersfrom Novo mesto made comments one percent waspositive 11 percent were neutral and 24 percent werenegative The most remarkable comment included ahalf-joking threat that ldquowe will send a couple of bus-loads of our Roma to your hometown and yoursquoll seewhat itrsquos likerdquo Unsolicited statements were not nearlyas common in the other two towns In Murska Sobotaone subject remarked upon receiving the treatmentthat she did not see why it would matter that her partnerwas a Roma or a non-Roma In Cakovec we receivedthree negative comments the most remarkable ldquoThehell will he sendrdquo referred to the possibility that aRomani partner may return some of the endowmentto his partner in the trust game

A concern arises in light of these findings Sendersfrom Cakovec and Novo mesto might have sent lessto Roma partners because the Roma in Cakovec andNovo mesto are actually less likely to reciprocate insuch a context compared to the Roma in Murska Sob-ota A comparison of responses by Romani receivershowever indicates that this is not the case There isno statistically significant difference between what theRoma from all three locations returned to their part-ners as a proportion of the total pot The Roma playedconsistently

These findings have three implications First the EUaccession process does not necessarily lower discrimi-nation such that it will be lower in an accessing statethan in an EU member Second discrimination can beremarkably low even absent in an EU member stateThird and finally inclusive organizing that aims to im-prove Romanon-Roma relations helps reduce groundlevel discrimination

A few caveats are in order Crucially the findingsdo not assess the absolute ground level effect of theEU accession process The sample does not includeobservations from a location that has not yet been af-fected by the accession process or one that is entirelyoutside the purview of the EU Without such a base-line an evaluation of the efficacy of the process as awhole is nearly impossible to make Therefore whilethe findings demonstrate that the EU accession processdoes not necessarily lower ground level discriminationbelow that in an EU member they do not address theclaim that the EU accession process as a whole is inef-

fective in reducing ground level discrimination againstthe Roma

Next while the results demonstrate that discrimina-tion can be remarkably low in an EU member statethis is not universally the case It is abundantly clearthat many Roma who live in EU member states East-ern and Western do not consistently enjoy enviablerights and equal treatment At the state level Frenchdeportations of immigrant Romani populations areself-evidently intolerant (Erlanger 2010) as are Italyrsquosefforts to destroy makeshift dwellings in settlementsoutside Milan (ERRC et al 2008) Hateful acts inspiredby personal bigotry range from creating fictional ad-ministrative barriers that impede obtaining social ben-efits (State Department 2009a) to participating in im-promptu (State Department 2009b) or planned (StateDepartment 2007) demonstrations that usually takeplace in Romani settlements and frequently involveMolotov cocktails (Tkach 2010) To this collection ofspecific events this study adds statistically significantevidence that ground-level discrimination can still ex-ist within EU member states and that it varies withincountries likely at the town level

Finally the results do not ensure that we would ob-serve the same effect of inclusive Roma organizingthroughout the EU space or beyond it or that wewould observe the same if inclusive organizing wererandomly assigned They also do not speak conclu-sively to the degree of a direct link between the EUand Roma organizing Roma issues are a high prioritywithin the EU space (European Commission 2015) Inthe context of violence against women the combina-tion of local autonomous feminist activism and regionalor international normative mechanisms is remarkablyeffective at spurring change (Htun and Weldon 2012)Drawing on that and considering the depth of EUinvolvement in the transnational Roma advocacy net-work possible EU effects cannot be ruled out Ac-tivists and experts from across Slovenia however aredivided on the depth and importance of the EU inlocal organizing (Tahirovic 2011) Some claim that theEU plays a strong role in motivating activists (Balazek2012 Klopcic 2012) while others maintain that its roleis negligible (Horvat-Muc 2011a Roser 2011) Whetheror not the EU in any form is the primary driver be-hind the observed lack of discrimination is at presentuncertain and more work is required to clarify thatrelationship32 In the meantime the absence of animusin Murska Sobota is striking

This finding speaks to a facet in the literature on thepromotion of human rights norms (Brysk 1993 Keckand Sikkink 1998 Khagram et al 2002 Sikkink 1993)that focuses on softening the public mood (Amentaet al 2010) The result thus contributes to scholarshipthat establishes the necessity of a local presence innorms promotion (Htun and Weldon 2012 Murdie andDavis 2012) and offers a clarification not just any typeof organizing will do That does not mean that the only

32 Neither the NGOs nor the Slovene Office for National Minori-ties were forthcoming with records of financial support given to theNGOs in Murska Sobota and Novo mesto

542))amp)())amp-13$))amp$+()$$)( ))))((+)


type of local action that helps improve human rights isinclusive organizing rather it means that local actionought to be tailored to its target audience In targetingindividuals from Murska Sobota inclusive organizingappears to be appropriate

Focusing on the behavior of individuals this findingalso bears upon literature in psychology on contactprejudice and discrimination Crucially and unlikethose in many psychological studies the subjects inthis study were not treated with direct contact duringthe course of participation The study instead offersan assessment of long-term intergroup contactmdashas itdevelops naturally on the groundmdashas a measure in-tended to reduce discrimination and so provides a linkbetween scores of convincing results from intergroupinteractions in laboratory settings and theories of hu-man rights norm promotion

CONCLUSION AND FUTURE DIRECTIONS

The main contribution of this article lies in its origi-nal microlevel data about human rights behaviors andits exploration of how that data link to broader hu-man rights institutions This study finds no support forthe idea that a macrolevel combination of incentivesand norms in the form of EU accession temporar-ily reduces individual-level discrimination against theRoma It does however provide support for the hy-pothesis that ground level organizing geared towardsimproving Romanon-Roma relations a microlevelprocess helps reduce discrimination

Questions remain Is the effect of ground level or-ganizing observed in Murska Sobota generalizableWould we observe a similar effect if the treatmentof local level organizing or a simulation thereofwere randomly assigned Would non-Roma froma town entirely outside the purview of the EUdiscriminate against the Roma Would non-Romafrom such a town exhibit nondiscriminatory atti-tudes if the town had strong local level organiz-ing like that in Murska Sobota Do these findingsgeneralize to other groups that face discriminationelsewhere

These unresolved questions urge us to dig deeperMuch research remains to be done on the effects ofmacro- and microlevel processes on human rights out-comes Recent focus on police brutality in the US andin Brazil demands research in which individual abusesindividual perpetrators and institutions intended tocurb them are systematically examined As humanrights monitors recognize more types of violations asabuse scholars must examine them as well We mightexplore if and how human rights institutions lead toimproved prison conditions or how they reduce thenumber of hate crimes We might ask do individu-als respond differently to nascent human rights normsbased on how many people support them Are localadvocacy groups more successful at persuasion thanestablished international NGOs In answering someof these questions gathering original data and takingadvantage of natural or quasi experiments might help

scholars overcome the natural scarcity of human rightsdata33

With respect to the Roma this study helps illuminatethe relationships between people of Europersquos largestethnic minority and of two European nationalitiesThe findings suggest that Roma inclusion strategiesought to include those that aim to promote friendlycontact between Roma and non-Roma Most currentgroups focus justifiably on improving Romani accessto education employment healthcare and eliminatingvarious other injustices there are very few organiza-tions that focus on intergroup contact Discriminationat the level of the individual however is at the heart ofmany violations barriers to employment or access tosocial services can often be traced to bigoted individu-als acting with impunity Intergroup contact offers oneavenue towards eliminating individual bigotry coupledwith other interventions contact strategies could leadto sustainable improvement Developing and fundingsuch strategies is therefore an obvious next step inRoma inclusion efforts

REFERENCES

Ajdic Karmen 2008 Diplomsko delo Romska Naselja kot PrimerProstorske Segregacije v Sloveniji Ljubljana Univerza v LjubljaniFakulteta za druzbene vede

Allport Gordon Willard 1954 The Nature of Prejudice CambridgeMA Addison-Wesley

Amenta Edwin Neal Caren Elizabeth Chiarello and Yang Su 2010ldquoThe Political Consequences of Social Movementsrdquo Annual Re-view of Sociology 36 287ndash307

Anonymous 2012 Anonymous government official from the Officefor National Minorities in Croatia Interview by author Voicerecording Zagreb

Anonymous 2013 Anonymous official from the Office of the Gov-ernment of the Czech Republic Interview by author NotesPrague

Anonymous 2014 Anonymous European Commission official In-terview with author Notes Brussels

Bajric Bajro 2012 Interview by author Voice recording CakovecBalazek Dusica 2012 Interview by author Voice recording Novo

mestoBaldus David C George Woodworth David Zuckerman

Neil Alan Weiner and Barbara Broffitt 1998 ldquoRacial Discrimina-tion and the Death Penalty in the Post-Furman Era An Empiricaland Legal Overview with Recent Findings from PhiladelphiardquoCornell Law Review 83 1638ndash770

Baluh Stane 2012 Interview by author Voice recording LjubljanaBarany Zoltan 2002 The East European Gypsies Regime Change

Marginality and Ethnopolitics Cambridge UK Cambridge Uni-versity Press

Berg Joyce John Dickhaut and Kevin McCabe 1995 ldquoTrust Reci-procity and Social Historyrdquo Games and Economic Behavior 10122ndash42

Blair Irene V Bernadette Park and Jonathan Bachelor 2003 ldquoUn-derstanding Intergroup Anxiety Are Some People More Anxiousthan Othersrdquo Group Processes amp Intergroup Relations 6 (2) 151ndash69

Blascovich Jim Wendy Berry Mendes Sarah B HunterBrian Lickel and Neneh Kowai-Bell 2001 ldquoPerceiver Threatin Social Interactions With Stigmatized Othersrdquo Journal ofPersonality and Social Psychology 80 253ndash67

Boisjoly Johanne Greg J Duncan Michael Kremer Dan M Levyand Jacque Eccles 2006 ldquoEmpathy or Antipathy the Impact ofDiversityrdquo The American Economic Review 96 1890ndash905

33 See Dunning (2012) for a an extensive overview of such method-ology in political science and related disciplines

543))amp)())amp-13$))amp$+()$$)( ))))((+)


Branton Regina Gavin Dillingham Johanna Dunaway andBeth Miller 2007 ldquoAnglo Voting on Nativist Ballot Initiatives ThePartisan Impact of Spatial Proximity to the US-Mexico BorderrdquoSocial Science Quarterly 88 882ndash97

Brewer Marilynn B 1999 ldquoThe Psychology of Prejudice IngroupLove or Outgroup Haterdquo Journal of Social Issues 55 429ndash44

Brewer Marilynn B 2007 ldquoThe Social Psychology of IntergroupRelations Social Categorization Ingroup Bias and OutgroupPrejudicerdquo In Social Psychology Handbook of Basic Principleseds Arie W Kruglanski and Tory E Higgins New York GuilfordPress 695ndash715

Brown Rupert and Miles Hewstone 2005 ldquoAn Integrative Theoryof Intergroup Contactrdquo Advances in Experimental Social Psychol-ogy 37 255ndash343

Brysk Alison 1993 ldquoFrom Above and Below Social Movementsthe International System and Human Rights in Argentinardquo Com-parative Political Studies 26 259ndash85

Cahn Claude 2002 Roma Rights Race Justice and Strategies forEquality New York NY International Debate Education Asso-ciation

Cohen Jacob 1988 Statistical Power Analysis for Behavioral Sci-ences Hillsdale NJ Lawrence Erlbaum Associates

Cook Stuart W 1971 The Effect of Unintended Interracial ContactUpon Racial Interaction and Attitude Change Final Report Wash-ington DC Office of Education (DHEW) Bureau of Research

Courthiade Marcel 2003 ldquoThe Ganetic City of Kannauj OriginalCradle-town of the Rromani Peoplerdquo In Evropa Slovenija inRomi Zbornik Referatov na Mednarodni Konferenci v Ljubljani15 februarja 2002 eds Vera Klopcic and Miroslav PolzerLjubljana Slovenia Institut za narodnostna vprasanja145ndash68

Cox James C 2004 ldquoHow to Identify Trust and Reciprocityrdquo Gamesand Economic Behavior 46 260ndash81

Daniel Stanko 2010 Interview by author Notes BudapestDavies Kristin Linda R Tropp Arthur Aron Thomas F Pettigrew

and Stephen C Wright 2011 ldquoCross-Group Friendships and In-tergroup Attitudes A Meta-Analytic Reviewrdquo Personality andSocial Psychology Review 15 (4) 332ndash51

Davis David R Amanda Murdie and Coty Garnett Steinmetz2012 ldquoMakers and Shapers Human Rights INGOs and PublicOpinionrdquo Human Rights Quarterly 34 199ndash224

De Witte Bruno 2003 ldquoThe Impact of Enlargement on the Constitu-tion of the European Unionrdquo In The Enlargement of the EuropeanUnion ed Marise Cremona New York Oxford University Press209ndash52

Deutsch Morton and Mary Evans Collins 1951 Interracial Hous-ing Minneapolis MN University of Minnesota Jones Press Inc

Dimitrova Antoaneta and Dimiter Toshkov 2009 ldquoPost-accessionCompliance between Administrative Co-ordination and Polit-ical Bargainingrdquo In European Integration Online Papers edsFrank Schimmelfennig and Florian Trauner 13 1ndash18

Djuric Rajko 2007 Povijest Roma Prije i Poslije Auschwitza Za-greb Croatia Prosvjeta

Dovidio John F Samuel L Gaertner Jason A NierKerry Kawakami and Gordon Hodson 2004 ldquoContemporaryRacial Bias When Good People do Bad Thingsrdquo In The SocialPsychology of Good and Evil ed Arthur G Miller New YorkGuilford Press 141ndash67

Dunning Thad 2012 Natural Experiments in the Social Sciences ADesign-Based Approach Cambridge UK Cambridge UniversityPress

Eberhardt Jennifer L Paul G Davies Valerie J Purdie-Vaughnsand Sheri Lynn Johnson 2006 ldquoLooking Deathworthy PerceivedStereotypicality of Black Defendants Predicts Capital-SentencingOutcomesrdquo Psychological Science 17 383ndash6

Eckel Catherine C and Rick K Wilson 2004 ldquoIs Trust a RiskyDecisionrdquo Journal of Economic Behavior and Organization 55447ndash65

Erjavec Karmen Sandra B Hrvatin and Barbara Kelbl 2000 Mi oRomih Diskriminatorski Diskurz v Medijih v Sloveniji LjubljanaSlovenia Open Society Institute

Erlanger Steven 2010 ldquoExpulsion of Roma Raises Questions inFrancerdquo The New York Times August 19 2010

European Commission 2009 Croatia 2009 Progress Report Euro-pean Commission

European Commission 2010 Communication from the Commissionto the European Parliament and the Council Enlargement Strategyand Main Challenges 2010ndash2011

European Commission 2011 ldquoPress Release EU Closes Ac-cession Negotiations with Croatiardquo httpeuropaeurapidpressReleasesActiondoreference=IP11824

European Commission 2012a ldquoTackling Discrimination EUand Roma Projects and Fundingrdquo httpeceuropaeujusticediscriminationromaproject-fundingindex enhtm

European Commission 2012b ldquoRoma Summitsrdquo httpeceuropaeujusticediscriminationromaroma-summitsindex enhtm

European Commission 2014 ldquoEnlargement Strategy andProgress Reportsrdquo httpeceuropaeuenlargementcountriesstrategy-and-progress-reportindex enhtm

European Commission 2015 ldquo10 Priorities Justice andFundamental Rightsrdquo httpeceuropaeuprioritiesjustice-fundamental-rightsindex_enhtm

European Roma Rights Centre 1997 Profession Prisoner Roma inDetention in Bulgaria Budapest European Roma Rights Centre

European Roma Rights Centre 2001 ldquoBritish Officials ContinuePolicy of Stopping Roma at Czech Airport Czech Roma andERRC sue UK Governmentrdquo httpwwwerrcorgcikkphpcikk=1289

European Roma Rights Centre 2004 The Situation of Roma inan Enlarged European Union Budapest European Roma RightsCentre

European Roma Rights Centre et al 2008 Security a la ItalianaFingerprinting Extreme Violence and Harassment of Roma in ItalyBudapest Hungary Fo-Szer Bt

Eurostat 2015 ldquoEurostat Statistics Explained Consumer Prices- Inflation And Comparative Price Levelsrdquo httpeceuropaeueurostatstatistics-explainedindexphptitle=Consumer prices -inflation and comparative price levelsampoldid=223383

Fariss Christopher 2014 ldquoRespect for Human Rights has ImprovedOver Time Modeling the Changing Standard of AccountabilityrdquoAmerican Political Science Review 108 297ndash318

Finnemore Martha and Kathryn Sikkink 1998 ldquoInternationalNorm Dynamics and Political Changerdquo International Organiza-tion 52 (4) 887ndash917

Gelazis Nida M 2004 ldquoThe European Union and the StatelessnessProblem in the Baltic Statesrdquo European Journal of Migration andLaw 6 225ndash42

Gerber Alan S and Donald P Green 2012 Field Experiments De-sign Analysis and Interpretation New York W W Norton ampCompany

Green Donald P and Janelle S Wong 2009 ldquoTolerance andthe Contact Hypothesis A Field Experimentrdquo In The Politi-cal Psychology of Democratic Citizenship eds Eugene BorgidaChristopher M Federico and John L Sullivan New York NYOxford University Press

Hafner-Burton Emilie and Kiyoteru Tsutsui 2007 ldquoReplicationdata for Human Rights Practices in a Globalizing World TheParadox of Empty Promisesrdquo httphdlhandlenet1902110487UNF3EMcZWT0FrM1S+Zv8GzR1ng== Emilie M Hafner-Burton [Distributor] V1 [Version]

Hafner-Burton Emilie M 2005 ldquoTrading Human Rights How Pref-erential Trade Agreements Influence Government RepressionrdquoInternational Organization 59 (3) 593ndash629

Hafner-Burton Emilie M 2008 ldquoSticks and Stones Naming andShaming the Human Rights Enforcement Problemrdquo InternationalOrganization 62 (4) 689ndash716

Hancock Ian F 2002 We are the Romani People Hertfordshire Uni-versity of Hertfordshire Press

Hendrix Cullen S and Wendy Wong 2012 ldquoWhen Is the Pen TrulyMighty Regime Type and the Efficacy of Naming and Shamingin Curbing Human Rights Abusesrdquo British Journal of PoliticalScience 43 651ndash72

Hewstone Miles Mark Rubin and Hazel Willis 2002 ldquoIntergroupBiasrdquo Annual Review of Psychology 53 575ndash604

Ho Daniel E Kosuke Imai Gary King and Elizabeth A Stuart2007a ldquoMatching as Nonparametric Preprocessing for ReducingModel Dependence in Parametric Causal Inferencerdquo PoliticalAnalysis 15 (3) 199ndash236

Ho Daniel E Kosuke Imai Gary King and Elizabeth A Stu-art 2007b ldquoMatchit Matching as Nonparametric Preprocessing

544))amp)())amp-13$))amp$+()$$)( ))))((+)


for Parametric Causal Inferencerdquo Journal of StatisticalSoftware

Ho Daniel E Kosuke Imai Gary King and Elizabeth A Stu-art 2011 ldquoMatchIt Nonparametric Preprocessing for ParametricCausal Inferencerdquo Journal of Statistical Software 42 1ndash28

Hoffman Elizabeth Kevin McCabe Keith Shachat andVernon Smith 1994 ldquoPreferences Property Rights andAnonymity in Bargaining Gamesrdquo Games and Economic Behav-ior 7 346ndash80

Hojsik Marek 2010 Interview by author Notes BratislavaHollyer James 2010 ldquoConditionality Compliance and Domestic

Interests State Capture and EU Accession Policyrdquo The Review ofInternational Organizations 5 (4) 387ndash431

Holt Charles A and Susan K Laury 2002 ldquoRisk Aversion andIncentive Effectsrdquo American Economic Review 92 1644ndash55

Horvat-Muc Jozek 2010 20 let Romani Union Murska Sobota 1990ndash2010 RD Romani Union

Horvat-Muc Jozek 2011b Romska Skupnost v Sloveniji Zgodov-ina in Kultura Romov Murska Sobota Slovenija Zveza RomovSlovenije Romani Union

Horvat-Muc Jozek 2011a Interview by author Voice recordingMurska Sobota

Htun Mala and S Laurel Weldon 2012 ldquoThe Civic Origins of Pro-gressive Policy Change Combating Violence against Women inGlobal Perspective 1975ndash2005rdquo American Political Science Re-view 103 548ndash69

Hughes Krista 2014 ldquoUS cuts swaziland from trade benefits re-instates madagascarrdquo httpwwwreuterscomarticle20140626us-usa-trade-africa-idUSKBN0F12T920140626

Jackman Mary R and Marie Crane 1986 ldquoSome of my best friendsare black Interracial friendship and whitesrsquo racial attitudesrdquoPublic Opinion Quarterly 50 459ndash86

Jovanovic Djordje 2010 Interview by author Notes BudapestKanizaj Igor 2004 ldquoPredstavljenost nacionalnih manjina u

Hrvatskim dnevnim novinama komparativni pregled 2001-2003rdquoPoliticka Misao 41 (2) 30ndash46

Karba Pavla 2010 Program Osnovna Sola Drzavljanska in Do-movinska Vzgoja ter Etika Ljubljana Ministrstvo za solstvo insport Zavod RS za solstvo

Kasambala Tiseke 2014 ldquoDispatches Swazilandrsquos dismal humanrights record just got worserdquo httpwwwhrworgnews20140725dispatches-swaziland-s-dismal-human-rights-record-just-got-worse

Keck Margaret E and Kathryn Sikkink 1998 Activists BeyondBorders Advocacy Networks in International Politics Ithaca NYCornell University Press

Kelley Judith Green 2004 Ethnic Politics in Europe The Power ofNormas and Incentives Princeton NJ Princeton University Press

Kenrick Donald 2001 ldquoFormer Yugoslavia A Patchwork of Des-tiniesrdquo In Between Past and Future The Roma of Central andEastern Europe ed Will Guy Hertfordshire UK University ofHertfordshire Press 93ndash116

Khagram Sanjeev James V Riker and Kathryn Sikkink 2002ldquoFrom Santiago to Seattle Transnational Advocacy GroupsRestructuring World Politicsrdquo In Restructuring World PoliticsTransnational Social Movements Networks and Norms edsSanjeev Khagram James V Riker and Kathryn Sikkink Min-neapolis MN University of Minnesota Press 3ndash23

Klopcic Vera 2007 Polozaj Romov v Sloveniji Romi in GadzeLjubljana Slovenia Institut za narodnostna vprasanja

Klopcic Vera 2012 Interview by author Voice recordingKolakovic Ines 2013 ldquoEmployment and Wages 2012rdquo Croatian

Bureau of Statistics of the Republic of Croatia 15022013Laar Colette Van Shana Levin Stacey Sinclair and Jim Sidanius

2005 ldquoThe Effect of University Roommate Contact on EthnicAttitudes and Behaviorrdquo Journal of Experimental Social Phsy-chology 41 329ndash45

Lah Lenart Katja Rutar and Irena Svetin 2011 ldquoTrg delaLabour marketrdquo In Statisticne Informacije Rapid Reportsed Statisticni Urad Republike Slovenije Ljubljana SloveniaStatisticni Urad Republike Slovenije number 11 1ndash23

Levitz Philip and Grigore Pop-Eleches 2010 ldquoWhy No Backslid-ing The European Unionrsquos Impact on Democracy and Gover-nance Before and After Accessionrdquo Comparative Political Studies43 457ndash85

Mackie Gerry 1996 ldquoEnding Footbinding and Infibulation A Con-vention Accountrdquo American Sociological Review 61 (6) 999ndash1017

Matras Yaron 2000 ldquoRomani Migrations in the Post-communistEra Their Historical and Political Significancerdquo Cambridge Re-view of International Affairs 13 (2) 32ndash50

McDowell Allen and Nicholas J Cox 2004 ldquoHow do you fit a modelwhen the dependent variable is a proportionrdquo httpwwwstatacomsupportfaqsstatlogithtml

Mendes Wendy Berry Jim Blascovich Brian Lickel andSarah Hunter 2002 ldquoChallenge and Threat During Social In-teractions With White and Black Menrdquo Personality and SocialPsychology Bulletin 28 939ndash52

Mesojedec Silvo 2012 Interview by author Voice recordingMeyer-Sahling Jan-Hinrik 2008 ldquoThe Changing Colours of the Post-

Communist State The Politicization of the Senior Civil Service inHungaryrdquo European Journal of Political Research 47 1ndash33

Miller Norman 2002 ldquoPersonalization and the Promise of ContactTheoryrdquo Journal of Social Issues 58 387ndash410

Ministarstvo unutarnjih poslova Republike Hrvatske 2006 ldquoProjektldquopolicija i gradjanirdquo uspjesno zavrsen httpwwwmuphr2850aspx

Mirwaldt Katja 2010 ldquoContact Conflict and Gepgraphy What Fac-tors Shape Cross-border Citizen Relationsrdquo Political Geography29 434ndash43

Moody James 2001 ldquoRace School Integration and Friendship Seg-regation in Americardquo American Journal of Sociology 107 679ndash716

Morton Rebecca and Kenneth C Williams 2010 Experimental Po-litical Science and the Study of Causality From Nature to the LabNew York Cambridge University Press

Murdie Amanda and David R Davis 2012 ldquoShaming and Blam-ing Using Events Data to Assess the Impact of Human RightsINGOsrdquo International Studies Quarterly 56 1ndash16

Nielsen Richard 2014 ldquoCase Selection via Matchingrdquo SociologicalMethods amp Research 45 (3) 1ndash29

Oravec Laco 2010 Interview by author NotesPage-Gould Elizabeth Rodolfo Mendoza-Denton and

Linda Tropp 2008 ldquoWith a Little Help From My Cross-GroupFriend Reducing Anxiety in Intergroup Contexts ThroughCross-Group Friendshiprdquo Journal of Personality and SocialPsychology 95 (5) 1080ndash94

Paluck Elizabeth L and Donald P Green 2009 ldquoPrejudice Reduc-tion What Works A Review and Assessment of Research andPracticerdquo Annual Review of Psychology 60 339ndash67

Papke Leslie E and Wooldridge Jeffrey 1996 ldquoEconometric Meth-ods for Fractional Response Variables with an Application to401(k) Plan Participation Ratesrdquo Journal of Applied Econometrics11 619ndash32

Pettigrew Thomas F 1998 ldquoIntergroup Contact Theoryrdquo AnnualReview of Psychology 49 65ndash85

Pettigrew Thomas F and Linda R Tropp 2006 ldquoA Meta-analyticTest of Intergroup Contact Theoryrdquo Journal of Personality andSocial Psychology 90 (5) 751ndash83

Phare 2005 Phare Project Fiche 2005 Roma Support Project Croa-tia

Pokos Nenad 2005 ldquoDemografska analiza Roma na temeljustatistickih podatakardquo In Kako Zive Hrvatski Romi edMaja Stambuk Zagreb Croatia Institut drustvenih znanosti IvoPilar 35ndash53

Polzer-Srienz Miriam 2003 ldquoRomi in evropska integracja Primer-java stanja v Sloveniji Avstriji in na Hrvaskemrdquo In EvropaSlovenija in Romi Zbornik Referatov na Mednarodni Konfer-enci v Ljubljani 15 februarja 2002 eds Vera Klopcic andMiroslav Polzer Ljubljana Slovenia Institut za narodnostnavprasanja 93ndash104

Pridham Geoffrey 2008 ldquoThe EUrsquos Political Conditionality andPost-Accession Tendencies Comparisons from Slovakia andLatviardquo Journal of Common Market Studies 46 365ndash87

Remec Matija 2005 ldquoZivljenjska raven Level of livingrdquo InStatisticne Informacije Rapid Reports Statisticni Urad Repub-like Slovenije Ljubljana Slovenia Statisticni Urad RepublikeSlovenije 1ndash10 number 21

Republika Hrvatska - Drzavni Zavod za Statistiku 2001 ldquoPopisstanovnistva kucanstava i stanovardquo httpwwwdzshr

Ripka Stepan 2010 Interview by author Notes

545))amp)())amp-13$))amp$+()$$)( ))))((+)


Rothbart Myron and Oliver P John 1993 ldquoIntergroup Relationsand Stereotype Change A Social-Cognitive Analysis and SomeLongitudinal Findingsrdquo In Prejudice Politics and the Ameri-can Dilemma eds Paul M Sniderman Philip E Tetlock andEdward G Carmines Stanford CA Stanford University Press

Roser Janja 2011 Interview by author NotesRS Urad za Narodnosti 2015 ldquoRepublika Slovenija Urad za Nar-

odnosti Romska skupnost ustavno-pravni polozajrdquo httpwwwungovsisimanjsineromska skupnostustavno pravni polozaj

Sandreli Monika 2012 Interview by author Voice recordingSchechter Laura A 2007 ldquoTraditional Trust Measurement and the

Risk Confound An Experiment in Rural Paraguayrdquo Journal ofEconomic Behavior and Organization 62 272ndash92

Schimmelfennig Frank Stefan Engert and Heiko Knobel 2005 TheImpact of EU Political Conditionality In The Europeanizationof Central and Eastern Europe eds Frank Schimmelfennig andUlrich Sedelmeier Ithaca NY Cornell University Press 29ndash50

Schutz Heidi and Bernd Six 1996 ldquoHow Strong is the Relation-ship between Prejudice and Discrimination A Meta-analytic An-swerrdquo International Journal of Intercultural Relations 20 441ndash62

Scicluna Henry 2007 ldquoAnti-Romani Speech in Europersquos PublicSpace The Mechanism of Hate Speechrdquo Roma Rights QuarterlyJournal of the European Roma Rights Centre 3 47ndash55

Siftar Vanek 1970 Cigani Minulost v Sedanjosti Murska SobotaPomurska Zalozba

Siftar Vanek 1989 ldquoRomi vceraj pojutrisnjemrdquo Znamenje 2122ndash37

Sikkink Kathryn 1993 ldquoHuman Rights Principled Issue-networksand Sovereignty in Latin Americardquo International Organization47 (3) 411ndash41

Sikkink Kathryn 2011 The Justice Cascade How Human RightsProsecutions Are Changing World Politics New York WW Nor-ton amp Company Inc

Simmons Beth A 2009 Mobilizing for Human Rights InternationalLaw in Domestic Politics 1st ed New York Cambridge UniversityPress

Simmons Beth Frank Dobbin and Geoffrey Garrett 2008 TheGlobal Diffusion of Markets and Democracy 1st ed New YorkCambridge University Press

Sinclair Betsy 2012 The Social Citizen Peer Networks and PoliticalBehavior Chicago The University of Chicago Press

Slezak Hrvoje 2009 ldquoProstorska segregacija romskog stanovnistvau Medjimurskoj Zupanijirdquo Hrvatski Geografski Glasnik 71 (2)65ndash81

Statisticni urad Republike Slovenije 2002 ldquoPopis Prebivalstvardquohttpwwwstatsi

Statisticni urad Republike Slovenije 2011 ldquoPovprecne mesecneplace po Skd dejavnost statisticna regija mesec place meritverdquohttpwwwstatsi

Strukelj Pavla 1980 Romi na Slovenskem Ljubljana SloveniaCankarjeva Zalozba v Ljubljani

Tahirovic Haris 2011 Interview by author NotesTancer Mladen 1994 Vzgoja in Izobrazevanje Romov na

Slovenskem Maribor Slovenia Zalozba Obzorja MariborTancer Mladen 1997 Romi v Sloveniji In Strpnost do Manjsin

Zbornik Referatov Mednarodnega Znanstvenega Simpozija edsZlatko Tisljar and Alojz Sirec Inter-kulturo Maribor Pedagoskafakulteta Maribor 94ndash9

Tancer Mladen 2003 Izobrazevanje Romov v Sloveniji In EvropaSlovenija in Romi Zbornik Referatov na Mednarodni Konfer-enci v Ljubljani 15 februarja 2002 eds Vera Klopcic andMiroslav Polzer Ljubljana Slovenia Institut za narodnostnavprasanja 67ndash76

Tesser Lynn M 2003 ldquoThe Geopolitics of Tolerance MinorityRights Under EU Expansion and East-Central Europerdquo EastEuropean Politics and Societies 17 (3) 483ndash532

The Central Intelligence Agency 2002 Balkan Battlegrounds A Mil-itary History of the Yugoslav Conflict 1991ndash1995 Washington DCUS Central Intelligence Agency

The University of Minnesota Human Rights Center 2011 ldquoHumanRights Library Ratification of International HumanRights Treatiesrdquo httpwww1umneduhumanrtsresearchratification-indexhtml

Thorat Sukhdeo and Joel Lee 2005 ldquoCaste Discrimination andFood Security Programmesrdquo Economic and Political Weekly 40(39) 4198ndash201

Tichy Brano 2010 Interview by author NotesTkach Andrew 2010 ldquoBurned Girl a Symbol of Roma Hate

and Hoperdquo httparticlescnncom2010-06-25worldromaprejudice 1 arson-attack-molotov-cocktail-attack-anti-romas=PMWORLD

Tucker Joshua A Alexander C Pacek and Adam J Berinsky 2002ldquoTransitional Winners and Losers Attitudes Toward EU Member-ship in Post-Communist Countriesrdquo American Journal of PoliticalScience 46 (3) 557ndash71

Tudija Milena 2012 Interview by author Voice recordingUS Department of State 2001 ldquo2000 Human Rights Reports Slove-

niardquo httpwwwstategovjdrlrlshrrpt2000eur864htmUS Department of State 2002 ldquo2001 Human Rights Reports Slove-


niardquo httpwwwstategovjdrlrlshrrpt200218391htmUS Department of State 2004 ldquo2003 Human Rights Reports Slove-

niardquo httpwwwstategovjdrlrlshrrpt200327864htmUS Department of State 2006 ldquo2006 Human Rights Reports

Slovak Republicrdquo URL httpwwwstategovjdrlrlshrrpt200678838htm

US Department of State 2009a ldquo2008 Human Rights Reports Bul-gariardquo httpwwwstategovjdrlrlshrrpt2008eur119072htm

US Department of State 2009b ldquo2008 Human Rights ReportsCzech Republicrdquo httpwwwstategovjdrlrlshrrpt2008eur119075htm

US Department of State 2010 ldquo2009 Human Rights ReportsCroatiardquo httpwwwstategovjdrlrlshrrpt2009eur136025htm


US Department of State 2012a ldquo2011 Human Rights ReportsCroatiardquo httpwwwstategovjdrlrlshrrpthumanrightsreportindexhtmdlid=186339

US Department of State 2012b ldquo2011 Human Rights ReportsSloveniardquo httpwwwstategovjdrlrlshrrpthumanrightsreportindexhtmdlid=186405

US Department of State 2013a ldquo2012 Human Rights ReportsCroatiardquo httpwwwstategovjdrlrlshrrpthumanrightsreportindexhtmyear=2012ampdlid=204271

US Department of State 2013b ldquo2012 Human Rights ReportsSloveniardquo httpwwwstategovjdrlrlshrrpthumanrightsreportindexhtmyear=2012ampdlid=204339

Vachudova Milada Anna 2005 Europe Undivided DemocracyLeverage amp Integration After Communism Oxford UK OxfordUniversity Press

Vachudova Milada Anna 2009 ldquoCorruption and Compliance inthe EUrsquos Post-Communist Members and Candidatesrdquo Journal ofCommon Market Studies 47 43ndash62

Vican Dijana and Ivan Milanovic Litre 2006 Nastavni Plan i Pro-gram za Osnovnu Skolu Zagreb Ministarstvo znanosti obrazo-vanja i sporta

Vlada RS 2014 ldquoTretje porocilo Vlade Republike Slovenije opolozaju romske skupnosti v Sloveniji - priloga 4rdquo Porocilo o iz-vajanju Zakona o romski skupnosti v Republiki Sloveniji (Uradnilist RS t 332007) in Nacionalnega programa ukrepov za RomeVlade Republike Slovenije za obdobje 2010-2015

Vratusa Anton 2008 ldquoPanonski prostor in ljudje med dvema trome-jama Kulturno-zgodovinski in druzbeno-ekonomski orisrdquo In Nar-odne manjsine 6 Ziveti z mejo Panonski Prostor in Ljudje ob DvehTromejah Zbornik Referatov na Znanstvenem Posvetu v MurskiSoboti 9-11 novembra 2007 eds Vera Klopcic and Anton VratusaLjubljana Slovenska akademija znanosti in umetnosti Institut zanarodnostna vprasanja 18ndash32

Vugrincic Marijan and Robert Siladi 2008 Iz Zivota RomaKreativnost Roma Doprinos Kulturi i Turizmu Cakovec CroatiaMedjimurski savez sportske rekreacije ldquoSport za sverdquo

Wagner Ulrich Oliver Christ and Thomas F Pettigrew 2008 ldquoPrej-udice and Group-Related Behavior in Germanyrdquo Journal of SocialIssues 64 (2) 403ndash16

546))amp)())amp-13$))amp$+()$$)( ))))((+)

WHO ARE THE ROMA AND WHY USE THEIR CASE HERE

HYPOTHESIS I THE EU ACCESSION PROCESS

Why EU Accession may Influence Individuals



CASE SELECTION



Identification


MEASUREMENT

Games


RESULTS


REFERENCES


countries (Kasambala 2014) When the United Statesrevoked Swazilandrsquos AGOA status due to insufficientprotections of workersrsquo and human rights the decisionaccordingly received much attention (Hughes 2014)The two strategies while chiefly operating at state andinternational levels at least partly rely on domesticaudiences to mobilize in favor of rights change (Keckand Sikkink 1998 Simmons 2009) An individual actorobserving or participating in public discourse howevermight not distinguish between incentives and normpromotion but simply perceive that rights should bechanged In studying individuals then the effects oftwo strategies that often appear together ought to beconsidered together as well

An alternative strategy to change discriminatory be-havior much smaller in scope than the approachesjust mentioned stems from NGO efforts that encour-age contact between the majority population and themarginalized group in order to foster inclusion Whilethe merits of intergroup contact remain contestedlaboratory-based scholarship in psychology stronglysuggests that intergroup contact reduces prejudice anddiscriminatory intent (Allport 1954 Dovidio et al2004 Miller 2002 Pettigrew 1998 Pettigrew and Tropp2006 Wagner et al 2008) Recent evidence from fieldexperiments offers additional support when whitechildren (Green and Wong 2009) or college students(Boisjoly et al 2006 Laar et al 2005) are randomlyassigned to diverse groups for short-term or extendedinteractions they show lower levels of prejudice Ef-forts on the part of NGOs to increase contact betweenthose who transgress and their targets at the individuallevel may then decrease prejudice and discriminationat that level as well Most research on intergroup con-tact is conducted in the American context predomi-nantly examining prejudice of whites towards blacksby studying discrimination in Europe this article offersa departure from this tradition

I conduct a test of the intergroup contact strategyand the combination of incentives and norm promotionin the context of discrimination against the Roma inSlovenia and Croatia The Roma commonly referredto by the disfavored term ldquoGypsiesrdquo (Hancock 2002)are the largest ethnic minority in Europe Historicallythe Roma have been variously enslaved deportedforcibly assimilated and subjected to genocide (Barany2002) Today many remain segregated and face dis-crimination in schools on the job market in hospitalsin police stations and on the street (European RomaRights Centre (ERRC) 2004) This study examines theRoma because they are a clear case of a minority thathas suffered discrimination abuse and exclusion forcenturies In such a seemingly intractable case identi-fying interventions that promote inclusion would showpromise for studying and curtailing marginalization invarious other communities that historically have beendenied or have themselves resisted inclusion

My empirical strategy involves two approaches aborder study and a within-country design I use a bor-der study to examine the combination of incentivesand norms The EU accession process is arguably thestrongest mechanism of human rights conditionality

and is deemed particularly powerful in the contextof Roma rights (Schimmelfennig et al 2005) To gainmembership EU candidate states must enact legisla-tive and policy changes that protect Roma rights atstate and local levels Conditionality demands are ac-companied by norm advocacy and naming and sham-ing via NGOs and official EU channels I leverage thisprocess and measure discrimination against the Romain two towns one in Croatia and one in Slovenia thatmatch closely on 43 Roma and human rights relatedfactors at state regional and local levels At the timeof investigation Croatia was at the height of the EUaccession process and thus faced enormous pressures toimprove its treatment of the Roma In contrast Slove-nia had been an EU member for eight years and facedno pressure to improve Roma rights

I test the intergroup contact strategy in the context ofdiscrimination against the Roma in two towns in Slove-nia that match exceptionally well on 12 localregionalRoma and human rights-related factors but differin type of NGO action Action in the first town isinclusivemdashtargeting both Roma and non-Romamdashandpromotes intergroup contact NGO action in the sec-ond town is not inclusive and focuses on service provi-sion to Romani settlements Accordingly roughly halfof randomly sampled non-Roma from the first townare familiar with the NGO in the other town the sameholds for only two percent NGO leadersrsquo personal id-iosyncrasies completely define the organizationsrsquo fociThis mitigates the endogeneity concern that the twotypes of NGOs developed because of beliefs that theirtypes would be particularly effective in their respectiveenvironments

To test the two hypotheses I use simple gamesintended to capture other-regarding behavior whichmap remarkably well onto the stereotype that moti-vates anti-Roma sentiment Discrimination against theRoma is largely motivated by the belief that the Romaare cheaters and thieves The trust game (Berg et al1995) played with money elicits behavior by non-Romathat demonstrates distrust towards the Roma given thestrong link between distrust and discrimination in thisparticular case the method likely captures discrimina-tory behavior

I find no support for the hypothesis that a high-level combined strategy of incentives and norm pro-motion affects individual behavior At the height of theEU accession process Croats discriminated against theRoma while Slovenes just across the border under nopressure to improve their treatment of the Roma didnot I find support however for the intergroup contacthypothesis Slovenes from a town with inclusive NGOaction treated Roma no differently than non-Romawhereas Slovenes from a town with noninclusive NGOaction discriminated against the Roma

My findings suggest that (1) while powerful at thestate level incentive-based and traditional norm pro-moting strategies inherent in the EU accession processmay not necessarily have effects that reach the indi-vidual and (2) NGOs promoting intergroup contactbetween Roma and non-Roma can help reduce dis-crimination In the context of changes at the individual

531))amp)())amp-13$))amp$+()$$)( ))))((+)














532))amp)())amp-13$))amp$+()$$)( ))))((+)
















533))amp)())amp-13$))amp$+()$$)( ))))((+)











CASE SELECTION



534))amp)())amp-13$))amp$+()$$)( ))))((+)












535))amp)())amp-13$))amp$+()$$)( ))))((+)




Identification




536))amp)())amp-13$))amp$+()$$)( ))))((+)








MEASUREMENT

Games






537))amp)())amp-13$))amp$+()$$)( ))))((+)






radic radic radic


359 227 12-28


10 10 10


radic radic radic



radic radic radic


radic radic radic



radic radic radic


018 019 019


radic radic radic




1169 1927 2176


73 85 85


radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic



radic radic radic


171 17 17


radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic

538))amp)())amp-13$))amp$+()$$)( ))))((+)

Ana Bracic


TABLE 1 Continued



radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic


2008 2007 2007




radic radic











539))amp)())amp-13$))amp$+()$$)( ))))((+)






RESULTS










540))amp)())amp-13$))amp$+()$$)( ))))((+)


















541))amp)())amp-13$))amp$+()$$)( ))))((+)












542))amp)())amp-13$))amp$+()$$)( ))))((+)










REFERENCES

















543))amp)())amp-13$))amp$+()$$)( ))))((+)


















































544))amp)())amp-13$))amp$+()$$)( ))))((+)






















































545))amp)())amp-13$))amp$+()$$)( ))))((+)





















































546))amp)())amp-13$))amp$+()$$)( ))))((+)






CASE SELECTION



Identification


MEASUREMENT

Games


RESULTS


REFERENCES














532))amp)())amp-13$))amp$+()$$)( ))))((+)
















533))amp)())amp-13$))amp$+()$$)( ))))((+)











CASE SELECTION



534))amp)())amp-13$))amp$+()$$)( ))))((+)












535))amp)())amp-13$))amp$+()$$)( ))))((+)




Identification




536))amp)())amp-13$))amp$+()$$)( ))))((+)








MEASUREMENT

Games






537))amp)())amp-13$))amp$+()$$)( ))))((+)






radic radic radic


359 227 12-28


10 10 10


radic radic radic



radic radic radic


radic radic radic



radic radic radic


018 019 019


radic radic radic




1169 1927 2176


73 85 85


radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic



radic radic radic


171 17 17


radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic

538))amp)())amp-13$))amp$+()$$)( ))))((+)

Ana Bracic


TABLE 1 Continued



radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic


2008 2007 2007




radic radic











539))amp)())amp-13$))amp$+()$$)( ))))((+)






RESULTS










540))amp)())amp-13$))amp$+()$$)( ))))((+)


















541))amp)())amp-13$))amp$+()$$)( ))))((+)












542))amp)())amp-13$))amp$+()$$)( ))))((+)










REFERENCES

















543))amp)())amp-13$))amp$+()$$)( ))))((+)


















































544))amp)())amp-13$))amp$+()$$)( ))))((+)






















































545))amp)())amp-13$))amp$+()$$)( ))))((+)





















































546))amp)())amp-13$))amp$+()$$)( ))))((+)






CASE SELECTION



Identification


MEASUREMENT

Games


RESULTS


REFERENCES
















533))amp)())amp-13$))amp$+()$$)( ))))((+)











CASE SELECTION



534))amp)())amp-13$))amp$+()$$)( ))))((+)












535))amp)())amp-13$))amp$+()$$)( ))))((+)




Identification




536))amp)())amp-13$))amp$+()$$)( ))))((+)








MEASUREMENT

Games






537))amp)())amp-13$))amp$+()$$)( ))))((+)






radic radic radic


359 227 12-28


10 10 10


radic radic radic



radic radic radic


radic radic radic



radic radic radic


018 019 019


radic radic radic




1169 1927 2176


73 85 85


radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic



radic radic radic


171 17 17


radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic

538))amp)())amp-13$))amp$+()$$)( ))))((+)

Ana Bracic


TABLE 1 Continued



radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic


2008 2007 2007




radic radic











539))amp)())amp-13$))amp$+()$$)( ))))((+)






RESULTS










540))amp)())amp-13$))amp$+()$$)( ))))((+)


















541))amp)())amp-13$))amp$+()$$)( ))))((+)












542))amp)())amp-13$))amp$+()$$)( ))))((+)










REFERENCES

















543))amp)())amp-13$))amp$+()$$)( ))))((+)


















































544))amp)())amp-13$))amp$+()$$)( ))))((+)






















































545))amp)())amp-13$))amp$+()$$)( ))))((+)





















































546))amp)())amp-13$))amp$+()$$)( ))))((+)






CASE SELECTION



Identification


MEASUREMENT

Games


RESULTS


REFERENCES











CASE SELECTION



534))amp)())amp-13$))amp$+()$$)( ))))((+)












535))amp)())amp-13$))amp$+()$$)( ))))((+)




Identification




536))amp)())amp-13$))amp$+()$$)( ))))((+)








MEASUREMENT

Games






537))amp)())amp-13$))amp$+()$$)( ))))((+)






radic radic radic


359 227 12-28


10 10 10


radic radic radic



radic radic radic


radic radic radic



radic radic radic


018 019 019


radic radic radic




1169 1927 2176


73 85 85


radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic



radic radic radic


171 17 17


radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic

538))amp)())amp-13$))amp$+()$$)( ))))((+)

Ana Bracic


TABLE 1 Continued



radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic


2008 2007 2007




radic radic











539))amp)())amp-13$))amp$+()$$)( ))))((+)






RESULTS










540))amp)())amp-13$))amp$+()$$)( ))))((+)


















541))amp)())amp-13$))amp$+()$$)( ))))((+)












542))amp)())amp-13$))amp$+()$$)( ))))((+)










REFERENCES

















543))amp)())amp-13$))amp$+()$$)( ))))((+)


















































544))amp)())amp-13$))amp$+()$$)( ))))((+)






















































545))amp)())amp-13$))amp$+()$$)( ))))((+)





















































546))amp)())amp-13$))amp$+()$$)( ))))((+)






CASE SELECTION



Identification


MEASUREMENT

Games


RESULTS


REFERENCES












535))amp)())amp-13$))amp$+()$$)( ))))((+)




Identification




536))amp)())amp-13$))amp$+()$$)( ))))((+)








MEASUREMENT

Games






537))amp)())amp-13$))amp$+()$$)( ))))((+)






radic radic radic


359 227 12-28


10 10 10


radic radic radic



radic radic radic


radic radic radic



radic radic radic


018 019 019


radic radic radic




1169 1927 2176


73 85 85


radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic



radic radic radic


171 17 17


radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic

538))amp)())amp-13$))amp$+()$$)( ))))((+)

Ana Bracic


TABLE 1 Continued



radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic


2008 2007 2007




radic radic











539))amp)())amp-13$))amp$+()$$)( ))))((+)






RESULTS










540))amp)())amp-13$))amp$+()$$)( ))))((+)


















541))amp)())amp-13$))amp$+()$$)( ))))((+)












542))amp)())amp-13$))amp$+()$$)( ))))((+)










REFERENCES

















543))amp)())amp-13$))amp$+()$$)( ))))((+)


















































544))amp)())amp-13$))amp$+()$$)( ))))((+)






















































545))amp)())amp-13$))amp$+()$$)( ))))((+)





















































546))amp)())amp-13$))amp$+()$$)( ))))((+)






CASE SELECTION



Identification


MEASUREMENT

Games


RESULTS


REFERENCES




Identification




536))amp)())amp-13$))amp$+()$$)( ))))((+)








MEASUREMENT

Games






537))amp)())amp-13$))amp$+()$$)( ))))((+)






radic radic radic


359 227 12-28


10 10 10


radic radic radic



radic radic radic


radic radic radic



radic radic radic


018 019 019


radic radic radic




1169 1927 2176


73 85 85


radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic



radic radic radic


171 17 17


radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic

538))amp)())amp-13$))amp$+()$$)( ))))((+)

Ana Bracic


TABLE 1 Continued



radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic


2008 2007 2007




radic radic











539))amp)())amp-13$))amp$+()$$)( ))))((+)






RESULTS










540))amp)())amp-13$))amp$+()$$)( ))))((+)


















541))amp)())amp-13$))amp$+()$$)( ))))((+)












542))amp)())amp-13$))amp$+()$$)( ))))((+)










REFERENCES

















543))amp)())amp-13$))amp$+()$$)( ))))((+)


















































544))amp)())amp-13$))amp$+()$$)( ))))((+)






















































545))amp)())amp-13$))amp$+()$$)( ))))((+)





















































546))amp)())amp-13$))amp$+()$$)( ))))((+)






CASE SELECTION



Identification


MEASUREMENT

Games


RESULTS


REFERENCES








MEASUREMENT

Games






537))amp)())amp-13$))amp$+()$$)( ))))((+)






radic radic radic


359 227 12-28


10 10 10


radic radic radic



radic radic radic


radic radic radic



radic radic radic


018 019 019


radic radic radic




1169 1927 2176


73 85 85


radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic



radic radic radic


171 17 17


radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic

538))amp)())amp-13$))amp$+()$$)( ))))((+)

Ana Bracic


TABLE 1 Continued



radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic


2008 2007 2007




radic radic











539))amp)())amp-13$))amp$+()$$)( ))))((+)






RESULTS










540))amp)())amp-13$))amp$+()$$)( ))))((+)


















541))amp)())amp-13$))amp$+()$$)( ))))((+)












542))amp)())amp-13$))amp$+()$$)( ))))((+)










REFERENCES

















543))amp)())amp-13$))amp$+()$$)( ))))((+)


















































544))amp)())amp-13$))amp$+()$$)( ))))((+)






















































545))amp)())amp-13$))amp$+()$$)( ))))((+)





















































546))amp)())amp-13$))amp$+()$$)( ))))((+)






CASE SELECTION



Identification


MEASUREMENT

Games


RESULTS


REFERENCES






radic radic radic


359 227 12-28


10 10 10


radic radic radic



radic radic radic


radic radic radic



radic radic radic


018 019 019


radic radic radic




1169 1927 2176


73 85 85


radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic



radic radic radic


171 17 17


radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic

538))amp)())amp-13$))amp$+()$$)( ))))((+)

Ana Bracic


TABLE 1 Continued



radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic


2008 2007 2007




radic radic











539))amp)())amp-13$))amp$+()$$)( ))))((+)






RESULTS










540))amp)())amp-13$))amp$+()$$)( ))))((+)


















541))amp)())amp-13$))amp$+()$$)( ))))((+)












542))amp)())amp-13$))amp$+()$$)( ))))((+)










REFERENCES

















543))amp)())amp-13$))amp$+()$$)( ))))((+)


















































544))amp)())amp-13$))amp$+()$$)( ))))((+)






















































545))amp)())amp-13$))amp$+()$$)( ))))((+)





















































546))amp)())amp-13$))amp$+()$$)( ))))((+)






CASE SELECTION



Identification


MEASUREMENT

Games


RESULTS


REFERENCES


TABLE 1 Continued



radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic


radic radic radic


2008 2007 2007




radic radic











539))amp)())amp-13$))amp$+()$$)( ))))((+)






RESULTS










540))amp)())amp-13$))amp$+()$$)( ))))((+)


















541))amp)())amp-13$))amp$+()$$)( ))))((+)












542))amp)())amp-13$))amp$+()$$)( ))))((+)










REFERENCES

















543))amp)())amp-13$))amp$+()$$)( ))))((+)


















































544))amp)())amp-13$))amp$+()$$)( ))))((+)






















































545))amp)())amp-13$))amp$+()$$)( ))))((+)





















































546))amp)())amp-13$))amp$+()$$)( ))))((+)






CASE SELECTION



Identification


MEASUREMENT

Games


RESULTS


REFERENCES






RESULTS










540))amp)())amp-13$))amp$+()$$)( ))))((+)


















541))amp)())amp-13$))amp$+()$$)( ))))((+)












542))amp)())amp-13$))amp$+()$$)( ))))((+)










REFERENCES

















543))amp)())amp-13$))amp$+()$$)( ))))((+)


















































544))amp)())amp-13$))amp$+()$$)( ))))((+)






















































545))amp)())amp-13$))amp$+()$$)( ))))((+)





















































546))amp)())amp-13$))amp$+()$$)( ))))((+)






CASE SELECTION



Identification


MEASUREMENT

Games


RESULTS


REFERENCES


















541))amp)())amp-13$))amp$+()$$)( ))))((+)












542))amp)())amp-13$))amp$+()$$)( ))))((+)










REFERENCES

















543))amp)())amp-13$))amp$+()$$)( ))))((+)


















































544))amp)())amp-13$))amp$+()$$)( ))))((+)






















































545))amp)())amp-13$))amp$+()$$)( ))))((+)





















































546))amp)())amp-13$))amp$+()$$)( ))))((+)






CASE SELECTION



Identification


MEASUREMENT

Games


RESULTS


REFERENCES












542))amp)())amp-13$))amp$+()$$)( ))))((+)










REFERENCES

















543))amp)())amp-13$))amp$+()$$)( ))))((+)


















































544))amp)())amp-13$))amp$+()$$)( ))))((+)






















































545))amp)())amp-13$))amp$+()$$)( ))))((+)





















































546))amp)())amp-13$))amp$+()$$)( ))))((+)






CASE SELECTION



Identification


MEASUREMENT

Games


RESULTS


REFERENCES










REFERENCES

















543))amp)())amp-13$))amp$+()$$)( ))))((+)


















































544))amp)())amp-13$))amp$+()$$)( ))))((+)






















































545))amp)())amp-13$))amp$+()$$)( ))))((+)





















































546))amp)())amp-13$))amp$+()$$)( ))))((+)






CASE SELECTION



Identification


MEASUREMENT

Games


RESULTS


REFERENCES


















































544))amp)())amp-13$))amp$+()$$)( ))))((+)






















































545))amp)())amp-13$))amp$+()$$)( ))))((+)





















































546))amp)())amp-13$))amp$+()$$)( ))))((+)






CASE SELECTION



Identification


MEASUREMENT

Games


RESULTS


REFERENCES






















































545))amp)())amp-13$))amp$+()$$)( ))))((+)





















































546))amp)())amp-13$))amp$+()$$)( ))))((+)






CASE SELECTION



Identification


MEASUREMENT

Games


RESULTS


REFERENCES





















































546))amp)())amp-13$))amp$+()$$)( ))))((+)






CASE SELECTION



Identification


MEASUREMENT

Games


RESULTS


REFERENCES

reaching the individual: eu accession, ngos, and human rights · human rights monitors look harder...

Documents