Session by xiaohuicaicai


									      Statistique, Développement et Droits de l‘Homme

           Session C-Pa 5b

Statistics in the Combat of Racial
Discrimination: A Necessary Tool

     James A. GOLDSTON

                    Montreux, 4. – 8. 9. 2000
                             Statistique, Développement et Droits de l‘Homme

Statistics in the Combat of Racial Discrimination: A Necessary
Deputy Director, Open Society Institute
400 West 59 Street
New York, NY 10019, USA
T. + 1 212 548 0600 F. + 1 212 548 4679


      Statistics in the Combat of Racial Discrimination: A Necessary Tool

       The tension between the right to information and the right to privacy has been widely
recognised, if not definitively resolved. Nowhere is the difficulty of reconciling these tensions more
pronounced than in the field of race relations. Fundamental to the task of promoting civil rights and
non-discrimination throughout Europe is accurate documentation of the subordinated position of
racial and ethnic minorities in many areas of public life. Statistical information is a prerequisite for
the formulation of government policy. It is particularly crucial in addressing, and providing
evidence to support, claims of racial discrimination. And yet, if statistics are needed to document
the condition of minorities and prove legal violations, many understandably fear the abusive
purposes to which statistics have been - and can be - put. Some actively oppose renewed efforts to
gather information on the number of ethnic minorities in schools, courthouses and prisons. Others,
mistrustful of the willingness and/or capacity of government and other officials to maintain the
confidentiality of collected information, counsel non-cooperation with surveys and census counts.
Some data protection laws are interpreted so as to hinder the collection of race- or ethnic-coded
statistics. The unfortunate consequence is that non-discrimination advocates are handicapped by
scarce reliable information, and allegations of racial discrimination often lack the persuasive
power which statistical evidence would provide. The paper will address the need for race- and
ethnic- based statistics, and evaluate the extent to which the gathering and maintenance of such
statistics are permissible under prevailing legal standards and useful for advocates who seek legal
redress for discrimination, as well as for governments seeking to promote anti-discrimination


      Statistiques dans le combat de la discrimination raciale : un outil nécessaire

     La tension entre le droit à l’information et le droit à la vie privée a été largement reconnue
même si elle n’a pas été résolue définitivement. Dans aucune autre domaine, la difficulté de
réconcilier ces tensions n’est aussi prononcée que dans le domaine des relations interraciales.
Dans le travail de promotion des droits civils et de la non discrimination à travers l’Europe, la
documentation précise de la position subordonnée des minorités raciales et ethniques dans de
                                          Montreux, 4. – 8. 9. 2000
                                    Statistique, Développement et Droits de l‘Homme

nombreux secteurs de la vie publique, est un facteur fondamental. L’information statistique est un
pré-requis pour la formulation de politiques gouvernementales. Son rôle est particulièrement
crucial pour aborder et fournir la preuve des affirmations de discrimination raciale. Pourtant, si
l’on a besoin des statistiques pour documenter la condition des minorités et pour prouver les
violations de la loi, nombreux sont ceux qui craignent, cela se comprend, que les statistiques soient
utilisées à d’autres fins que celles initialement proposées. Certains s’opposent vivement aux efforts
renouvelés visant à la réunion d’informations sur le nombre des minorités ethniques dans les
écoles, les tribunaux et les prisons. D’autres, doutant de la volonté et/ou de la capacité du
Gouvernement et d’autres services officiels à conserver la confidentialité de l’information collectée,
préconisent la non adhésion aux décomptes des enquêtes et recensements. Certaines lois sur la
protection des données sont interprétées comme étant un obstacle à l’établissement de statistiques à
indicateurs raciaux ou ethniques Conséquence malheureuse, les avocats de la non discrimination
sont handicapés par le manque d’information fiable et les allégations de discrimination raciale
manquent souvent du pouvoir de persuasion qu’une preuve statistique pourrait donner. Cet exposé
traitera de la nécessité de disposer de statistiques basées sur les races et les ethnies et évaluera
dans quelle mesure l’établissement et le maintien de telles statistiques est admissible dans le cadre
des standards légaux actuels, et utile pour les avocats qui recherchent des recours légaux à la
discrimination ainsi que les Gouvernements qui cherchent à promouvoir des politiques contre la

1. Introduction

        The tension between the right to information and the right to privacy has been widely recognised,
if not definitively resolved. Nowhere is the difficulty of reconciling these interests more pronounced
than in the field of race relations. Fundamental to the task of promoting civil rights and non-
discrimination throughout Europe is accurate documentation of the subordinated position of racial and
ethnic minorities in many areas of public life. Statistical information is a prerequisite for the
formulation of government policy. It is particularly crucial in addressing, and providing evidence to
support, claims of racial discrimination. And yet, if statistics are needed to document the condition of
minorities and prove legal violations, many understandably fear the abusive purposes to which
statistics have been – and can be – put. Some actively oppose renewed efforts to gather information on
the number of ethnic minorities in schools, courthouses and prisons. Others, mistrustful of the
willingness and/or capacity of government and other officials to maintain the confidentiality of
collected information, counsel non-cooperation with surveys and census counts. Some data protection
laws are interpreted so as to hinder the collection of race- or ethnic-coded statistics. 1 The unfortunate
consequence is that non-discrimination advocates are handicapped by scarce reliable information, and
allegations of racial discrimination often lack the persuasive power which statistical evidence would
        These problems, long apparent, have become more pressing with, on the one hand, a widely
reported resurgence in recent years in acts of racial and ethnic discrimination and violence across much
of Europe, and, on the other, renewed efforts – by governments, civil society and Europe-wide
institutions – to reverse this trend. Within the confines of this relatively brief paper, I offer some initial
thoughts about these issues.2

  Throughout this paper, the phrases “race- or ethnic-coded statistics”, “race or ethnic statistics”, “race statistics”, “race-
or ethnic-coded data”, and “statistics broken down by race or ethnicity” are used interchangeably to refer to informat ion
that is differentiated as to the race and/or ethnicity of the persons to whom it pertains.
  This paper is limited in length to comply with the requirements for contributed papers of the International Conference
on Statistics, Development and Hu man Rights. An expanded version of this paper will be published after the
Conference‟s conclusion.
                                                  Montreux, 4. – 8. 9. 2000
                                  Statistique, Développement et Droits de l‘Homme

       After a generation in relative abeyance,” one observer has noted, “„the political discourse of
racism‟ has once again forced itself onto Europe‟s agenda…. A resurgence in right-wing parties of a
racist nature has been evident across most continental European nations in the 1990s.” 3 A
government report published this March noted a rise in “racist and a nti-Semitic violence” in France
in 1999. 4 No lesser an authority than European Union Enlargement Commissioner Gunter
Verheugen has acknowledged – in his former capacity as Minister of State at the German Foreign
Office – “an alarming increase in racist and xenophobic tendencies in all European countries,
including Germany.”5 And, in its “Human Rights Agenda for the European Union for the Year
2000,” the EU‟s own Comite des Sages expressed concern that, “Within the Union, large-scale
discrimination persists in various forms. Racism and xenophobia are thriving.” 6
       EU residents who were born outside the EU “have been subjected to greatly increased levels
of racially motivated attacks” over the past decade. 7 Heightened violence against foreign-born
immigrants and racial minorities has been documented in Germany, 8 the United Kingdom, 9 and
other countries. News outlets from across the political spectrum have reported that “Europe‟s
difficulties in integrating millions of immigrants are feeding xenophobic movements….”10
       And in Central and Eastern Europe, where ten countries are lining up to join the EU in the
next several years, the situation appears no better. A survey this spring commissioned by the United
Nations High Commission for Refugees revealed a rise in “overt xenophobic phenomena” in
Hungary, where 38% of those polled said Hungary should not receive any refugees. 11 In February, a
racist Czech computer game appeared on the Internet. Entitled “Shoot your Gypsy down,” the game
purportedly allows players to fire at Roma who are trying to hide behind a brick wall. 12 A few
months earlier, researchers from the STEM agency reported the results of a study indicating that
90% of ethnic Czechs denied that Roma in the Czech Republic suffer discriminatory treatment. 13 A
similar poll conducted at the same time by the TNS Polling Institute in Bratislava concluded that
more than three of five Slovaks say they favor segregating the Roma minority from the majority

  C. Gearty, “Racis m, Religious Intolerance, Xenophobia,” in P. A lston, ed., The EU and Hu man Rights (1999), p. 328.
  “France Finds Rise in Racist Violence,” International Herald Tribune, March 16, 2000.
   G. Verheugen, “The Answer Is – Europe,” in European Monitoring Centre on Racism and Xenophobia, European
Media Conference – Cu ltural Diversity against Racism: Docu mentation (1999), p. 22.
   The Co mite des Sages consisted of Judge Antonio Cassese, Catherine Lalu miere, Peter Leuprecht, and Mary
Robinson. See Comite des Sages, “Leading by Examp le: A Hu man Rights Agenda for the European Union for th e Year
2000” (1998), para. 11. See also P. Alston and J.H.H. Weiler, “The Eu ropean Union and Human Rights: Final Project
Report on an Agenda for the Year 2000” (1998), para. 32 (noting “a resurgence in racist and xenophobic behavior, a
failure to fu lly live up to equality norms or to eliminate various types of discrimination”).
  C. Gearty, supra, p. 329.
   See, e.g., The Economist, August 5-11, 2000, p. 32 (reporting a “growing number of attacks on foreigners” in
Germany); John Schmid, “German Split Persis ts in Jobless Data and Attacks on Foreigners,” International Herald
Tribune, August 9, 2000, p. 1 (“Almost daily reports of skinhead and neo -Nazi activ ity have riveted attention on the
increase in hate crimes, which is concentrated in the five fo rmer East German states”). See also Hu man Rights
Watch/Helsinki, ‟Germany for Germans‟: Xenophobia and Racist Vio lence in Germany (1995).
   See “The Stephen Lawrence Inquiry: Report of an Inquiry by Sir William Macpherson of Cluny” (1999); Hu man
Rights Watch/Helsinki, “Racist Violence in the United Kingdom” (1997).
    Y. Trofimov, “Berlusconi urges anti-immig ration bill,” The Wall Street Journal Europe, March 31 – April 1, 2000, p.
   Nepszabadsag, March 10, 2000, p. 5.
    Transitions Online, Week in Review, 21 – 27 February 2000. The “wall” was apparently an allusion to a wall
constructed in the town of Usti nad Labem in 1999 at the behest of ethnic Czech residents who sought to restrict the
movement of Ro ma in the town. After much public acrimony, the wall was pulled down after crit icis m fro m the
European Union and Ro ma rights activists.
   Rad io Prague, December 29, 1999.
                                               Montreux, 4. – 8. 9. 2000
                                  Statistique, Développement et Droits de l‘Homme

population, including in separate schools. 14 Of course, throughout much of Central and Eastern
Europe, the Roma remain “the one people … whom it is largely acceptable to persecute.” 15
       In response, Europe‟s official institutions have not been silent. Article 13 of the Treaty
establishing the European Community (as modified by the Amsterdam Treaty) authorises the
European Council, acting unanimously on a proposal from the Commission and after consulting the
European Parliament, to “take appropriate action to combat discrimination based on sex, racial or
ethnic origin, religion or belief, disability, age or sexual orientation.” The new Article 29 of the
TEC establishes for the first time that one of the central objectives of the Union is to provide
citizens with a high level of safety within an area of freedom, security and j ustice by developing
common action in the field of police and judicial cooperation in criminal matters and by preventing
and combating racism and xenophobia.
       In 1998, the European Monitoring Centre on Racism and Xenophobia opened in Vienna,
charged with studying the extent and development of racism, xenophobia and anti-Semitism in the
EU, with gathering and analysing information at European level, with establishing a racism and
xenophobia information network, and with making recommendations on policy to European
institutions and EU Member States. In June 2000, pursuant to Article 13 of the TEC, the European
Union adopted a directive (the “Race Directive”) to require member states to outlaw through
legislation and enforcement mechanisms racial discrimination. 16 The Race Directive marks a
significant step forward and will give advocates of anti-discrimination important legal tools to
combat discriminatory practices. One month earlier, the Committee of Ministers of the Council of
Europe adopted Protocol 12 of the European Convention on Human Rights. This provision, once
ratified by the governments of ten member States, will for the first time give rise to an independent
right (enforceable by the European Court of Human Rights) not to be subjected to discrimination on
a host of grounds, including race. Protocol 12 would constitute a major advance over Article 14 of
the Convention, which presently prohibits discrimination only in the enjoyment of other rights of
the Convention.
       And yet, in seeking to capitalize on these new region-wide legal norms, many governments and
non-governmental advocates for equality will be constrained by the scarcity in many countries of
accurate data broken down by race or ethnicity which might document claims of discrimination and/or
serve as reliable measures of progress. Race statistics are not widely available in Europe. Neither
Eurostat – the official statistical body of the European Union – nor the International Labour
Organisation in Geneva appears to maintain statistics broken down by race or ethnicity. Moreover,
many governments simply do not gather such statistics.
       The first Annual Report of the European Union‟s Monitoring Centre on Racism and
Xenophobia notes that, “although it is possible to obtain a qualitative view of racism and
xenophobia, at least in their most spectacular manifestations, it is very difficult to arrive at precise
and reliable statistical data. Quantitative assessments pose many problems that have to be overcome

    Radio Free Europe/Radio Liberty, December 30, 1999, citing a December 28, 1999 article in the Czech daily,
“Hospodarske noviny.”
   S. Erlanger, “Life is getting much worse for Gypsies, Europe‟s scorned minority,” International Herald Tribune, April
3, 2000, p. 6. Of course, anti-Ro ma sentiment is by no means confined to Europe. In March, a Canadian judge acquitted
six men of promoting hatred against the Roma people. Accused of demonstrating outside a motel housing Czech Ro ma
asylum applicants, the men were reported to have worn swastikas, made stiff -armed salutes, and held up placards urging
motorists to “Honk if you hate Gypsies.” The Ontario Co urt judge found no evidence that “Roma” and “Gypsy” meant
the same thing. J. Saunders, The Globe and Mail, March 25, 2000, p. A-10.
    Council Direct ive 2000/43/ EC of 29 June 2000, implementing the principle of equal treatment between persons
irrespective of racial or ethnic origin, published in Official Journal of the European Co mmun ities, Vo l. 43, 19 July
                                               Montreux, 4. – 8. 9. 2000
                                  Statistique, Développement et Droits de l‘Homme

if the acts involved are to be categorised, the victims defined and the social environment in question
along with the vulnerable individuals and groups identified.” 17
       Information published by the Council of Europe in 1998 18 indicates that official information
broken down by race/ethnicity is rare. Thus, of 37 countries which took part in a comprehensive
survey which formed the basis of the study, only four countries – Bulgaria, Cyprus, the
Netherlands, and United Kingdom – included information about ethnicity in the last census or in the
official population register. 19 In only the Netherlands and the United Kingdom were surveys
containing information on ethnicity common. 20 An additional eleven countries – all in Central and
Eastern Europe and the Balkans – asked about “national group” at the last census: Russia, Ukraine,
Romania, Belarus, Hungary, Slovak Republic, Latvia, Estonia, Croatia, the Former Yugoslav
Republic of Macedonia, Slovenia. Lithuania asked about “nationality.” 21 A number of these
countries included information on national origin in various surveys, registration data and migration
statistics. 22
       The dearth of race statistics poses serious problems for governments seeking to fulfill legal
obligations and to victims of discrimination who strive to prove their cases in court or before
administrative agencies. To take one example, the newly minted EU Race Directive defines its core
principle – “the principle of equal treatment” – as the absence of “direct or indirect discrimination
based on racial or ethnic origin.”23 While “direct discrimination” often requires the existence of a
law or practice which treats people differently on its face (“no blacks may apply”), “indirect
discrimination” is often more subtle. Indirect discrimination occurs “where an apparently neutral
provision, criterion or practice would put persons of a racial or ethnic origin at a particular
disadvantage compared with other persons.” 24
       Hence, laws (which are in force in a number of Central and Eastern European countries)
establishing “special schools” for the “mentally deficient” might be held to discriminate indirectly
against Roma students, if it could be shown that administrators applying such laws
disproportionately over-assign Roma to these schools, absent educational justification. And yet, to
establish the existence of racially disproportionate assignment patterns – i.e., that the laws in
practice “put persons of a racial or ethnic origin at a particular disadvantage compared with other
persons” – it may well be necessary to obtain statistics documenting school assignments for
students of different race/ethnicity. Indeed, the Race Directive expressly authorizes the use of
statistical evidence to prove indirect discrimination: “The appreciation of the facts from which it
may be inferred that there has been direct or indirect discrimination is a matter for national judicial
or other competent bodies, in accordance with rules of national law or practice. Such rules may
provide in particular for indirect discrimination to be established by any means including on the
basis of statistical evidence.”25

   Eu ropean Monitoring Centre on Racism and Xenophobia, Annual Report 1998, p. 5.
   Werner Haug, Youssef Courbage, Paul Co mpton, The demographic characteristics of national minorities in certain
European states (Council o f Europe: 1998). The data contained in this work was derived fro m a “survey to determine
the statistical sources of informat ion, existing or potential, concerning religion, language, national group and ethnic
group” sent to the Statistical Offices of 43 European countries in 1995. The Council of Europe Group of Specialists
prepared the survey on the demographic situation of national minorities, with the aid of Eurostat and the United Nations
Economic Co mmission for Europe. Id., p. 23.
   Id., pp. 54-55.
   Id., p. 55. United Kingdom surveys including information on ethnic groups addressed households and the working
population, among other sectors. In the Netherlands, information concern ing the place of birth of father/mother may be
found in the population register, emp loyment surveys, and immigrat ion and vital statistics. Id., pp. 55-56.
   Id., pp. 42-47.
   Id., pp. 48-49.
   Race Directive, Art. 2(1).
   Race Directive, Art. 2(2)(b).
   Race Directive, Preamb le, para. 15 (emphasis supplied).
                                               Montreux, 4. – 8. 9. 2000
                                 Statistique, Développement et Droits de l‘Homme

       Without such statistics, many victims of racial discrimination are unable to pursue and obtain
redress for legitimate claims. 26 Moreover, as noted above, governments can hardly be expected to
comply with international obligations to eradicate racial discrimination without data showing the
racial impact of policies in the fields of, inter alia, employment, housing, education, and criminal
       And still, numerous concerns about race- or ethnic-coded data collection persist, often within
the very minority communities on whose behalf the supporters of such data purport to advocate.
Among the most common objections voiced are the following:
        Legacy of Abuse: The past legacy of intentional misuse of race- and ethnic-coded data to
          the detriment of minorities and the continuing potential for abuse – perhaps most evident
          in the field of crime statistics – mandate caution in the collection and dissemination of
          such information.
        Reinforcing Stereotypes: Closely related to the foregoing is the concern that the very
          effort to gather and publish race statistics will reinforce negative stereotypes about
        Legal Restriction: It is argued that international law and/or the domestic legislation of
          numerous countries in Europe restrict the gathering and maintenance of race statistics.
        Right to Privacy: Data collection interferes with the right to privacy, which includes the
          right to decide what kinds of information are collected and maintained about each
        Choice of Identity: Data collection interferes with the exclusive right of an individual to
          claim, maintain and express his/her ethnic identity by, effectively, imposing an external
          choice of identity upon the individual. The very question of ethnic identity (who is
          “Roma”; who is “Hungarian”) is a question which can only be answered by each
          individual, rather than by independent investigators or the state.
       Given limited space, I will address only the first three of those concerns here.

2. Legacy of Abuse/Reinforcing Stereotypes

      Those concerned about race statistics point to the legacy of abuse of such data in Nazi-era
Europe and by some of the Communist regimes in power before 1989 in Central and Eastern
Europe. Indeed, the stereotyping of one pan-European minority – the Roma -- as an “asocial” or
“criminal” group formed the intellectual underpinning of the Porraimos, or Romani Holocaust,
during the Second World War. Indeed, there are some indications that controversial practices have
not been completely eliminated. Thus, the annual report for 2000 of the United Nations Special
Rapporteur on Racism noted information from Germany “that Sinti and Roma minorities are being
specially registered in the databases and records of the Bavarian police as Roma/Sinti type, gypsy
type or the old Nazi term Landfahrer (vagrant). The Central Council of German Sinti and Roma has
been informed of the report of the Bavarian Data Protection Commissioner of 16 December 1998
which states that Sinti and Roma are being registered generally on special police files without

   Among the many instances in which race statistics have been successfully used to demonstrate unjustifiable racial
disparities are two studies released this year – by, respectively, the United States government and Human Rights Watch
– which document pervasive and shocking differences of treatment accorded racial minorities in the United States
criminal justice system. See Fo x Butterfield, “Racist Disparities are Pervasive in Justice System, Report Say s,” The
New Yo rk Times, April 26, 2000 (describing report by U.S. Depart ment of Justice showing that “[b]lack and Hispanic
youths are treated more severely than white teenagers charged with comparable crimes at every step of the juvenile
justice system” and noting that, “for those charged with drug offenses, black youths are 48 times more likely than
whites to be sentenced to juvenile prison”); Human Rights Watch, Punishment and Prejudice: Racial Disparities in the
War on Drugs (June 2000) (finding, inter alia, that although five times as many whites use drugs as blacks, black drug
offenders are imp risoned at far higher rates than whites).
                                              Montreux, 4. – 8. 9. 2000
                                  Statistique, Développement et Droits de l‘Homme

reason or legal basis by their personal details and even the number plates of their cars and further
data. The police justify this storage as supposedly vorbeugende Verbrecenbekampfung (preventive
crime combat) and explain that Sinti and Roma could be a public danger.”27
       Others have claimed that customs officers have employed the practice of racial profiling to
identify and/or deport members of certain ethnic minorities. Czech Airlines was discovered to have
been identifying Roma with the letter “R” on their records, in order to assist this process of sifting
out “unwanted” migrants. 28
       A recent survey of media coverage of Roma in Slovakia reported several examples in which
race statistics were used to document an alleged Romani propensity to commit cr ime. Thus, one
cited article from the local press noted, “… [S]ince 1994 Romani criminality accounted for 22
percent of the total number of investigated offences…. Romani youth account for more than 30
percent of investigated criminal offences committed by young people.”29 And, “Of all morality-
breaking offenders Roma comprised 56 percent.” 30
       The legacy of misuse of ethnic and race-coded data is such that even actions apparently
motivated by the desire to assist ethnic minorities have fallen under the suspic ion of those they are
ostensibly aimed at helping. Thus, in November 1999, the Ministry of Education in the Czech
Republic sent a letter to elementary and special schools nationwide, asking that files be maintained
on the educational achievements of Roma students. The government said that special monitoring
was needed to improve the education of Roma. The reaction among some Roma leaders was hostile.
“If somebody wants to differentiate people according to the color of their skin, that‟s clear
discrimination, even if made for good purposes,” one Roma activist was reported to have
commented. “We have a certain historical experience with genocide, and have no guarantee
democracy will stay here forever. What if neo-Nazis take over?”31
       And yet, race statistics are not a necessary pre-condition to prejudice, discrimination, or the
fostering of negative stereotypes. Thus, for example, statistics are not needed for negligent and/or
consciously biased authors to selectively note and publish the ethnicity of persons arrested for
alleged criminal activity. 32 Nor are they required for political leaders to inflame racial prejudice
through insensitive and/or openly hostile public statements. 33

    E/CN.4/2000/ 16, 10 February 2000, para. 37. The UN Report goes on to note that, in December 1996, when a
question was raised in the German Bundestag about racial classifications in police records, the “Federal Govern ment
stated that … doing away with such classifications altogether does not come into consideration because of their
indispensable nature for police work….” Id., para. 39.
   Pro ject on Ethnic Relations, “Ro ma and the Law: Demythologizing the Gypsy C riminality Stereotype” (1999), p. 12.
   Praca, 25 August, 1998, p. 4, quoted in Slovak Helsinki Co mmittee, “Image of the Roma in Selected Slovak Media”
(2000), pp. 19-20.
   Id., p. 20.
   Associated Press, “Controversy flares over Gypsy lists in Czech schools,” 3 November, 1999.
   Numerous examples are cited in Slovak Helsinki Co mmittee, “Image of the Roma in Selected Slovak Media” (2000),
pp. 20-21.
   Thus, in early August 2000, a parliamentary deputy representing the Slovak National Party publicly declared that
"unadaptable Roma" must be placed in "reservations" to reduce the crime rate, CTK reported. The deputy reportedly
said that 1) the state must stop providing social benefits to "people who harm it," deeming the payment of such benefits
to Ro ma "inhu mane to the rest of the population;" 2) p lacing Ro ma in reservations would be "completely normal" as "in
America there are also reservations for the Indians;" and 3) if Slovakia does not place "unadaptable Ro ma" into
reservations now, "they will place us there 20 years fro m now." Three days later, the chairwo man of the Slovak
National Party (SNS) said that she saw no need to apologize for the deputy‟s comments, sinc e, in her view, the failure
to solve the "genuine Ro ma problem" suits those Roma activ ists who "should come up with some proposals instead of
strong words that solve nothing." The chairwo man added that "the SNS is interested in a thorough solution of the
problem of the Ro many ethnic group, because it is not the Gypsies, but the rest of Slovakia's population that is
discriminated against." RFE/ RL (Un )Civ il Societ ies, Vol. 1, No. 13, 10 August 2000.
                                               Montreux, 4. – 8. 9. 2000
                                   Statistique, Développement et Droits de l‘Homme

       Indeed, some purported abuses of “statistics” seem more made up than grounded in t he reality
of authentic data. 34 The alleged abuse is often more accurately a reflection of the absence of
statistics, which provides a vacuum to be exploited by creative, often malign, invention.
       Some observers have noted what they perceive is a “double sta ndard” with regard to race
statistics. Thus, on the one hand, when advocates of non-discrimination ask for data concerning,
inter alia, the number of Roma and non-Roma children in certain schools, or the number of ethnic
minorities represented in the police force, government officials not uncommonly claim that such
information does not exist. 35 On the other hand, this alleged non-existence of such data has not
prevented even officials representing governments with restrictive data protection laws from on
occasion citing to a broad range of race- or ethnic-coded statistics. 36 At their worst, law
enforcement agencies in some countries appear willing to use and/or publicize ethnic- and race-
coded data when it serves the purpose of highlighting what is alleged to be disproportionate
criminal activity on the part of certain ethnic minorities. Information provided by Romanian police
about a series of raids in the spring of 2000 was emblematic of this tendency: “In the last five days,
police have raided Bucharest‟s seediest areas, in a crackdown on violent crime. Of 2,000 people
arrested, police have said, most of them were Roma.” 37

3. Legal Restrictions on the Collection of Race- and Ethnic-Coded Data

      Some of the resistance to intensified collection and maintenance of race- and ethnic-coded
data is premised on the understanding that European regional law and the domestic law in a number
of countries prohibit or severely restrict such activities. A number of surveys of domestic legislation
presently underway will yield results later this year. For now, I limit myself to noting that regional
data protection regulations in Europe reasonably distinguish between individual, identifiable data
and collective, anonymous data in ways which safeguard privacy and do not impede the good- faith
collection and dissemination of race statistics for legitimate governmental, scientific, and/or public
interest objectives. 38

   Thus, the same Slovak National Party parliamentary deputy referred to in footnote 34, supra, was further quoted as
saying it has been "statistically proven" that most "retarded people" come fro m among the Roma and as asking, "What
is humane about morons being allowed to give birth to more morons and raise the pe rcentage of morons and crazies in
the nation?" See also Slovenska republika , 12 December, 1998, p. 3, quoted in Slovak Helsinki Co mmittee, “Image of
the Ro ma,” p. 23 (“There are crimes 95 percent of wh ich are committed by our fellow Ro mani citizens. It is not just
specific sexual crimes, incest or sex abuse of the under-aged”).
   Thus, the 1999 Progress Report of the European Commission on Hungary observes: “According to the Hungarian
Govern ment, identify ing the ethnicity of offenders is not allo wed under the data protection law and thus no statistical
evidence on discrimination is available.” (European Co mmission, “Regular Report fro m the Commission on Progress
on Accession,” 13 October 1999).
   At a meeting in Bucharest in January 2000 sponsored by the Embassy of Finland, a representative of the Hungarian
government‟s Office for National and Ethnic Minorities offered a range of statistics concerning the Roma. A mong the
data provided were the following: “some 60 percent of the Roma live in rural areas, abou t 30 percent in towns and the
remain ing 10 percent in Budapest”; “more than 70 percent of Ro ma children finish the eight classes of primary school”;
“[o]nly one-third of all Ro ma children attend secondary education”; the unemployment rate within the Ro ma
community “is about five times higher” than for others. Judith Solymosi, “Ro ma Policy in Hungary between 1998 –
2000,” in Embassy of Finland in Romania, International Expert Sy mposium on Roma Questions (28-29 January, 2000),
p. 98.
   Associated Press, “Gypsies march to protest racism,” 23 March, 2000. Indeed, racial stereotyping is not unique to
Europe. A United States police-training manual (B.H. Carter, Gypsies, Travelers and Thieves , South Carolina Criminal
Justice Academy, 1987) states that it is “structured to provide police personnel with Gypsy lifestyle and criminal
activities wh ich contribute to effective investigative and enforcement techniques.” The manual suggests how physical
characteristics may be used to identify “European Gypsies.” See Project on Ethnic Relations, Ro ma and the Law
(2000), p. 12.
   Thus, for example, the European Convention for the Protection of Individuals with Regard to Automatic Processing
of Personal Data (ETS 108) applies only to “personal data”, which is defined as “any informat ion relating to an
                                                Montreux, 4. – 8. 9. 2000
                                    Statistique, Développement et Droits de l‘Homme

      The above understanding of existing European law is bolstered by the practice of European
and international bodies in repeatedly encouraging governments to gather and provide race- and
ethnic-coded information necessary to measure compliance with international anti-discrimination

3.1 European Union

       The EU Race Directive, discussed above, offers perhaps the most recent and clearest
statement that race- and ethnic-coded statistics are essential tools in the combat of discrimination.
However, European Union organs have on previous occasions made clear that such information is,
not only permissible, but indeed desirable. Thus, the European Parliament‟s most recent Resolution
concerning racism in the context of accession “calls on the candidate countries to collect, as a basis
for policy action, reliable monitoring data on ethnic, linguistic and religious minority groups
including immigrants and refugees, the number and outcome of racist acts reported and prosecuted,
and the performance of minority groups in the economic and social spheres.” 39 The Regulation
establishing the European Union Monitoring Centre on Racism and Xenophobia authorized the
Centre, inter alia, to “collect, record and analyse information and data, including data resulting from
scientific research,” to “build up cooperation between the suppliers of information and develop a
policy for concerted use of their databases in order to foster … the wide distribution of their
information,” and to “develop methods to improve the comparability, objectivity and reliability of
data at Community level by establishing indicators and criteria that will improve the consistency of
information.”40 And the 1998 Annual Report of the European Union Monitoring Centre stated that,
in certain countries, such as the United Kingdom, “As far as the public authorities and public
opinion are concerned, the best means of combating racist and xenophobic acts and attitudes is to
start with a clear picture: in other words, to compile statistics based on the best possible criteria so
as to obtain a close-up view of the real nature of these phenomena. The aim is not to compare such

identified or identifiable individual.” See ETS 108, Arts. 2(a), 3(1). Moreover, this Convention expressly exempts from
the requirement of “additional safeguards” certain “automated personal data files used for statistics or for scientific
research purposes when there is obviously no risk of an infringement of the privacy of the data subjects.” ETS 108, Art.
9(3). The Convention‟s Explanatory Report notes that this exemption “leaves the possibility of restricting the exercise
of the data subjects‟ rights [to additional safeguards] with regard to data processing operations which pose no risk.
Examples are the use of data for statistical work, insofar as these data are presented in aggregate form and stripped of
their identifiers.” ETS 108, Explanatory Report, para. 59. Similarly, the Recommendation of the Committee of
Ministers to Member States Concerning the Protection of Personal Data Collected and Processed for Statistical
Purposes (No. R(97)18) (the “Reco mmendation”) distinguishes b etween “personal” and “anonymous” data (as to which
“identificat ion requires an unreasonable amount of time and manpower”) (Principle 1); applies to statistical results only
“to the extent that they permit identificat ion of data subjects” (Principle 2.1); allows for the processing of “sensitive
data” (including those relating to racial/ethnic origin) for “statistical purposes”, where the data subjects are not
“identifiab le” (Princip le 4.8); and provides that “an important public interest” may justify the co llection or processing of
“sensitive data”, even absent the consent of the data subject, where such consent would ordinarily be required (Principle
6.2). In this regard, the Exp lanatory Memorandum to the Recommendation notes that an exception to the consen t
requirement for the collection and processing of “sensitive data” for statistical purposes might be “justified by major
public interest, as where statistical information is needed to, [inter alia] … develop aid to social groups in difficulty.
Such examp les, to which many more might be added, relate to matters which affect society‟s essential interests and in
which the state has responsibilit ies.” (Explanatory Memorandum, para. 85(b)). It seems plausible that the governmental
obligation to eradicate racial discrimination might, under some circu mstances, qualify as such a “major public interest.”
   European Parliament, Report on the communication fro m the Co mmission countering racis m, xenophobia and anti-
semit ism in the candidate countries” (COM (1999)256-C5-0094/1999-1999/ 2099(COS)) (28 February, 2000), para. 13.
   Council Regulat ion (EC) No. 1035/ 97, of 2 June 1997, Arts. 2(a), 2(b), 2(f). The Regulation further notes that the
“Centre shall apply to its processing and exchange of data under this Regulation the provisions laid down in Directive
95/ 46/ EC [the European Union‟s data protection directive].” Id., Art. 5(1).
                                                  Montreux, 4. – 8. 9. 2000
                                    Statistique, Développement et Droits de l‘Homme

figures between different countries, but to be able to approach them on the basis of the same
definitions, something that has yet to be done.”41

3.2 European Commission against Racis m and Intole rance

       The European Commission against Racism and Intolerance (ECRI), a monitoring organ of the
Council of Europe, has consistently sought statistical information from governments on the effects
of government policies on different racial and/or ethnic groups.
       ECRI‟s General Policy Recommendation No. 1 on “Combating Racism, Xenophobia,
Antisemitism and Intolerance,” recommends that, “since it is difficult to develop and effectively
implement policies … without good data,” governments should “collect, in accordance with
European laws, regulations and recommendations on data protection and protection of privacy,
where and when appropriate, data which will assist in assessing and evaluating the situation and
experiences of groups which are particularly vulnerable to racism, xenophobia, anti-Semitism and
       A subsequent ECRI recommendation “stress[es] that statistical data on racist and
discriminatory acts and on the situation of minority groups in all fields of life are vital for the
identification of problems and the formulation of policies,” and notes ECRI‟s “convi[ction] that
such statistical data should be supplemented by data on attitudes, opinions and perceptions.” 43 The
Recommendation goes further to “recommend” that “governments of member States … take steps
to ensure that national surveys on the experience and perception of racism and discrimination from
the point of view of potential victims are organised,” and sets forth guidelines for the conduct of
such surveys, which note the importance of “[g]ood population statistics including information
about variables such as place of birth, ethnic origin, religious confession, mother tongue,
citizenship, etc.”44
       ECRI‟s Country-by-Country Reports have specifically suggested that authorities gather
statistical information on racial/ethnic minorities for the formulation and evaluation of government
policies. In considering Poland, ECRI noted: “There is a system of data collection on ethnic and
national groups in Poland, based on individual declarations of persons belonging to minority
groups. However, most sources indicate that it is impossible to determine accurately the size of any
ethnic or national group in Poland, since post World War II censi have not included questions
pertaining to ethnic identity. Without accurate and up to date statistics in this field, it is impossible
to draw up suitable policies or evaluate their effectiveness. Therefore, an absolute priority must be
the establishment of a reliable system of data-collection in this field, in accordance with European
laws, regulations and recommendations on data-protection and protection of privacy, in order to
determine the real figures of ethnic and national minorities, immigrant groups, etc.” 45 Reports on
other countries have evinced similar concerns. 46

   Eu ropean Monitoring Centre on Racism and Xenophobia, “Annual Report 1998,” pp. 11 -12.
   ECRI, General Po licy Reco mmendation No. 1, 4 October 1996.
   ECRI, General Po licy Reco mmendation No. 4, 6 March 1998.
   Id., Appendix, para. 6.
   ECRI, “ECRI‟s Country-by-Country Approach: Volu me 1”, 1997, p. 86.
    See, e.g., Report on France (ECRI, “ECRI‟s Country -by-Country Approach: Volu me III”, 1998, pp. 23, 28)
(highlighting “the need to improve statistical records in the areas of discrimination, racial harassment and the size of
minority groups” and observing that “[i]t is difficult to obtain an exact evaluation of the size of the different ethnic
groups in France, because, in accordance with the Republican ideal where every national is a citizen, there is officially
no categorisation of ethnic or racial groups in the statistics. The main categories used are thus „foreigners‟ and
„immig rants.‟ It is therefore difficult to collect data in order to analyse discrimination based on race, colour or ethnic
origin. Possible ways of dealing with this gap in statistical informat ion might be considered”); Report on the
Netherlands (ECRI, “ECRI‟s Country-by-Country Approach: Vo lu me III”, 1998, p. 47) (“Despite the widespread
collection of statistics in a number of areas, there seems to be a lack of reliab le, harmon ised and comparable statistics as
                                                 Montreux, 4. – 8. 9. 2000
                                   Statistique, Développement et Droits de l‘Homme

       Finally, ECRI has urged even those countries whose data protection legislation purportedly
prohibits the dissemination of race or ethnic statistics to pursue alternative avenues for the
collection of such data. Thus, its Report on the Czech Republic advises: “The Czech authorities
state that data on demographic composition other than census results cannot be published in the
Czech Republic with regard to legislation designed to protect personal data and privacy. It is
suggested that steps should be taken to improve information on the Roma/Gypsy community at the
level of local authorities, research institutions and non-governmental organisations in order to
facilitate the planning of social policies in relation to the Roma/Gypsy community.” 47 And even
Hungary, whose data protection legislation is among the most restrictive in Europe, is urged to
“work towards establishing a system of collection of data and information in accordance with”
European standards, “bearing in mind the importance of accurate and up-to-date statistics for
drawing up policies and evaluating their effectiveness.”48

3.3 United Nations Committee on the Elimination of Racial Discrimination

       In response to widespread government refusals to provide demographic information broken
down by race or ethnicity, the United Nations Committee on the Elimination of Racia l
Discrimination (CERD) has gone so far as to make clear, in its “General Guidelines Regarding the
Form and Contents of Reports to be Submitted by States Parties under Article 9, para. 1, of the
Convention,” the following: “8. The ethnic characteristics of the country are of particular
importance in connection with the International Convention on the Elimination of All Forms of
Racial Discrimination. Many States consider that, when conducting a census, they should not draw
attention to factors like race lest this reinforce divisions they wish to overcome. If progress in
eliminating discrimination based on race, colour, descent, national and ethnic origin is to be

regards the situation of minority groups in all areas of social an d economic activ ity…. Further efforts should be made to
collect more co mparab le and reliable statistics, using a standard national form of categorisation of ethnic origin as the
basis of all relevant studies, after full consultation about its acceptability and in full accordance with European laws,
regulations and recommendations on data protection and protection of privacy”); Report on Slovakia (ECRI, “ECRI‟s
Country-by-Country Approach: Volu me III”, 1998, p. 64) (“As a rule, members of minority groups are now identified
as a result of their voluntary declaration under the census. Changes in the statistics on the situation or minority groups
over the past ten years cannot be considered reliable because some types of problems have not been studied in a
continuous fashion and the data on various trends have not been systematically collected and published. Furthermo re,
people belonging to a minority or religious group are often reluctant to register as such for fear of encountering
discrimination and harassment. Lastly, the inadequate amount and scope of statistics on minority groups is partly due to
the simple fact that the members of those groups, especially those of the Roma/ Gypsy community, do not behave in a
uniform manner when it co mes to registering. It is hoped that a reliable system of data collect ion will be developed in
accordance with European standards and recommendations on data protection and the protection of privacy”); Report
on Russia (ECRI, “ECRI‟s Country-by-Country Approach: Vo lu me IV”, 1999, p. 41) (noting that existing data “do not
include statistics on the condition of ethnic minorities or the extent to which they are affected by social and economic
policies – for examp le, in the shape of breakdowns of the experience of the various ethnic minorit ies in respect of
emp loyment, housing, health, social services, education and law and order. Experience in other countries has
demonstrated the usefulness of such statistics in monitoring the relative position of minorit ies in society and in
constituting policies wh ich might affect them”); Report on Spain (ECRI, “ECRI‟s Country -by-Country Approach:
Vo lu me IV”, 1999, p. 75) (“One statistical area which should be improved is data concerning the Roma/ Gypsy
community. Consideration should be given to various ways of obtaining information about the size and situation of the
Ro ma/ Gypsy community at national and regional/local level”); Report on United Kingdom (ECRI, “ECRI‟s Country -
by-Country Approach: Volu me IV”, 1999, p. 89) (endorsing as “valuable and imp ortant” the collection of informat ion
“on ethnic origin as such,” including a question in the 1991 census on “respondents‟ ethnic origin”); Report on Austria
(ECRI, “ECRI‟s Country-by-Country Approach, Volu me V”, 1999, p. 17) (noting the absence of reliable data about
minority groups, and the consequent difficulty of developing policies “without good data,” and recommending the
establishment of a “reliable system of data collection,” in accordance with Eu ropean standards).
   ECRI, “ECRI‟s Country-by-Country Approach, Volu me I”, p. 17.
   Id., p. 44.
                                                Montreux, 4. – 8. 9. 2000
                                     Statistique, Développement et Droits de l‘Homme

monitored, some indication is needed of the number of persons who could be treated less favourab ly
on the basis of these characteristics. States which do not collect information on these characteristics
in their censuses are therefore requested to provide information on mother tongues … as indicative
of ethnic differences, together with any information about race, colour, descent, national and ethnic
origins derived from social surveys.”
       Similarly, CERD General Recommendation IV (1973) expressly endorses the use of race
statistics by inviting “States parties to endeavour to include in their reports under article 9 relevant
information on the demographic composition of the population.” These general guidelines have
been frequently reinforced by CERD requests to individual governments for statistical information
on the racial/ethnic composition of the population. 49
       The CERD‟s interest in obtaining race statistics – and the invalidity of commonly deployed
government justifications for the failure to collect race statistics -- were highlighted in a recent
dialogue with the Portuguese government from 1999. In response to Committee requests for
demographic information broken down by race or ethnicity, the government of Portugal maintained
a) that its Constitution “prohibited the conducting of surveys on the racial or ethnic component of
the population,” b) that the United Nations Population Commission had recommended that the
category of “race” should be optional in government censi, and c) that Portugal‟s “multiracial
tradition and the absence of racial prejudices” made it “improper to establish statistics in terms of
the race, religion … of the persons concerned.”50
       The Committee rejected all these arguments, observing that the UN Population Commission
recommendation that the category of “race” should be optional in censi “concerned only the

    See, e.g., CERD Concluding Observations on Italy, 18 March 1999, CERD/ C/54/Misc.32.Rev.3, para. 16 (“The
Co mmittee reco mmends that the State party include in its next report statistical data on the ethnic compo sition of the
country”); CERD Concluding Observations on Czech Republic, 30 March 1998, CERD/C/ 304/Add.47, para. 18 (“The
Co mmittee reco mmends that the State party provide, in its next report, mo re specific statistical data on minority
representation in local, regional and State administrations as well as informat ion on their situation in the fields of
education, employ ment and health. The Co mmittee would also welco me mo re data on minorities‟ po lit ical, economic
and cultural rights. Information is also requested concerning the proportions of ethnic minority groups and aliens
residing in the country”); CERD Concluding Observations on Poland, 15 October 1997, CERD/C/ 304/Add.36, para. 16
(suggesting that “the State Party take all appropriate measures to compile more precise informat ion” with regard to
“statistical information on minorit ies”, and “include such data in the next periodic report”); CERD Conclud ing
Observations on the Former Yugoslav Republic of Macedonia, 15 October 1997, CERD/C/ 304/Add.38, para. 6 (noting
“with concern” the “lack of statistical information on the invocation of the various protections against racial
discrimination”); CERD Concluding Observations on Ukraine, 30 March 1998, CERD/C/ 304/Add.48, para. 8
(expressing “concern” at “the inadequacy of demographic data on the different ethnic groups living in the State party”);
CERD Concluding Observations on Hungary, 28 March 1996, CERD/C/ 304/Add.4, para. 16 (“the absence of
demographic data on the minorities in different districts of the count ry makes any evaluation of activities intended for
their benefit difficu lt. Equally, the lack of data on the representation of minorit ies in the local authorities and the lack of
recent data on the situation of minorit ies in the fields of education, culture, the media and emp loyment is regretted”);
id., para. 22 (reco mmending “that the State party provide in its next report, statistical data on the minorities in the
different districts, on their representation in the local authorities, as well as recent dat a on their situation in the fields of
education, culture, the media and emp loyment”); CERD Concluding Observations on Spain, 28 March 1996,
CERD/C/ 304/Add.8 (“the Co mmittee stresses the necessity for comp lete and up -to-date statistical data to be included in
the next report on the exact ethnic composition of the Spanish population and on the socio -economic characteristics of
each ethnic group”); CERD Concluding Observations on Romania, 22 September 1995, A/50/18, para. 275
(reco mmending “that the Government systematically collect data on foreigners resident in Ro mania”); CERD
Concluding Observations on Bulgaria, 23 April 1997, CERD/ C/304/Add.29, para. 14 (“Adequate indicators and other
means of monitoring the economic and social living conditions of [the Roma] should be developed. The Committee
requests the State party to provide detailed information on such measures in its next report. The Co mmittee also
recommends that the State party provide, in the next report, such statistical data and informat ion as are available on the
situation of all minorities” with respect to equal protection of the laws).
   Su mmary Record of the 1311th Meeting of the CERD, Consideration of Portugal, 4 March 1999, CERD/ C/SR.1311,
para. 14; Su mmary Record of the 1312th Meeting of the CERD, Consideration of Portugal, 4 March 1999,
CERD/C/Sr.1312, para. 5.
                                                   Montreux, 4. – 8. 9. 2000
                                   Statistique, Développement et Droits de l‘Homme

methodology of State census-taking. None of the optional categories could be a legal barrier for a
State to gather information on the demographic composition of its population.” 51 The Committee
concluded by urging that Portugal “in its next periodic report provide detailed and relevant
information on the demographic composition of the Portuguese population….” 52

3.4 Other United Nations Organs

       Although they have had less occasion to address the question, other United Nations
monitoring bodies have also asked governments to p rovide race- or ethnic-coded information. In its
Concluding Observations of 19 November 1998, the UN Committee Against Torture requested that
“Hungary should include in its next periodic report all relevant statistics, data and information on:
a) the number of complaints about ill treatment; the proportion they represent in relation to the total
number of cases investigated and, in particular, the proportion of Roma complaints, detainees and
prisoners….”. 53 During its consideration of Romania, the United Nations Human Rights Committee
“members … wanted to know the number of complaints lodged by Roma alleging mistreatment by
authorities or private individuals.”54 And on more than one occasion, the UN Committee on the
Rights of the Child (CRC) has asked for data specifically concerning ethnic minorities. 55

3.5 Other Bodies

      The recent “Report on the situation of Roma and Sinti in the OSCE Area” of the OSCE High
Commissioner on National Minorities of the Organization for Security and Cooperation in Europe
highlighted the need for race- and ethnic-coded data: “Statistical data on the ethnic composition of
populations is an important tool for establishing patterns of discrimination, for facilitating efforts to
enforce legal prohibitions of discrimination and for assessing the efficacy of anti-discrimination and
other policies aimed at improving the conditions of Roma.” 56
      In calling for the adoption of a directive covering non-discrimination and equal treatment, the
Final Report of the Project on the European Union and Human Rights – the principle reference
document used by the Comite des Sages in the preparation of their Human Rights Agenda for the
EU for the Year 2000 – suggested that such a provision should “require employers to monitor the

   Su mmary Record o f the 1311th Meeting of the CERD, supra, para. 14.
   CERD Concluding Observations, 8 April 1999, CERD/ C/304/Add.67, para. 15.
   CAT Concluding Observations, 19 Nove mber 1998, A/54/ 55, para. 85.
   United Nations Press Release HR/CT/ 99/ 18, 20 July 1999.
   See, e.g., Concluding Observations on Bulgaria, 24 January 1997, CRC/C/ 15/Add.66, para. 9 (exp ressing “concern[]
about the need to strengthen the State party‟s capacity to collect and process data to evaluate progress achieved and to
assess the impact of policies adopted on children, in part icular the most vulnerable groups of children”); id., para. 23
(reco mmending that “the State party give priority give attention to the development of a system of data collection and to
the identification of appropriate disaggregated indicators with a view to addressing all areas of the Convention and all
groups of children in society. Such mechanisms … can be used as a basis for designing programmes to imp rove the
situation of children, particularly those belonging to the most disadvantaged groups, including … children belonging to
minority groups, especially Ro ma”); Concluding Observations on Czech Republic, 27 October 1997,
CRC/C/ 15/Add.81, para. 10 (expressing “concern [] about the need to strengthen the State party‟s limited capacity to
develop specific disaggregated indicators to evaluate progress achieved and assess the impact of existing policies on all
children, in particular children belonging to minority groups”); id., para. 29 (reco mmending that priority be given to
identification of “appropriate disaggregated indicators” to address this need); Concluding Observations on Hungary, 5
June 1998, CRC/ C/15/Add.87, para. 9 (express ing “concern[] at the lack of disaggregated statistical data covering all
   Organization for Security and Cooperation in Europe, High Co mmissioner for National M inorit ies, Report on the
Situation of Ro ma and Sinti in the OSCE Area, 10 March 2000, text at note 179.
                                                Montreux, 4. – 8. 9. 2000
                                   Statistique, Développement et Droits de l‘Homme

composition of the workforce in terms of gender, race and disability to establish a workplace equal
opportunities policy.” 57
      In short, while more research on the legal parameters is needed, there appears to be broad
agreement among international and European expert bodies that, with the proper safeguards in place
to protect against abuse, race and ethnic statistics are an essential component of any effective anti-
discrimination policy.

4. Concluding Thoughts

       The history of Europe in the twentieth century casts a long shadow over any discussion of
race statistics and complicates efforts to extricate the issue of numbers from broader questions of
race, ethnicity, and justice. And too, the subjective nature of racial and ethnic identity makes it
difficult to achieve consensus on its contours. Indeed, “[i]t is one of the special features of ethnic
identity that its form and expression can change and be modified. Ethnic identities alter under the
influence of historical events, politics, the media, education and succeeding generations, a tendency
that is encouraged by migration, mixed marriages, assimilation processes, etc.” 58 One expert group
has gone so far as to suggest, “Minority statistics constitute a great illusion. Under the appearance
of scientific certainty and mathematical precision, they conceal, advertently or inadvertently, a
world of differing degrees and conceptions of identity, diverging definitions, and unstable
classifications.”59 Thus, even well- intentioned efforts to count the numbers of ethnic minorities will
face a number of obstacles, including the fear of some minority members to self- identify; the desire
to identify with another, more powerful ethnic group; the incentive of some to artificially inflate or
deflate counts for political purposes; and, in a number of countries, the poor quality of public
databases produced by poverty and/or decades of neglect. These persistent difficulties have been
complicated by political factors in much of Central and Eastern Europe.
       And still, the reality of enduring racism, and the need for knowledge – cold, hard facts -- to
struggle against it, mandate persistence. Even if race and ethnicity are socially and historically
constructed, they are living concepts which continue to affect (and distort) the distribution of
resources and power in many contemporary societies. Hence, the desire to identify as “human” and
no more – though admirable to some – does not erase the existence of prejudice or the need to
measure its effects. To ignore racial categories – to stop asking about race – is no defence to racism.
       At the same time, it is, of course, not the data themselves which are value- laden, but their
potential use or misuse by government authorities and others. Thus, the mere existence of statistics
showing that, to take a hypothetical example, fifty percent of all the prison inmates in country X are
members of minority group Y, even though minority Y forms only ten percent of the nationwide
population, is not prejudicial or racist. Controversy arises in the interpretation of this data, and t he
conclusions to be drawn therefrom. Thus, some may view the hypothetical data as evidence of the
disproportionate criminality of the Y minority. To others, however, the same statistics constitute
definitive proof that the criminal justice system in country X operates in a discriminatory manner.
Often, debates which appear to be about race statistics are really about more fundamental,
independent questions of racial prejudice and political power.

   P. Alston and J.H.H. Weiler, “The European Union and Human Rights: Final Project Report on an Agenda for the
Year 2000” (1998), para. 205.
   “Council of Europe Demographics Study,” p. 15. See id., p. 14 (The concepts of “national or ethnic groups or
communit ies” are “mult idimensional: cultural, historical and territorial. The co mp lexity of the terms and the lack o f a
precise and generally recognised definition mean that it is difficu lt to make the concepts operational”).
   Robert Schuman Centre for Advanced Studies, European University Institute, and Forward Studies Unit, European
Co mmission, Final Report of the Reflection on the Long-Term Implications of EU Enlargement: the Nature of the New
Border (1999), p. 81.
                                                Montreux, 4. – 8. 9. 2000
                                   Statistique, Développement et Droits de l‘Homme

      In seeking to develop policies capable of, on the one hand, generating reliable race- and
ethnic-coded data, and, on the other, safeguarding individual privacy, the following considerations
should be taken into account:
       Distinguish between race statistics – which are social indicators – and the social and
          political values which affect how they are interpreted and used.
       Distinguish between, on the one hand, hate speech and its contribution to the perpetuation
          of racial stereotypes, and, on the other, the misuse of race statistics. Develop effective
          means of addressing the former, including the promulgation and enforcement of codes of
          conduct for police and other government authorities in labeling, referring to, and
          characterizing ethnic minority groups.
       Separate the issue of crime statistics from other statistics. For a variety of reasons, race-
          coded crime statistics often raise far more controversy than analogous statistics in the
          fields of employment, education or political participation. Although race-coded crime
          statistics are essential to document, and/or correc t, systematic racial bias in criminal
          justice systems, the frequent misuse of race statistics in the criminal justice field in
          particular, warrants a particularized approach.
       Distinguish between, on the one hand, the individual‟s right to choose and express his/her
          own identity and, on the other, collective race- or ethnic-coded data which, if generated
          and maintained in accordance with international data protection procedures, are not
          traceable to any individual. 60 Thus, it should be possible for a school to maintain data
          showing that, say, 30 percent of the students are Roma, without divulging the identity of
          any single student reflected in the data.
       Involve members of all affected minority groups in the design, conduct and analysis of
          data collection from the very beginning of the process.
       Do not confuse the means and the end. Race statistics are needed not to satisfy an ideal
          notion of truth or science, but to assist in the achievement of very real, concrete goals:
          measuring and eliminating discrimination.
       Recall that, necessary as they are, race statistics are not a substitute for other methods to
          tackle discriminatory practices, including forthright political leadership, the adoption and
          implementation of comprehensive anti-discrimination legislation, and public education
          about the extent of racism and the need to overcome it.

   In this regard, see, e.g., Explanatory Memorandu m, Council of Europe Reco mmendation No. R(97)18 concerning the
Protection of Personal Data Co llected and Processed for Statistical Purposes, para. 10 (“It is therefore necessary to
make a fundamental distinction between the individual use of personal data and their collect ive use. This is a
differentiation of purpose, which is essential fro m the standpoint of the protection of individuals, their privacy and their
rights and freedoms”) (emphasis in original).
                                                 Montreux, 4. – 8. 9. 2000

To top