Race discrimination impact litigation in Eastern Europe - A discussion of legal strategies to confront racial discrimination

15 July 1997

James Goldston

Conversations with lawyers, judges, police officials, former defendants, and other persons familiar with the operation of the criminal justice system in a number of countries in Eastern Europe yield the common observation that Roma are discriminated against at many different points.

The kinds of discrimination Roma face are, however, generally manifestations of practice, not of law. Europe has few “Jim Crow” laws of the kind which, for many years in the United States, expressly treated African-Americans differently from whites — by mandating segregation in public schools, for example. Rather, in administering, applying and interpreting rules and statutes which are, on their face, racially neutral, public officials commonly treat Roma differently from — and, almost invariably, worse than — racial/ethnic majorities. Thus, in the criminal justice systems of some countries, while no law explicitly sanctions differential treatment on the basis of race or ethnicity, reports proliferate of Roma victims and defendants suffering discriminatory treatment.

Over the past several months, the ERRC has hosted discussions with sociologists and lawyers from around the region. The focus of these discussions has been how to obtain sufficiently compelling proof of race discrimination, as well as how to devise legal strategies for challenging it and obtaining remedies.

On April 17, together with the Constitutional and Legislative Policy Institute (COLPI), the ERRC held a discussion with lawyers and sociologists from Hungary concerning possible models of anti-discrimination litigation. On May 25, also in Budapest, the ERRC brought together lawyers from Bulgaria, the Czech Republic, Hungary and Romania to discuss the topic. During both sessions, participants simultaneously identified numerous obstacles to — and emphasised the need for — gathering the quantitative evidence necessary to bring certain broad-based discrimination lawsuits.

Roma enter the criminal justice system in two ways: as defendants and as victims. In both of these roles, the opportunity for discriminatory treatment appears at a number of different points. First, as defendants: (i) Pre-trial detention: Romani defendants who are arrested by the police for alleged involvement in criminal activity are, on a per capita basis, more frequently detained prior to trial than non-Roma arrested for alleged involvement in similar crimes; (ii) Conviction: Romani defendants who are indicted by the prosecution for certain crimes are, on a per capita basis, more frequently convicted at trial than non-Roma indicted and taken to trial for the same crimes; (iii) Sentence: Sentences imposed on Romani defendants who are convicted at trial of certain crimes are, on average, more severe than sentences imposed on non-Roma convicted of the same crimes.

Second, as victims: (i) Commencing an official investigation: Reported crimes of violence against Romani victims result in the opening of investigation by law enforcement bodies far less frequently, on a per incident basis, than do reported crimes of violence against non Romani victims; (ii) Arrest, indictment, and conviction: Reported crimes of violence against Romani victims result in the arrest, indictment and conviction of perpetrators far less frequently, on a per incident basis, than do reported crimes of violence against non- Europe has few “Jim Crow” Romani victims; (iii) Sentence: Sentences imposed upon perpetrators of crimes against Romani victims are, on average, less severe than sentences imposed upon the perpetrators of the same crimes against non-Romani victims.

And yet, there are few hard facts to document these widely held beliefs. A defendant who claims s/he has been victim of discriminatory treatment in pre-trial detention — i.e. that s/he has been detained prior to trial even though similarly situated non-Roma charged with identical crimes would not have been — must offer some substantiation. One form of proof might be the statements of law enforcement officials investigating the case or of the judge trying the case which reveal an intent to discriminate on the basis of race. A statement by a prosecutor along the lines of, “I am going to ask the judge to detain you prior to trial because I don’t like the colour of your skin,” would be an obvious, if unusually candid, example. Of course, such direct and compelling proof of discrimination is very rare.

In considering how to challenge systemic race discrimination, it may help to study examples of similar research and litigation that have been successfully carried out elsewhere. I am aware of analogous litigation in South Africa and India, but am most familiar with the civil rights litigation that has taken place over the past several decades in the United States. I offer for consideration one case that is now ten years old. The complainant in this case did not challenge a statute or regulation; rather, he argued that a governmental practice — the administration of the death penalty by the courts of one of the fifty states — was carried out in a racially discriminatory manner.

Warren McCleskey was a black man convicted of murdering a white police officer in the state of Georgia. At trial, he was sentenced to death. On appeal, McCleskey argued that his death sentence should be overturned on the grounds that the death penalty in Georgia was imposed in a racial discriminatory fashion. McCleskey did not argue that the judge or prosecutor or jury in his own case was biased against him. Instead, using quantitative evidence examining capital sentencing in the state of Georgia over time, he argued that, as a matter of statistical probability, there existed an unacceptable risk that racial considerations had infected the decision process concerning whether to impose the death penalty. In U.S. terms, McCleskey argued that his sentence had been imposed in violation of the Fourteenth Amendment of the Constitution, which guarantees the equal protection of the laws, and the Eighth Amendment, which prohibits cruel and unusual punishment. McCleskey grounded his argument not in any statute or administrative regulation, but in the bare language of these two constitutional provisions.

In support of his legal argument, McCleskey offered a comprehensive statistical study of over 2,000 murder cases in the state of Georgia in the 1970s. Considerable care was taken while conducting the survey to isolate non-racial factors which might have explained why the death sentence had been imposed, such as whether the defendant acted alone, the victim was a law enforcement officer, the murder was unusually gruesome, etc. After statistically taking account of such factors, the study reached the following conclusions:

  1. Disparate impact based on the race of the victim: defendants charged with killing white victims received the death penalty in % of the cases; defendants charged with killing black victims received the death penalty in 1% of the cases.
  2. Reverse disparate impact based on the race of the defendant: 4% of black defendants charged with murder received the death penalty; 7% of white defendants charged with murder received the death penalty.
  3. Disparate impact based on combined race of victim/defendant: The death penalty was imposed in 1% of the cases involving black defendants and black victims; 3% of the cases involving white defendants and black victims; 8% of the cases involving white defendants and white victims; and 22% of the cases involving black defendants and white victims.
  4. All in all, defendants charged with murder of white victims were 4.3 times as likely to receive the death penalty as defendants charged with murder of black victims.

The case went all the way to the Supreme Court, where McCleskey ultimately lost by one vote (5-4) in 1987. Though accepting the statistical validity of the evidence presented on McCleskey’s behalf, the Supreme Court majority ruled that the evidence was insufficient, because, under then-applicable standards of proof, McCleskey had failed to demonstrate convincingly that the criminal justice actors in his case had acted with discriminatory purpose.

Notwithstanding the fact that McCleskey’s statistical argument narrowly failed, McCleskey v. Kemp is important for several reasons:

First of all, McCleskey very nearly won. Indeed, he came so close that the deciding Justice, who wrote the majority opinion and retired soon afterward, later stated that his vote in the McCleskey case was the one vote in his long career which he would reverse if he subsequently had the opportunity. McCleskey’s narrow loss must be viewed in the unique political context of United States race discrimination law. Specifically, after more than two decades as a progressive leader in this area of law world-wide, the Supreme Court in recent years has reversed course markedly, so that the standard of proof required to show race discrimination in the United States is and was at the time of the McCleskey verdict, very stringent.

Moreover, the McCleskey model contains several of the elements of a race discrimination lawsuit which, when adapted to local conditions, might be brought on behalf of Roma in Europe, including: (i) a challenge to a govern mental practice, not a law, on grounds of race discrimination; (ii) a legal argument resting on broad constitutional provisions, rather than more specific implementing legislation; (iii) proof consisting largely of statistics; and (iv) a highly persuasive claim that race discrimination can exist even absent evidence of specific intent on the part of actors in a particular case.

McClesky v. Kemp is not the only United States case which may offer inspiration to European lawyers involved in anti-discrimination litigation. Other litigation may offer models of race discrimination challenges that, with appropriate modifications, could be applied successfully in Eastern Europe. At least two of the doctrinal arguments which courts in the United States have employed to strike down terrain statutes as unconstitutional might provide a foundation for challenges to race discrimination in this region.

First, United States courts have found unconstitutional state laws that use race-neutral terms – that is, they do not refer to race at all – but which are administered in ways that produce discrimination. Yick Wo v. Hopkins, an 1886 decision of the U.S. Supreme Court, involved a San Francisco ordinance requiring owners of hand laundries to obtain a permit to operate, supposedly to ensure the safety of the operation. The ordinance at issue did not refer to race, and it was conceded that the aim of the ordinance — promoting safety — was important. Nonetheless, the Supreme Court invalidated the ordinance as a violation of the U.S. Constitution’s equal protection clause, because in administering the ordinance, the city authorities had granted permits to almost every non-Chinese applicant who applied and had denied permits to every Chinese applicant.

Under the doctrine enunciated in Yick Wo, discrimination in administering neutral laws is unconstitutional. Thus, for example, if a state law authorises a literacy test for voting, but allows voting registrars to administer the test one way to whites and another to African Americans, the administration of the literacy test, though not the test itself, might be unconstitutional. Similarly, if a criminal procedure code sets forth criteria under which defendants may be detained prior to trial, but allows judges and prosecutors to interpret the criteria differently for Roma and for non Roma, such differential criteria in certain cases might violate the equal protection norms embodied in the International’ Covenant on Civil and Political Rights, the European Convention on the Protection of Human Rights and Fundamental Freedoms, and the constitutions of most Eastern European countries.

A litigator seeking to challenge race discrimination in pre-trial detention could use the Yick Wo doctrine in at least two ways. First, the actual pattern of results might demonstrate impermissible race discrimination. In Yick Wo, the observed pattern involved who was granted permits to operate laundries and who was denied; in a hypothetical challenge to pre-trial detention, one would seek to document who was detained and who was rot. Alternatively, the litigator might use Yick Wo to attack a race neutral law on grounds of vagueness. In the hypothetical East European case, one might challenge the portion of the criminal procedure code governing pre-trial detention not because pre-trial detention itself is inherently discriminatory, but ratter because the portion of the criminal procedure code governing pre-trial detention is so vague that it lends itself to racial discrimination in administration.

Yet another doctrinal strand from U.S. race discrimination litigation which may, with certain changes, be applied to Eastern Europe involves statutes that do not use racial terms, but that, even when administered even-handedly, impose disadvantages on African-Americans more frequently than on whites. In the United States, this doctrine was developed in a series of cases addressing the constitutionality of poll taxes — payments which, by law, a number of states in the South required of all voters. The poll taxes applied equally to all citizens African American and white. Poor whites as well as poor African-Americans were unable to vote because they could not afford to pay the taxes. However, poll taxes were struck down as unconstitutional, because in practice there were proportionately more poor African Americans than poor whites. Thus, even an even-handed administration of the poll tax excluded more African-Americans than whites from voting.

Similarly, in the context of a challenge to race discrimination in pre-trial detention in Eastern Europe, it may be that the statutory criteria for pre-trial detention (contained in the criminal procedure code) are being applied by judges and prosecutors even-handedly. Nonetheless, lawyers might challenge such criteria on the grounds that they result in the pre-trial detention of disproportionately more Roma than non-Roma, because in practice disproportionately more Roma criminal defendants satisfy the criteria for detention than do non-Roma defendants. Of course, this line of argument carries more weight where the procedure at issue, like the US poll tax, has a thinly-veiled discriminatory purpose.

All of the strategies delineated above involve arguments which emphasise that, notwithstanding the absence of discriminatory intent, there exists statistical evidence of discriminatory effect-i.e., that, overtime, Roma defendants are detained more frequently and for longer periods of time than similarly situated non-Roma defendants charged with identical crimes. Thus, though no one may have intended to discriminate, because pre-trial detention decisions are influenced by race on a systematic basis, there is an intolerable risk that race has affected the pre-trial decision in the defendant’s own case. Even absent direct proof of intent to discriminate in the defendant’s case, a substantial risk that race has infected the criminal justice process is incompatible with a government’s commitment to ensure that defendants charged with crimes are — and are seen to be — treated equally.

A race discrimination impact litigation project might be undertaken in two stages. In the first stage, the project would conduct the research necessary to develop a comprehensive body of statistics capable of documenting the existence of racial discrimination in pre-trial detention. The results of this research would be analysed and published to serve an independent public education function. In the second stage, if the research demonstrates the existence of discrimination, the project would sponsor litigation — in domestic and/or international fora — aimed at challenging the discriminatory practices and developing suitable remedies.

Participants in discussion sessions have identified a number of potential problems which such a project might encounter in its research phase. In the first place, participants expressed qualms over the very possibility that such a study could be conducted. A number of countries in the region impose restrictions on the maintenance, collection and publication of certain kinds of information. Criminal justice statistics are often incomplete, inaccessible and/or colour-blind; the law prohibits reference in such records to the race or ethnicity of a defendant or victim. Also, data protection laws in some countries restrict or prohibit the publication of information concerning individuals. In addition to the foregoing legal problems, there exists the underlying demographic problem of how to identify who is a member of a minority group and who is not. Indeed, many Roma, in particular, have been understandably reluctant to identify themselves as such.

A second problem is ethical: who can ensure that such statistics will not serve a purpose directly contrary to that intended? Statistics have been used at different times throughout the twentieth century to harm, rather than help, minority populations. Many prior efforts to gather statistics commenced with benevolent intentions, only to produce deleterious consequences for the intended beneficiaries. Finally, gathering and publicising statistics along racial lines can serve to entrench racial differences. By highlighting the relationship between Roma and criminality, they may reinforce negative stereotypes of and prejudices about Roma which pervade many countries in the region. Should these concerns be overcome, methodological problems remain. Any effort to conduct research in this area must address the following questions: Which stage in the criminal justice process — investigation, arrest, pre-trial detention, indictment, trial, conviction, sentence should be the focus of research? Which stage is likely to provide the most graphic and dramatic evidence of racial discrimination? In which legal position — that of defendant or victim — are Roma more frequently subjected to discrimination? What control factors must be isolated to ensure that any differences in result are in fact attributable to the race of the victim/defendant, and not to race-neutral characteristics? How large should the scope of the research be? Several countries? One county? One court? Over what period of time? How large must be the number of cases under research to be statistically significant?

By comparison, the McCleskey study focused on more than 2,000 cases over a six year period in every jurisdiction in the state of Georgia. The statistical study employed in the McCleskey case accounted for 230 different factors which could have explained the disparities at issue on non-racial grounds.

Apart from technical and ethical obstacles to this research, many lawyers raised doubts over the possibility of employing in Eastern Europe litigation strategies developed in other countries, many of them common law jurisdictions. Notwithstanding these serious concerns, it is important to remember that race discrimination litigation in such countries as the United States, India and South Africa was born out of legal struggle, with courageous lawyers working in often inauspicious circumstances. These countries may offer creative methods to challenge racially discriminatory practices.

The civil rights lawyers who launched the first modern race discrimination lawsuits in the United States in the 1930s and 1940s — in the years before the landmark Supreme Court ruling in Brown v. Board of Education (1954) — had to contend with many of the same problems advocates in Eastern Europe face today: judicial hostility to statistical evidence; unfamiliarity with class actions; inexperience at fashioning systemic remedies; and reluctance to address controversial claims of a “political” nature. American lawyers were, of course, operating in a common law system where precedent ruled. At the start, however, the precedent system was, at least in part, a hindrance: much of the early school desegregation litigation aimed at overcoming and, finally, reversing the well-established doctrine of “separate but equal” in public schools. By contrast, due to the relative absence of precedent in this region, Eastern European lawyers have the relative advantage of starting on a more level playing field.

It may ultimately be that the techniques used to attack race discrimination in other countries are not applicable in Eastern Europe. Given the severity of the problem, however, as well as the profound changes in so many areas of life which the region has witnessed in recent years, it would seem fool hardy not to try.


Challenge discrimination, promote equality


Receive our public announcements Receive our Roma Rights Journal


The latest Roma Rights news and content online

join us

Find out how you can join or support our activities