Content uploaded by Julie Ringelheim
Author content
All content in this area was uploaded by Julie Ringelheim on Sep 25, 2018
Content may be subject to copyright.
Ethnic Pro¢ling: A Rising Challenge for European
Human Rights Law
Olivier De Schutter
n
and Julie Ringelheim
nn
Ethnic pro¢ling, de¢ned as the use of racial, ethnic or religious background as a determining
criterion for the adoption of law enforcement decisions, has been rising signi¢cantly in Europe,
in particular in the wake of the terrorist attacks of 11 September 2001. This article examines
whether European human rights law is well equipped to deal with this challenge, and if not,
how it should be reformed. Against the widely held assumption that personal data protection
legislation is insu⁄ciently protective of ‘sensitive’ data relating to race or ethnicity, it explains
instead why combating ethnic pro¢ling has been made more di⁄cult, rather than less, by an
overly protective reading of the requirements of data protection laws. It then discusses the addi-
tional measures that European states could take to address more e¡ectively the human rights con-
cerns prompted by the development of ethnic pro¢ling.
In 1999, Mr Timishev, a Chechen lawyer living in Nalchik, in the Kabardino-
Balkaria Republic of the Russian Federation, travelled by car from the Ingushetia
Republic to Nalchik.When reaching the administrative border of the Kabardino-
Balkaria Republic, his car was stopped at a checkpoint and entry was refused to
him: tra⁄c police o⁄cers had received an oral instruction from the Ministry of
the Interior of Kabardino-Balkaria Republic not to admit persons of Chechen
ethnic origin.The Nalchik Town Court dismissed MrTimishev’s complaint that
this was discriminatory: in its view, the order was aimed at preventing the pene-
tration into towns and villages of persons having terrorist or antisocial aspirations.
International litigation followed. Five years after Mr. Timishev ¢led an applica-
tion against Russia, the European Court of Human Rights found that Russian
o⁄cers had violated the non discrimination provision of Article 14 ECHR in
combination with the freedom of movement guaranteed in Article 2 of Protocol
n14.The order, which barred passage to any person of Chechen ethnicity or per-
ceived as such,‘represented a clear inequality of treatment in the enjoyment of the
right to libertyof movement on account of one’s ethnic origin’.
1
In July 2001, in response to an in£ux ofasylum seekers of Czech nationality, the
vast majority of whomwere of Roma ethnic background, the United Kingdom
organised an operation at Prague Airport. Following an agreement with the
Czech Republic, British immigration o⁄cers were authorised to give or refuse
leave to enter the country to passengers before they boarded £ights to the UK.
n
Professor of Law, Universityof Louvain (UCL) and College of Europe (Natolin).
nn
Senior Fellow, National Fund for Scienti¢c Research, Centre for Legal Philosophy at theUniversity
of Louvain (UCL). The authors are grateful to Daniel Moeckli for provision of information and to the
two anonymous reviewers for comments.
1Timishev vRussia Eur Ct HR (2nd section), App Nos 55762/00 and 55974/00, judgment of 13
December 2005 (¢nal on 13 March 2006) y54.
r2008 The Authors. Journal Compilation r2008 The Modern Law Review Limited.
Published by Blackwell Publishing, 9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA
(2008) 71(3) 358^384
The avowed aim of the exercise was to bar fromentry travellers who intended to
visit the country for a purpose not within the Immigration Rules, including those
who wished to claim asylum in the UK. In practice, immigration o⁄cers operat-
ing at Prague Airport had a di¡erent attitude depending on whether passengers
appeared to be of a Roma background: Roma were subjected to longer and more
intrusive questioning, they were required to provide proof of matters whichwere
taken on trust from non-Roma, and a large proportion of them were refused leave
to enter. In e¡ect, the overwhelming majority of those who were barred from
entering the country were Roma. On 9 December 2004, the House of Lords held
that British immigration o⁄cers operating at Prague Airport had discriminated
on racial grounds, contrary to section 1(1)(a) of the Race Relations Act 1976,
against Roma seeking to travel from that airport to the United Kingdom, by
treating them more sceptically than non-Roma when determining whether to
grant them leave to enter the country.
2
In the wake of the 11 September 2001 terrorist attacks against New York and
Washington, the German authorities, in an attempt to identify ‘sleepers’ of terror-
ist organisations, decided to resort to the so-called Rasterfahndung method, ie the
screening by the police of personal data sets of public or private bodies in order to
track individuals presenting suspects’ features. The criteria established at the
national level for this operation included being male, Muslim, national ofor born
in one of 26 listed countries with predominantly Muslim population, a current or
former student, and legal resident in Germany. Numerous institutions, including
universities, employers, health and social insurance agencies, were required to
provide to the police the personal records of all individuals corresponding to the
de¢ned pro¢le. The operation did not result in any arrest or criminal charge for
terrorism-related o¡ences.
3
On 4 April 2006, the Federal Constitutional Court
ruled that the Rasterfahndung was in breach of the individual’s fundamental right
of self-determination over personal information (Article 2(1) and 1 of the Grund-
gesetz) and therefore was unconstitutional.
4
All these incidents have in common that in all three cases, persons were singled
out by law enforcement o⁄cers not because of their individual behaviour but
rather because of their ethnicity or religion. Ethnic or religious background was
used as a determining criterion for law enforcement decisions. O⁄cers assumed
that there was a correlation between membership in a particular ethnic or reli-
gious group and propensity to commit certain crimes or act in a way considered
undesirable. These incidents are but a few instances of a rising phenomenon in
Europe, that of ‘ethnic pro¢ling’. Increasingly, certain individuals come to be
2R (European Roma Rights Centre) vImmigration O⁄cer at PragueAirport (United Nations High Commis-
sioner for Refugees intervening) [2004] UKHL 55.
3 See D. Moeckli,‘Discrimination Pro¢les: Law Enforcement After 9/11and 7/7’ (2005) 5 European
Human Rights Law Review 517; D. Moeckli,‘Terrorist pro¢ling and the importance of a proactive
approach to human rights protection’ (16 December 2006), available at the Social Science
Research Network (SSRN): http://ssrn.com/abstract=952163 (last visited 18 February 2008);
J. Goldston, ‘Ethnic Pro¢li ng and Counter-Terrorism: Trends, Dangers and Alternatives’ (2006),
at 5^6. Available at http://www.soros.org/resources/articles_publications/articles/counterterrorism_
20060606 (last visited 18 February 2008).
4 Decision of 4 April 2006 (1BvR 518/02) (2006) 59 NeueJuristischeWochenschrift1939.
Olivier De Schutter and Julie Ringelheim
359
r2008 The Authors. Journal Compilation r2008 The Modern Law Review Limited.
(2008) 71(3) 358^384
targeted in law enforcement activities solely or partly on the ground of their
ethnicity or religion. Certainly, suspicious or discriminatory attitudes of the
police towards ethnic or religious minorities have long been observed in Eur-
opean countries as elsewhere. However, in the context of the ‘war on terror’ and
the continuous reinforcement of the ¢ght against illegal immigration, the in£u-
ence of ethnicity and religion on law enforcement activities has reached new pro-
portions, raising new concerns about the risks of interference with fundamental
rights of people belonging to certain ethnic or religious minorities ^ who, more
often than not, already su¡er from disadvantage and stigmatisation.
Ethnic or ‘racial’ pro¢ling has been widely studied and debated in the United
States since the 1990s. In Europe, by contrast, re£ection on this question and its
legal implications is just emerging. European and domestic case law is still scarce
and sometimes contradictory.This article aims to shed light on the legal challenges
raised by the development of ‘ethnic pro¢ling’ for European human rights law
and to re£ect on the adequate responses to this phenomenon.We argue that ethnic
pro¢ling is a form of discrimination prohibited under European human rights
law, whether or not a statistically signi¢cant relationship could be established
between membership in a particular racial or ethnic group and certain forms of
criminal behaviour. We also question the widely held assumption that personal
data protection legislation is insu⁄ciently protective of ‘sensitive’ data relating to
race or ethnicity. Instead, we explain why combating ethnic pro¢ling has been
made more di⁄cult, rather than less, by an overly protective reading of the
requirements of data protection laws. In order to arrive at these conclusions, we
re£ect, ¢rst, on the notion of ‘ethnic pro¢ling’itself. Second, we analyse the legal
framework stemming from present-day European human rights law, on the basis
of which the legality of ethnic pro¢ling practices must be assessed.Three bodies
of laws are essential in this regard: antidiscrimination law, personal data protec-
tion, and the regulation of police stop and search powers.The main legal sources
taken into account are European Community law and the law of the European
Convention on Human Rights. In addition, we consider the case-law of certain
domestic high courts. Third, we ask whether this framework provides su⁄cient
protection against potential human rights violations resulting from the targeting
of individuals on the basis of their ethnicity or religion in law enforcement deci-
sions.We conclude that it does not. The ¢nal section, therefore, will discuss the
additional measures that European states could take to address more e¡ectively
the human rights concerns prompted by the development of ethnic pro¢ling.
THE NOTION OF ETHNIC PROFILING
The debate on ethnic or ‘racial’ pro¢ling ¢rst arose in the United States. The
phrase ‘racial pro¢ling’, together with the more colloquial expression ‘driving
while black’,
5
was popularized in the1990s to describe the police practice of stop-
5 ‘Driving while black’ is an ironic allusion to the crime of ‘driving while intoxicated’. See D. Har-
ris,‘When Success Breeds Attack:The Coming Backlash Against Racial Pro¢ling Studies’ (2001)
6Michigan Journal of Race and Law 237, 237.
Ethnic Pro¢ling: A Rising Challenge for European Human Rights Law
360 r2008 The Authors. Journal Compilation r2008 The Modern Law Review Limited.
(2008) 71(3) 358^384
ping African American or Hispanic drivers in disproportionate numbers, under
the pretext of a minor tra⁄c infraction in order to look for evidence of other
crimes, in particular drug tra⁄cking. Such practice rested on the assumption that
these drivers were more likely than white motorists to be involved in criminal
activities.
6
The term then came to be used more generally to refer to the in£uence
of racial or ethnic factors in law enforcement decisions, whether in common stop
and search practices, anti-terrorism or other areas.
7
Still, some confusion surrounds this notion. A few words about the concept of
‘pro¢ling’ itself will help clarify the issue.‘Criminal pro¢ling’ is an investigative
method designed to facilitate the identi¢cation of criminals: it aims at determin-
ing the physical, behavioural or psychological characteristics of the perpetrator of
a speci¢c crime or of a group of crimes, and thus at constructing a practical
description of the criminal, on the basis of an analysis of hosts of elements, includ-
ing the nature of the o¡ence, the manner in which it was committed, as well as
information gathered on perpetrators of similar crimes in the past.
8
This techni-
que, developed initially in the US, was especially used from the1970s onwards in
investigations concerning serial killers. In the1980s, it was decided to apply it to a
broader category of crime, namely drug o¡ences. Law enforcement authorities
endeavoured in particular to develop pro¢les of ‘drug couriers’, ie people involved
in the transportation of drugs, in order to help o⁄cers deciding whom to consider
as suspects and stop on the road or on the street.
9
This new application of pro¢ling
marked a signi¢cant change compared to its use in relation to serial killers: while
in the latter case, pro¢les are descriptive of a particular criminal and based on ana-
lysis of a speci¢c crime which has already been committed and discovered by the police, in
the context of drug dealing or transporting, pro¢les are designed for proactive
detection of o¡ences as yet unknown tothepolice;theyaremeanttobepredictive of
crime.
10
More recently, the technique of pro¢ling has started to be applied in the
area of counter-terrorism. As noted by the Special Rapporteuron the protection
of human rights while countering terrorism, ‘in recent years, so-called terrorist
pro¢ling has become an increasingly signi¢cant component of States’ counter-
terrorism e¡orts.’
11
Signi¢cantly, the European Union has recommended to
members states to construct ‘terrorist pro¢les’ on the basis of characteristics such
as nationality, age, education, birthplace,‘psycho-sociological characteristics’, or
family situation, with a view to identifying terrorists before the execution of ter-
rorist acts as well as to prevent or reveal the presence of terrorists in their territory,
6 K. Meeks, DrivingWhile Black ^ WhatTo Do if You Are aVictim of Racial Pro¢ling (NewYork: Broad-
way Books, 2000) 4^7; D. Harris,‘Law Enforcement Stake in Coming to Grips with Racial Pro-
¢ling’ (2001) 3 Rutgers Race and Law Review 9, 9^10.
7 K. K. Russell,‘Racial Pro¢ling: A Status Report of the Legal, Legislative and Empirical Litera-
ture’ (2001) 3 Rutgers Race and Law Review 61.
8 D. A. Harris, Pro¢les in Injustice ^ Why Racial Pro¢ling Cannot Work (New York: The New Press,
2002) 16^18.
9ibid 18^26; R. S. Engel J. M. Calnon and T. J. Bernard,‘Theory and Racial Pro¢ling: Shortcom-
ings and Future Directions in Research’ (2002) 19 Justice Quarterly 249, 250^251.
10 Harris, Pro¢les in Injustice, n 8 above,19^20.
11 Report of the Special Rapporteur on the promotion and protection of human rights and funda-
mental freedoms while countering terrorism, Martin Scheinin, 29 January 2007, A/HRC/4/26, y
32 (hereinafter 29 January 2007 Report).
Olivier De Schutter and Julie Ringelheim
361
r2008 The Authors. Journal Compilation r2008 The Modern Law Review Limited.
(2008) 71(3) 358^384
in cooperation with the immigration services and the police.
12
Here too, pro¢les
have a predictive function: unlike serial killer pro¢les, they are not based on a
speci¢c, existing crime, but are designed to permit the discovery of those who
may be involved in future crimes, even before they commit a criminal act.Thus,
like most predictive pro¢les, they are based on generalisations about groups of
people.
13
When one of a pro¢le’s components is ethnicity or a related criterion, ie skin
colour, national origin or religion,‘ethnic pro¢ling’ can be said to occur.
14
‘Ethnic
pro¢ling’ is not necessarily a formalised policy: it may be based on a pro¢le for-
mally established by competent authorities (formal ethnic pro¢ling), but it may also
take the form of a de facto, informal, practice, based on o⁄cers’subjective memories
of signi¢cant experiences or assumptions about the typical features of o¡enders
(informal ethnic pro¢ling).
15
When a pro¢le is formally de¢ned, it may be used in
twoways bylaw enforcement o⁄cers: depending onthe circumstances, they will
seek to identify people meeting these characteristics either by automatic means,
through the screening of a vast number of individuals’ personal data, or directly
on the ground, on the basis of identity checks or visual observation.
Some authors or organisations de¢ne ‘ethnic pro¢ling’ in a way that suggests
that it would be limited to cases where ethnicity, skin colour, religion, or national
origin, are the only criteria taken into account for considering a person as suspect.
The American Civil LibertiesUnion, for instance, understands‘racial pro¢ling’as
‘any police or private security practice in which a person is treated as a suspect
because of his or her race, ethnicity, nationality, or religion. This occurs when
police investigate, stop, frisk, search, or use force against a person based on such
characteristics instead of evidence of a person’s criminal behaviour’.
16
Similarly,
before Canadian courts, racial pro¢ling has been described as
criminal pro¢ling based on race. Racial or colour pro¢ling refers to that phenom-
enon whereby certain criminal activity is attributed to an identi¢ed group in
society on the basis of race or colour resulting in the targeting of individual mem-
bers of that group. In this context, race is illegitimately used as a proxy for the crim-
inality or general criminal propensity of an entire racial group.
17
12 JHA Council of 28 and 29 November 2002, Council of the EU doc 14817/02 (press 875), Annex II,
21. For the recommendation itself, which is not mentioned in the summaryof the conclusions, see
Council of the EU doc. 11/11858/02.‘Terrorist pro¢le’ is de¢ned in this document as ‘a set of phy-
sical, psychological or behavioural variables, which have been identi¢ed as typical of persons
involved in terrorist activities and which may have some predictive value in that respect’.
13 D. Moeckli (2006), n 3 above,6.
14 ibid 6^7; Harris, n 8 above,11; Russell, n 7 above, 68.
15 On formal and informal pro¢ling, see Harris, ibid 26^30. See also J. D. Castiglione,‘Reconceptua-
lizing ‘‘Racial Pro¢ling’’ ’ 1 November 2006. Available at Social Science Research Network
Library http://papers.ssrn.com/sol3/papers.cfm?abstract_id=942774 (last visited 18 February 2008).
16 The Union adds that ‘racial pro¢ling’ . . .‘often involves the stopping and searching of people of
color for tra⁄c violations, known as‘DWB’ or ‘driving while black or brown’. Although normally
associated with African Americans and Latinos, racial pro¢ling and ‘DWB’ have also become
shorthand phrases for police stops of Asians, Native Americans, and, increasingly after 9/11, Arabs,
Muslims and South Asians’. See http://www.aclu.org/racialjustice/racialpro¢ling/index.html (last
visited 18 February 20 08).
17 RvRichards (1999), 26 CR (5th) 286 (Ont CA).This de¢ nition was proposed by the African Cana-
dian Legal Clinic before the Ontario Court of Appeals.
Ethnic Pro¢ling: A Rising Challenge for European Human Rights Law
362 r2008 The Authors. Journal Compilation r2008 The Modern Law Review Limited.
(2008) 71(3) 358^384
To be sure, treating a person as a suspect solely because of his or her ethnicity or
religion is discriminatory. However, whenethnic, racial, national, or religious back-
ground constitutes only one component of a pro¢le and is used in combination
with other factors bylaw enforcement o⁄cers when deciding to stop, search, arrest,
or put under surveillance a person, the risk of discrimination is no less present.
18
Indeed, the principle of non-discrimination requires that only in exceptional cir-
cumstances should the race or ethnicity, the religion, or the nationality of a person
in£uence the decision about how to treat or not to treat that person.
19
For the
purpose of this article, therefore,‘ethnic pro¢ling’ will be understood as the prac-
tice of using‘race’ or ethnic origin, religion, or national origin, as either the sole
factor, or one of several factors, in lawenforcement decision, on a systematic basis,
whether or not concerned individuals are identi¢ed by automatic means.
Whether formal or informal, based on the use of automatic means of proces-
sing information or not, ‘ethnic pro¢ling’ presupposes classifying people, impli-
citly or explicitly, along ethnic or related criteria.Yet, it is important to stress that
classifying individuals on the basis of such factors is not necessarily unlawful or
discriminatory. Such classi¢cation may serve a variety of purposes, including
monitoring the behaviour of public agents in order to ensure that they do not
commit direct or indirect discrimination. Similarly, the processing of sensitive
data that such procedures entail may or may not be compatible with legislation
on personal data protection: the legalityof data processing depends on the objec-
tives pursued and the means used. Importantly, as will be argued in the last sec-
tion, the collection of data on the ethnic background of those who are stopped,
searched, or arrested, may be a crucial tool to unmask discriminatory practices of
law enforcement o⁄cers, especially when they remain informal.
20
THE EXISTING LEGAL FRAMEWORK
Present-day European and international lawcontain no legally binding provision
dealing explicitly with the concept of ‘ethnic pro¢ling’. However, several human
rights guarantees, enshrined in European or international instruments, contribute
to de¢ning the legal framework against which the legality of such practices or
procedures must be assessed. Depending on the circumstances, ethnic pro¢ling
is likely to con£ict either with the right to privacy and personal data protection,
or with rules regulating the exercise by law enforcement o⁄cers of their powers,
in particular in stop-and-search procedures or in identity checks. In all cases, it
involves a di¡erential treatment based, in whole or in part, on ethnic or religious
criteria, thereby bringing intoplay antidiscrimination norms.We argue, however,
that the prohibition of ethnic pro¢ling as a speci¢c form of discrimination will
remain ine¡ective in a number of EU Member States as long as they adhere to an
overly rigid understanding of the requirements of data protection legislation,
18 See Russell, n 7 above, 65^68; Castiglione, n15 above, 30^32.
19 E.U. Network of Independent Experts on Fundamental Rights, Opinion No. 4 ‘Ethnic Pro¢ling’
(December 2006), at 10. Available at: http://ec.europa.eu/justice_home/cfr_cdf/index_en.htm (last
visited 18 February 2008).
20 ibid 9^12.
Olivier De Schutter and Julie Ringelheim
363
r2008 The Authors. Journal Compilation r2008 The Modern Law Review Limited.
(2008) 71(3) 358^384
which may result in an obstacle to the monitoring of the behaviour of law enfor-
cement authorities.
It is signi¢cant in this regard that, in its recent General Policy Recommenda-
tion N111 on combating racism and racial discrimination in policing (the recom-
mendation),
21
the Council of Europe Commission against Racism and
Intolerance (ECRI) does not mention the need to reinforce data protection
against the risks of ‘racial pro¢ling’,
22
but that it insists, rather, on the need to
‘monitor police activities in order to identify racial pro¢ling practices, including
by collecting data broken down by grounds such as national or ethnic origin, lan-
guage, religion and nationality in respect of relevant police activities’.
23
In addi-
tion, the General Policy Recommendation n111encourages the Member States to
‘introduce a reasonable suspicion standard, whereby powers relating to control,
surveillance or investigation activities can only be exercised on the basis of a sus-
picion that is founded on objective criteria’.
24
While they build on the existing
rules of international and European human rights law, these recommendations
make explicit, in addition, that the prohibition of ethnic pro¢ling as a form of
discrimination will only be e¡ective if certain supplementary conditions are
met, which the existing national rules do not always satisfy. We explore in this
section what these conditions are, and how, by taking them into account, we can
better combat the practice of ethnic pro¢ling. It is with this aim in mind that we
examine the content of the non-discrimination requirement, the regulationof the
power of law enforcement authorities to arrest individuals for the purpose of
identity searches, and data protection legislation in the EU Member States. In
relating these di¡erent bodies of law to the practice of ethnic pro¢ling, we seek
both to identify potential gaps in the existing legal framework and ^ where pro-
mising developments did take place under certain jurisdictions ^ where inspira-
tion may be found to identify solutions for these gaps.
Anti-discrimination law
Articles 2 and 26 of the International Covenant on Civil and Political Rights,
Protocol n112 to the European Convention on Human Rights (ECHR), as well
as the International Convention on the Elimination of All Forms of Racial Dis-
crimination (ICERD) all formulate a general prohibition of discrimination. Arti-
cle 14 of the ECHR also precludes discrimination based, inter alia,onrace,
religion, national origin, or association with a national minority, in the enjoyment
of the rights and freedoms set forth in the Convention. ICERD speci¢cally pro-
hibits racial discrimination in the enjoyment of ‘the right to security of person
and protection by the State against violence or bodily harm, whether in£icted
21 Adopted on 29 June 2007.
22 The ECRI de¢nes ‘racial pro¢ling’ as:‘The use by the police, with no objective and reasonable
justi¢cation, of grounds such as race, colour, language, religion, nationality or national or ethnic
origin in control, surveillance or investigation activities’. It recommends that the Member States
clearly de¢ne and prohibit racial pro¢ling by law.
23 Para 2 of the recommendation.
24 Para 3 of the recommendation.
Ethnic Pro¢ling: A Rising Challenge for European Human Rights Law
364 r2008 The Authors. Journal Compilation r2008 The Modern Law Review Limited.
(2008) 71(3) 358^384
by government o⁄cials or by any individual group or institution’ (article 5(b)),
‘the right to freedom of movement and residence within the border of the State’,
and ‘the right to leave any country, including one’s own, and to return to one’s
country.’ (article 5(d) i and ii).The Committee for the Elimination of Racial Dis-
crimination, tasked with monitoring this Convention, has explicitly stated that
racial or ethnic pro¢ling can result in a discrimination prohibited by the
ICERD.
25
In its General Recommendation No 31on the prevention of racial dis-
crimination in the administration and functioning of the criminal justice system
(2005), the Committee highlights an increase in the risks of discrimination in the
administration and functioning of the criminal justice system‘partly as a result of
the rise in immigration and population movements, which have prompted preju-
dice and feelings of xenophobia or intolerance among certain sections of the
population and certain law enforcement o⁄cials, and partly as a result of the
security policies and anti-terrorism measures adopted by many States, which
among other things have encouraged the emergence of anti-Arab or anti-Muslim
feelings, or, as a reaction, anti-Semitic feelings, in a number of countries’. To
counter this phenomenon, the Committee recommends that states take measures
‘to prevent questioning, arrests and searches which are in reality based solely on
the physical appearance of a person, that person’s colour or features or membership
of a racial or ethnic group, or any pro¢ling which exposes him or her to greater
suspicion.’
26
The question of ethnic pro¢ling has recently been the focus of renewed atten-
tion at international level. In his 2007 report, Martin Scheinin, the UN Special
Rapporteur on the promotion and protection of human rights and fundamental
freedoms, while countering terrorism, undertakes to assess whether ethnic pro¢l-
ing practiced in the area of counter-terrorism is compatible with the principle of
equal treatment. His analysis is based on the criteria set out by the anti-discrimi-
nation jurisprudence of the UN Human Rights Committee and the ICERD but
also of the European Court of Human Rights (ECtHR). According to the latter’s
well-known case-law, a di¡erence in treatment constitutes a discrimination either
when it lacks objective and reasonable justi¢cation, or when there is no reasonable
relationship of proportionality between the means employed and the legitimate
aim sought to be realised.
27
Applying these notions to pro¢ling in the context of
counterterrorism, the Special Rapporteur ¢rst asserts that insofar as it aims at pre-
venting terrorist attacks, the use of terrorist pro¢les based inter alia on ethnicity,
25 Committee on the Elimination of Racial Discrimination, Concluding Observations: Canada,UN
Doc CERD C/61/CO/3, para 338 (2002) (calling on Canada to ensure that applicationof Canada’s
Anti-Terrorism Act ‘does not lead to negative consequences for ethnic and religious groups,
migrants, asylum seekers and refugees, in particularas a result of racial pro¢ling’).
26 Committee o n the Elimination of Racial Discrimi nation, General recommendation No 31 on the
prevention of racial discrimination in the administration and functioning of the criminal justice
system (2005) para 20. Already in its General Recommendation No 30 on discrimination against
non-citizens (2004), the Committee recommends that states‘ensure that any measures taken in the
¢ght against terrorism do not discriminate, in purpose or e¡ect, on the grounds of race, colour,
descent, or national or ethnic origin and that non-citizens are not subjected to racial or ethnic
pro¢ling or stereotyping’ (para 10).
27 Eur Ct HR, Case ‘Relating to Certain Aspects of the Laws on the Use of Languages in Education in Belgium’
(Belgian Linguistic Case) (Appl nos 1474/62 et al) Judgment of 14 July 1968, 1EHRR 252 y10.
Olivier De Schutter and Julie Ringelheim
365
r2008 The Authors. Journal Compilation r2008 The Modern Law Review Limited.
(2008) 71(3) 358^384
national origin or religion, can be said to pursue a legitimate and compelling
social need.
28
But the bulk of his analysis consists in determining whether such
measure is proportionate to the objective pursued. Roughly said, the notion of
proportionality, as understood by the ECtHR, requires that a fair balance be pre-
served between the respect for individual rights and the protection of the public
interests at stake. This entails, in particular, that the means used are suitable to
advance the attainment of the objectives pursued,
29
and that among various
means to achieve these goals, the measure that entails the least restriction of rights
and freedoms is adopted.
30
In this regard, the Special Rapporteur observes that
terrorist pro¢les based oncharacteristics such as ethnicity, national origin and reli-
gion are regularly inaccurate and both over- and under-inclusive. Inpractice, most
terrorist pro¢les use ethnic appearance and national origin as proxies for religion,
typically Muslim religion, given that religious background is usually not readily
identi¢able. But ethnicityand national origin are poor proxies for religion: many
of those targeted on the groundof their physical appearance will not be Muslims.
And among those who are indeed Muslims, the overwhelming majority are in no
way involved in terrorism. Terrorist pro¢les that rely on this type of criteria will
thus a¡ect a great number of individuals who have nothing to dowith terrorism.
Further, by increasing the number of people considered as suspects by the police,
they divert important law-enforcement resources from more e¡ective work. At
the same time, terrorist pro¢les are under-inclusive: there are potential terrorists
who do not ¢t the pre-determined criteria and will therefore escape the attention
of law-enforcement agents working on the basis of such pro¢le. For the Special
Rapporteur, evidence suggests that ethnic or religious pro¢ling practices are ‘an
unsuitable and ine¡ective, and therefore a disproportionate means of countering
terrorism: they a¡ect thousands of innocent people, without producing concrete
results’.
31
In addition to its lack of e¡ectiveness, this method also entails adverse
e¡ects: by singling out certain people on the basis of their real or supposed ethni-
city, national origin, or religion in counter-terrorism actions, the authorities foster
stigmatisation of these groups of people which, in turn, is likely to increase their
feeling of alienation and induce deep mistrust of the police.
32
All these elements
together convince the Special Rapporteur that terrorist pro¢ling basedon ethnic,
religious, or national origin factors is a disproportionate measure to combat ter-
rorism, and therefore constitutes discrimination.
Somewhat paradoxically, the European Court of Human Rights itself proved
hesitant to condemn as discriminatory law enforcement actions that could be
deemed instances of ‘ethnic pro¢ling’, at least when it was confronted with an in-
formal practice, rather than a formalised policy. Cisse
ŁvFrance is a case in point. A
group of some two hundred illegal migrants, most of whomwere of African ori-
gin, had been occupying the St Bernard’s Church in Paris in June 1996.They re-
ceived the support of several human-rights organisations, some of whose activists
28 29 January 2007 Report, para 46.
29 Eur Ct HR, Observer a nd Guardian vUK (Appl no 13585/88) Judgment of 29 November 1991 y69.
30 Eur Ct HR, Informationsverein Lentia vAustria (Appl nos 13914/88 et al) Judgment of 24 November
1993 yy 39 and 43. See also Moeckli (2005), n 3 above, 525.
31 29 January 2007 Report, y54.
32 Id., yy 56^58.
Ethnic Pro¢ling: A Rising Challenge for European Human Rights Law
366 r2008 The Authors. Journal Compilation r2008 The Modern Law Review Limited.
(2008) 71(3) 358^384
decided to sleep on the premises in a demonstration of solidarity with their pre-
dicament. When, on 23 August 1996, the police executed an order for the total
evacuation of the premises, they set up a checkpoint at the church exit to verify
whether the aliens evacuated from the church had documentation authorising
them to stay and circulate in the territory. As described in the judgement of the
Court:
All the occupants of the church were stopped and questioned.Whites were imme-
diately released while the police assembled all the dark-skinned occupants, apart
from those on hunger strike, and sent them bycoach to an aliens’ detention centre
at Vi ncennes. Orders were made for the detention and deportation of almost all of
those concerned. More than a hundred were subsequently released by the courts on
a account of certain irregularities on the part of the police, which even extended to
making false reports regarding the stopping and questioning procedure.
33
This is a euphemistic description. In fact, it would seem that, as explained by the
applicant,‘only ‘‘blacks’’ had had their identities checked while ‘‘whites’’ had been
shown to the exit’, without the legality of their administrative situation being
checked.
34
In spite of this, the submission that the arrests were discriminatory
because the identity veri¢cations had allegedly been decided on the basis of the
skin colour of the individuals concerned, was dismissed at the admissibility stage.
The European Court contents itself with noting that: ‘the system set up at the
church exit for checking identities was intended to ascertain the identity of per-
sons suspected of being illegal immigrants. In these circumstances, it cannot con-
clude that the applicant was subjected todiscriminationbased on race or colour’.
35
This conclusion appears to follow from the choice of the Court, following the
submissions of the French government, to frame the issue as concerning exclu-
sively the deprivation of liberty of the individual applicant, without considering,
more broadly, how the screening was practiced by the French police on all the
occupants of the church.
The Court took a comparable stance a few months later in C
onka vBelgium.
36
At stake was an operation organised by the Belgian police to arrest dozens of
families of asylum seekers from Slovakia whose claim for asylum had been dis-
missed, in order to expel them from the country. Even though those targeted by
this operation were exclusively of a Roma ethnic background, the allegation of
discrimination was held inadmissible, on the ground that people subject to the
expulsion measure had not been selectedon the basis of their nationality or ethnic
origin, but rather on the basis of their belonging to a speci¢c immigration chan-
nel. Nevertheless, in its judgment, adopted by the shortest of margins (4 judges
against 3), one Chamber of the Court found Belgium guilty of having proceeded
33 Eur Ct HR (2nd sect), Cisse
ŁvFrance (Appl No 51346/99) Judgment of 9 April 2002 (¢nal on 9 July
2002), y13.
34 Eur Ct HR (2nd sect), Cisse
ŁvFrance (Appl No 51346/99), admissibility decision of 16 January
2001. This presentation of the facts by the applicant, while it does not correspond precisely to
the description given by the Court, was notcontradicted in the procedure by the defending State.
35 ibid.
36 Eur Ct HR (3d sect), C
onka and others, the Liguedes droits de l’homme vBelgium (Appl No 51564/99)
admissibility deci sion of 13 March 2001.
Olivier De Schutter and Julie Ringelheim
367
r2008 The Authors. Journal Compilation r2008 The Modern Law Review Limited.
(2008) 71(3) 358^384
to a collective expulsion of aliens, prohibited by article 4 of Protocol No 4, as the
contested measure had not been taken on the basis of a reasonable and objective
examination of the particular case of each individual alien of the group.
37
The
procedure, therefore, did not ‘a¡ord su⁄cient guarantees demonstrating that the
personal circumstances of each of those concerned had been genuinely and indi-
vidually taken into account.’
38
Yet, as we have seen, in the later case ofTimishev vRussia, decided on 13 Decem-
ber 2005, the Court took a much ¢rmer position. Here, the Court found that by
refusing entry to the Kabardino-Balkaria Republic to MrTimishev because of his
ethnic origin, Russian police o⁄cers violated the non-discrimination provision of
Article 14 ECHR, in combination with the freedom of movement guaranteed in
Article 2 of Protocol n14.The order issued by senior police o⁄cers, which barred
the passage of any person who actually was of Chechen ethnicity or was perceived
as such, represented a clear inequality of treatment in the enjoyment of the right
to liberty of movement, as no other ethnic communities were subject to similar
restriction. For the Court, discrimination on account of one’s actual or perceived
ethnicity has to be considered a form of racial discrimination; it is thus‘a particu-
larly invidious kind of discrimination’, one which‘in view of its perilous conse-
quences, requires from the authorities special vigilance and avigorous reaction.’
39
The Court further observes:
The Government did not o¡er any justi¢cation for the di¡erence in treatment
between persons of Chechen and non-Chechen ethnic origin in the enjoyment of
their right to liberty of movement. In any event, the Court considers that no di¡er-
ence in treatment which is based exclusively or to a decisive extent on a person’s
ethnic origin is capable of being objectively justi¢ed in a contemporary democratic
society built on the principles of pluralism and respect for di¡erent cultures.
40
In this important passage, the Court seems to exclude that anydi¡erence in treat-
ment on account of ethnicitycould be justi¢ed, where this leads to imposing dis-
advantages on the basis of this criterion alone or to a decisive extent.This entails that
when ethnicity is one among several factors that prompt law enforcement o⁄cials
to target a person, the measure will be held discriminatory if the ethnic back-
ground had a decisive (although not necessarily exclusive) in£uence on the deci-
sion-making process. Besides, while the Court in this case could dispense with
determining whether the measure was justi¢ed by a legitimate aim and propor-
tionate to it, since the government did not advance any justi¢cation for the o⁄-
cials’ decision, it nonetheless takes this opportunity to a⁄rm in unambiguous
terms that a di¡erence in treatment based on a person’s ethnic origin can never
be justi¢ed if it is detrimental to this person.This suggests that the proportionality
test becomes irrelevant in this context: treating individuals similarly situated dif-
ferently according to their real or supposed ethnicity should be considered as hav-
37 Eur Ct HR (3d sect), C
onka vBelgium (Appl No 51564/99) Judgment 5 February 2002 y59.
38 ibid y63.
39 Eur Ct HR (2nd sect), Timish ev vRussia (Appl Nos 55762/00 and 55974/00) Judgment of 13
December 2005 (¢nal on 13 March 2006) yy 54^57.
40 ibid y58.
Ethnic Pro¢ling: A Rising Challenge for European Human Rights Law
368 r2008 The Authors. Journal Compilation r2008 The Modern Law Review Limited.
(2008) 71(3) 358^384
ing such serious consequences in stigmatising ethnic minorities; feeding stereo-
types, prejudice, and exclusion; and creating divisiveness and resentment, that dif-
ferential treatment on this ground should in principle be deemed unlawful under
any circumstances.
41
The fundamental con£ict between ethnic pro¢ling and the core democratic
values of equality and respect for the individual, is well expressed by Baroness
Hale of Richmond, in R (European Roma Rights Centre) vImmigration O⁄cer at Pra-
gueAirport,
42
where the House of Lords held that immigration o⁄cers operating at
Prague had discriminated on racial grounds against Roma seeking to travel to the
United Kingdom by treating them more sceptically than non-Roma:
The whole point of the law is to require suppliers to treat each person as an indivi-
dual, not as a member of a group.The individual should not be assumed to hold the
characteristics which the supplier associates with the group, whether or not most
members of the group do indeed have such characteristics, a process sometimes
referred to as stereotyping.
43
Thus, instead of considering Roma travellers as inherently suspects:
Immigration o⁄cials should have been given careful instructions in how to treat all
would-be passengers in the same way, only subjecting them to more intrusive ques-
tioning if there was speci¢c reason to suspect their intentions fromthe answers they
had given to standard questions which were put to everyone. The system operated
by immigration o⁄cers at Prague Airport was inherently and systemically discrimi-
natory on racial grounds.
44
What should be noted at this stage is the contrast between the prohibition of eth-
nic pro¢ling as a form of discrimination on the one hand ^ as in the leading case
of Timishev ^ and, on the other hand, the considerable di⁄culties victims face
when having to prove any discrimination which they suspect has taken place ^
as in Cisse
Łor Conka. It is in this regard noteworthy that, in a recent judgment
where it was asked to ¢nd a violation of the non-discrimination requirement
with the right to education, the Grand Chamber of the European Court of
Human Rights has taken the view that‘when it comes to assessing the impact of
a measure or practice on an individual or group, statistics which appear on critical
examination to be reliable and signi¢cant will be su⁄cient to constitute the
prima facie evidence the applicant is required to produce’, resulting in shifting
the burden of proof to the respondent State, which ^ if presented with such
41 EU Networkof Independent Experts on Fundamental Rights, n19 above at11.
42 R (European Roma Rights Centre) vImmigration O⁄cer at Prague Airport (United Nations High Commis-
sioner for Refugees intervening) [2004] UKHL 55.
43 ibid at [74].This point is restated by Lord Hope of Craighead in GillanvCommissionerof Policefor the
Metropolis [2006] UKHL 12: ‘The whole point of making it unlawful for a public authority to
discriminate on racial grounds is that impressions about the behaviour of some individuals of a
racial group may not be true of the group as a whole. Discrimination on racial grounds is unlaw-
ful whether or not, in any given case, the assumptions on which it was based turn out to be justi-
¢ed (y43).
44 ibid Baroness Hale of Richmond, at [97].
Olivier De Schutter and Julie Ringelheim
369
r2008 The Authors. Journal Compilation r2008 The Modern Law Review Limited.
(2008) 71(3) 358^384
statistics ^ would then have to show that the di¡erence in treatment is not discri-
minatory.
45
However, for victims of discrimination to use this possibility, they
must have statistics to rely upon.The collection of data relating to the behaviour
of police o⁄cers, allowing identi¢cation, in particular, of the impacts of their
activities on di¡erent racial or ethnic groups, may therefore progressively be seen
as a condition for the e¡ective enforcement of the anti-discrimination norm itself.
The European Commission against Racism and Intolerance therefore recom-
mends the collection ofdata allowing the measurement of such impacts:‘For data
broken down by grounds such as national or ethnic origin, language, religion and
nationality tobe used to identifyand measure racial pro¢ling, such data should be
collected in respect of relevant police activities, including identity checks, vehicle
inspections, personal searches, home/premisessearches and raids. Data should also
be collected on the ¢nal results of these activities (in terms of prosecutions and
convictions) so as to be able to assess whether the ratio between checks carried
out and actual convictions is anydi¡erent for members of minority groups com-
pared to the rest of the population’.
46
We return below to the importance of such
data collection for an e¡ective policy aiming at combating ethnic pro¢ling.
47
One ¢nal remark may be in order. In contrast with the UN and Council of
Europe instruments mentioned above, EU law has not yet addressed the issue of
ethnic pro¢ling. This may seem paradoxical, since, after the 1997 Amsterdam
Treaty inserted the new article 13 in the EC Treaty, the ¢ght against discrimina-
tion based inter alia on ethnic origin and religion has become a priority of the
European Union. The two Equality directives adopted in 2000 on the basis of
article 13 have led all EU member states to improve sign¢cantly their anti-discri-
mination legislation. While the Racial Equality Directive (Council Directive
2000/43/EC
48
), prohibits discrimination based on racial or ethnic origin in various
areas (employment and occupation, social protection, social advantages, educa-
tion, and access to goods and services available to the public), the Employment
Equality Directive (Council Directive 2000/78/EC
49
) prohibits discrimination in
employment and occupation on various grounds, including religion or belief.
Each directive applies to both the public and private sectors, including public
bodies. However, their scope of application ratione materiae is limited and excludes
policing functions exercised by the law enforcement authorities.
50
Hence, these
45 Eur Ct HR (GC), D. H. a n d Oth ers vCzech Republic(Appl no 57325/00) judgment of 13 November
2007 yy 189^190. This judgment builds on earlier case -law in which the Court had recognized the
need to allow fora shifting of the burden of proof in anti-discrimination cases where circumstan-
cial evidence allows for the establishment of a presumption of discrimination (Eur Ct HR (GC)
Nachova and Others vBulgaria (Appl n143577/98 and 43579/98) judgment of 6 July 2005 yy 147 a nd
157); and where the Court had agreed to identify indirect discrimination on the basis of statistical
evidence (Hoogendijk vthe Netherlands (dec) no 58461/00, 6 January 2005).
46 Explanatory Memorandum to General Policy Recommendation n111, at [42].
47 See section headed: Strengthening the Means to Counter Ethnic Pro¢ling, below.
48 Council Directive 2000/43/EC of 29 June 2000 implementing the principle of equal treatment
between persons irrespective of racial orethnic origin (OJ L 180 of 19 July 2000 at 22).
49 Council Directive 2000/78/EC of 27 November 2000 establishing a general framework for equal
treatment in employment and occupation (OJ L 303 of 2 December 2000 at 16).
50 This restriction to the scope of application of these instruments seems to follow from their legal
basis in Article 13 of the EC Treaty. This provision empowers the Council of the European Union
Ethnic Pro¢ling: A Rising Challenge for European Human Rights Law
370 r2008 The Authors. Journal Compilation r2008 The Modern Law Review Limited.
(2008) 71(3) 358^384
directives have no bearing on the question of ethnic pro¢ling in the context of
law enforcement activities.
The regulation of stop and search powers
As we have seen, General Policy Recommendation n111 adopted by the European
Commission against Racism and Intolerance recommends in paragraph 3 the intro-
duction of a ‘reasonable suspicion’ standard, according to which ‘powers relating to
control, surveillance or investigation activities can only be exercised on the basis of
a suspicion that is founded on objective criteria’.
51
If followed, this recommendation
would have a signi¢cant impact on many instances of ethnic pro¢ling as documen-
ted in existing studies; and it would bring about a culture change in the practices of
law enforcement authorities. In many European states, law enforcement o⁄cers
enjoy a large discretion in the ful¢lment of duties such as identity checks and ‘stop-
and-search’ arrests.This situation obviously increases the risk of discriminatory pro-
¢ling being practiced byo⁄cers acting on the ground insofar as this discretion could
enable them to target disproportionately or arbitrarily persons of certain ethnic
appearances when exercising their powers. Only the power to arrest individuals is
explicitly regulated by the European Convention on Human Rights: under Article
5(1), it must be subject to strict safeguards. No one can be deprived of his liberty save
in certain well-de¢ned situations and ‘in accordance with a procedure prescribed by
law.’ The ECtHR has recently insisted that ‘where deprivation of liberty is con-
cerned, it is particularly important that the general principle of legal certainty be
satis¢ed’.
52
This entails that the conditions for the deprivation of liberty must be
clearly de¢ned under domestic law, and that the law itself must be foreseeable in its
application.The standard of ‘lawfulness’set by the Convention requires that the law
be su⁄ciently accessible and precise to allow people to reasonably foresee the conse-
quences a given action may yield. Furthermore, an essential element of the lawful-
ness of a detentionwithin the meaning of Article 5(1) is the absence of arbitrariness.
53
In a remarkable decision dated 23 March 2006,
54
the Constitutional Court of
Slovenia developed the view that similar requirements apply to the exercise by
to legislate in order to combat discrimination on the grounds listed ‘within the limits of the
powers conferred by it upon the Community’ (Art 13(1) of the EC Treaty).
51 This is inspired by the European Code of Police Ethics, which provides that‘[p]olice investiga-
tions shall, as a minimum, be based upon reasonable suspicion of an actual or possible o¡ence or
crime’ (para 47). The European Code of Police Ethics was the subject of Recommendation
(2001)10 of the Committee of Ministers of the Council of Europe, adopted on 19 September
2001 at the 765th meeting of Ministers’ Deputies. See also, for a similar recommendation, EU
Networkof Indepe ndent Experts on Fundamental Rights, Opinion No 4 ‘Ethnic Pro¢ling’ (Decem-
ber 2006) at 26, stating that the EU Member States should‘de¢ ne with the greatest clarity possible
the conditions under which law enforcement authorities may exercise their powers in areas such
as identity checks or stop-and-search procedures’.
52 Eur Ct HR (2nd section) Enhorn vSweden (Appl No 56529/00) Judgment of 25 January 2005 y36.
53 ibid; Eur Ct HR (Grand Chamber) Chahal vUnited Kingdom (22414/93) Judgment of 15 November
199 6 y118.
54 Judgment U-I-152/03.The text of the judgment, which includes the opinion of the Ombudsman,
is available in English on the website of the Slovenian Constitutional Court: http://www.us-rs.si/
en/index.php (last vi sited 18 February 2008).
Olivier De Schutter and Julie Ringelheim
371
r2008 The Authors. Journal Compilation r2008 The Modern Law Review Limited.
(2008) 71(3) 358^384
the police of its power to stop individuals for the purpose of checking their iden-
tity. The case originated in a petition of the Ombudsman for Human Rights,
who challenged aprovision of the Police Act (Zpol) allowing policemen to estab-
lish the identity of a person‘who by his or her behaviour, conduct, and appearance
or by being situated in a certain place orat a certain time arouses the suspicion that
they will perpetrate, are perpetrating, or have perpetrated a minor or criminal
o¡ence . . .’.
55
The Ombudsman argued that when the police oblige a person to
showher personal identity documents, they interfere with her freedom of move-
ment and rights to privacy. Hence, such power must conform to the principle of
legal certainty. The view of the Ombudsman was that:‘Police authority must . . .
be de¢ned precisely in the law such that the police o⁄cer and the individual
clearly understand what the conditions are on the basis of which the use of police
authority is allowed for the establishment of identity’. For the Ombudsman, the
impugned provision did not meet this condition. Its formulation was so general
and imprecise that it enabled the police touse their power in an arbitrary manner.
In particular, the rule that a person could be suspected merely for reasons of his or
her appearance could easilygive rise to abuse by the police.
The Court fully endorsed the position of the Ombudsman: it held that the
statute did not determine with su⁄cient precision when an individual circum-
stance can constitute a su⁄cient basis to give rise to suspicion. Consequently,
the provision did not permit clearly discerning the limits between admissible
and inadmissible conduct of state authorities. In practice, this indeterminacy
permitted the police authority to establish identity without justi¢cation.
56
The Court, therefore, declared the provision unconstitutional and ordered the
legislature to remedy the unconstitutionality within a year of the publication of
its decision by specifying the criteria which enable the police to suspect that a
person will ‘perpetrate, is perpetrating, or has perpetrated a minor or criminal
o¡ence’, in particular when the suspicion is based on a person’s appearance or his
or her being situated in a certain place.
57
This case-law thus o¡ers a reading of the
requirements of the principleof legality in the exercise of the powers of the police
to stop a person which closelycorresponds tothat recommended by the European
Commission against Racism and Intolerance when, citing the 2001 European
Code of Police Ethics, it insists on the introduction of a ‘reasonable suspicion’
standard, prior to a person being stopped, searched, put under surveillance or
investigated.
Personal data protection
In the context of counterterrorism strategies, police and intelligence services are
increasingly resorting to data mining techniques, ie the computerised analysis of
vast databases of personal information in order to identify people whose data
55 Article 35.1of the Police Act (O⁄cial Gazette RS, Nos 49/98, 66/98 ^ corr, 93/01, 56/02, 79/03,43/04
^ o¡ consol text, 50/04,102/04 ^ o¡ consol text, 14/05 ^ corr o¡ consol text, 53/05, 70/05 ^ o¡ consol
text, 98/05 and 3/06 ^ o¡ co nsol text) (SPol).
56 Judgment U-I-152/03 at [15].
57 ibid at [18].
Ethnic Pro¢ling: A Rising Challenge for European Human Rights Law
372 r2008 The Authors. Journal Compilation r2008 The Modern Law Review Limited.
(2008) 71(3) 358^384
match a pre-de¢ned suspects’ pro¢le.
58
Such operations, which entail the
collection, analysis and storage of information on a large number of individuals
by public authorities, raise many concerns as to the dangers of misuse, manipula-
tion, or unwanted disclosure of the data processed. In the case of antiterrorism
these concerns are heightened by the fact that terrorist pro¢les typically
include criteria related to religion and national origin, which are especially
sensitive data.
In all EU countries, however, the legislation imposes restrictions on the cir-
cumstances and the way in which personal data can be processed by states’ autho-
rities. These rules, designed to protect the individuals’ private life vis-a
'-vis the
processing of personal data concerning them, provide critical safeguards against
ethnic pro¢ling carried out through automatic means. At the European level,
the main piece of EU legislation regarding personal data protection is Directive
95/46/EC of 24 October 1995 on the protection of individuals with regard to
the processing of personal data and on the free movement of such data.
59
This
instrument, however, does not cover processing operations concerning public
security, defence, State security and the activities of the State in areas of criminal
law.
60
But the 1981Council of Europe Convention for the Protection of Indivi-
duals with regard to Automatic Processing of Personal Data,
61
which all EU
member states are party to, applies generally to the private and public sectors,
including law enforcement authorities. Moreover, the ECtHR has made clear
that article 8 of the European Convention on Human Rights, which guarantees
the right to respect for private life, is applicable to instances of processing of
personal data.
62
The 1981 Convention sets out fundamental safeguards that must be ensured to
preserve the rights of individuals whose personal data are being processed by
automatic means. Although the Convention applies only to the automatic proces-
sing of personal data and does not include the mere collection of data within
the notion of ‘processing’ where not accompanied by the storage of the data
collected,
63
it has laid the foundations for the later developments of personal
58 D. J. Solove,‘Data Mining and the Secur ity-Liberty Debate’ (2007^08) 74 University of Chicago Law
Review 278.
59 OJL 281 of 23 November 1995, 31.
60 Article 3(2) a nd Preamble, Recital 13.
61 CETS, No 108.
62 See eg Eur Ct HR, Leander vSweden 26 March 1987, S A 116, 22 y48; Eur Ct HR (GC), Rotaru v
Romania (Appl n128341/95) Judgement of 4 May 2000 yy 43^45. But see, for the limits of this
protection, Eur Ct HR, Zdanoka vLatvia (Appl n158278/00) partial inadmissibility decision of 6
March 2003.
63 In both these respects, the EU Data Protection Directive has a broader scope of application: this
instrument applies to the processing of personal data, which is de¢ned as ‘any operation or set of
operations which is performed upon personal data, whether or not by automatic means, such as
collection, recording, organization, storage, adaptation or alteration, retrieval, consultation, use,
disclosure by transmission, dissemination or otherwise making available, alignment or combina-
tion, blocking, erasure or destruction’ (Art 2(b)) of Directive 95/46/CE of the European Parlia-
ment and the Council of 24 October 1995 on the protection of individuals with regard to the
processing of personal data and on the free movement of such data, OJL 28 of 23 November
1995, 31). It is not required that the processing take place by automatic means; nor is it required
that the data, once collected, are stored for further use: the mere ‘collection’ of data constitutes
processing covered by the directive.
Olivier De Schutter and Julie Ringelheim
373
r2008 The Authors. Journal Compilation r2008 The Modern Law Review Limited.
(2008) 71(3) 358^384
data protection in European legislation. The basic requirements de¢ned in this
instrument include the following: personal data must be processed fairly and law-
fully;
64
they must be collected for speci¢ed, explicit and legitimate purposes, and
cannot be used in a way incompatible with those purposes;
65
the data collected
must be adequate, relevant and not excessive in relation to the purpose for which
they are processed;
66
they must be kept in a form which permits identi¢cation of
data subjects for no longer than is necessary for the purposes for which they were
collected or for which they are further processed;
67
and the data subject should
have the right to access to and rectify data concerning him or her.
68
In addition,
a key principle established by the 1981 Convention, and restated in Directive 95/
46/EC, is that data relating to religion or racialor ethnic origin are partof a special
category of so-called ‘sensitive’ data, which call for a higher level of protection,
given the risks of discrimination entailed by the processing of such data. Accord-
ingly, under article 6 of the 1981Convention, personal data revealing racial origin
or religious beliefs, as well as personal data concerning political opinions, health
or sexual life ‘may not be processed automatically unless domestic law provides
appropriate safeguards’. Article 9 allows for derogations to this guarantee pro-
vided that they are de¢ned in national legislation and are necessary in a demo-
cratic society inter alia in the interests of public safety, the elimination of criminal
o¡ences, or of protecting the rights and freedoms of others. Nonetheless, the
Committee of Ministers of the Council of Europe has de¢ned more speci¢c
requirements for the processing of sensitive data carried out by the police: its
1987 Recommendation regulating the use of personal data in the police sector
prohibits the collection of data on individuals solely on the basis that they have a
particular racial origin, particular religious convictions or belong to particular
movements or organisations which are not illegal. Furthermore, collection of data
concerning these factors should only be carried out if itis ‘absolutely necessary for the
purposes of a particular inquiry.’
69
More recently, the European Commission has also recognised the necessity
of specifying further the rules applying to the processing of personal data in the
context of law enforcement activities under Title VI of the EU Treaty, which
relates to police cooperation and judicial cooperation in criminal matters. In
October 2005, it proposed the adoption of a Framework Decision on the protec-
tion of personal data processed in the framework of police and judicial coopera-
tion in criminal matters. Such Decision is designed to apply to the processing, by
automatic means or by other means, of personal data which form part (or are
intended to form part) of a ¢ling system, by a competent authority and for the
purpose of the prevention, investigation, detection and prosecution of criminal
64 Council of Europe Convention for the Protection of Individuals with regard toAutomatic Pro-
cessing of Personal Data (CETS, No 108), article 5(a).
65 Article 5(b).
66 Article 5(c).
67 Article 5(d).
68 Article 8(c).
69 Principe 2.4 of the Basic Principles contained in the Appendix to the Recommendation Rec
(87)15 addressed by the Committee of Ministers to the Member States of the Council of Europe,
regulating the use of personal data in the police sector, adopted by the Committee of Ministers on
17 September 1987, at the 401st meeting of the Ministers’ Deputies (our emphasis).
Ethnic Pro¢ling: A Rising Challenge for European Human Rights Law
374 r2008 The Authors. Journal Compilation r2008 The Modern Law Review Limited.
(2008) 71(3) 358^384
o¡ences.
70
It would require the EU member states to prohibit the processing of
personal data revealing racial or ethnic origin as well as religious beliefs, except
where it is provided by a law and is absolutely necessary for the ful¢lment of the
legitimate task of the authority concerned for the purpose of the prevention,
investigation, detection or prosecution of criminal o¡ences or if the data subject
has given his or her explicit consent to the processing, and suitable safeguards are
provided.
71
But, apart from the fact that this framework decision is yet to be
approved by the Council, it allows for the processing of sensitive data, including
data relating to the race, ethnic origin, or religion of individuals, for the preven-
tion of criminal o¡ences in general, without this having to relate to speci¢c inqui-
ries into o¡ences which have been committed, as would be required under the
Basic Principles regulating the use of personal data in the police sector, adopted
by the Committee of Ministers of the Council of Europe in 1987.
72
Therefore,
under the draft framework decision, ethnic pro¢ling as de¢ned above is not per
se illegal; it merely has to be accompanied by adequate legal safeguards if resorted
to by the EU Member States’ law enforcement authorities.
The questionwhether pro¢ling based on automatic means and involving eth-
nic or religious criteria is compatible with the right to privacy has been raised
before the German Federal Constitutional Court (Bundesverfassungsgericht)inthe
Rasterfahndung case. As mentioned above, between 2001 and 2003, the German
police authorities conducted a nationwide dragnet investigation for Islamic ter-
rorist‘sleepers’.The operation involved the screening of a huge amount of perso-
nal data held by various public or private institutions, with a view to identifying
people corresponding to a de¢ned pro¢le, which included mainly being male,
aged18 to 40, a student or former student, Muslim, and native or national of spe-
ci¢ed countries with apredominantly Muslim population.These datawere com-
bined with personal information available on general registers, and a nationwide
¢le listing all individuals corresponding to the‘terrorist sleepers’ pro¢le was estab-
lished. This operation mobilised an enormous amount of personnel and time
within the police force. According to certain sources, data of 8.3 million persons
were analysed by the Germanpolice.
73
Approximately 32,000 persons were iden-
ti¢ed as potential terrorist‘sleepers’ and subject to closer examination. However,
this programme did not lead to the arrest of any single terrorist suspect.
74
70 Proposal for a Council framework decision on the protection of personal data processed in the
framework of police and judicial cooperation in criminal matters (COM(2005) 475 ¢nal), article
3(1).
71 ibid article 6.
72 See above n 69.
73 D. Moeckli,‘Terrorist pro¢ling and the importance of a proactive approach to human rights pro-
tection’ 16 December 2006 available at the Social Science Research Network (SSRN): http://
ssrn.com/abstract=952163 (last vis ited 18 February 20 08).
74 G. Kett-Straub,‘Data Screening of Muslim Sleepers Unconstitutional’ (2006) 7 German LawJournal
967, 970^971; D. Moeckli,‘Discriminatory Pro¢les: Law Enforcement After 9/11 and 7/7’ (2005) 5
European Human Rights Law Review 517, 520^522. See also F. Bignami,‘European versus American
Liberty: A Comparative Privacy Analysis of Anti-Terrorism Data-Mining’ Duke Law School Legal
Studies Research Paper Series Research Paper No 135, January 2007, 31^32, available at the Social
Science Research Network (SSRN): http://papers.ssrn.com/sol3/papers.cfm?abstract_id=955024
(last visited 18 February 2008). (Hereinafter: Resource Guide).
Olivier De Schutter and Julie Ringelheim
375
r2008 The Authors. Journal Compilation r2008 The Modern Law Review Limited.
(2008) 71(3) 358^384
In its landmark judgment of 4 April 2006, the German Constitutional Court
found the programme to be unconstitutional. It declared it incompatible with the
fundamental right of self-determination over personal information, which pro-
tects individuals against the collection, transfer, storage or processing of personal
information by state authorities, including the police. The Court admitted that
the operation pursued a legitimate aim, namely safeguarding the national security,
and that in theory data-mining could be a suitable means to achieve this goal. But,
in the present context, it held that the interference with the right of informational
self-determination resulting from this operation was disproportionate.The bur-
den imposed on constitutional rights was particularly serious, among other rea-
sons because the screening exercise focused on a speci¢c religious community,
namely the Muslim community, and was therefore likely to have a ‘stigmatising
impact’ onthose concerned and to‘increase the risk of beingdiscriminated against
in working and everyday life’.
75
In the viewof the Court, such far-reaching con-
sequences for individuals’ rights could only be considered acceptable if there were
actual facts demonstrating a concrete, ie imminent and speci¢c, endangerment
(konkrete Gefahr) of national security or individual rights.
76
By contrast, a general
and indeterminate threat as the one constantly existing since 11 September 2001,
did not provide a su⁄cient basis for justifying the operation.
77
We have serious doubts, however, whether this episode, regrettable as it is,
should be seen as calling for the reinforcement of personal data protection laws
in Europe. Not only did the legal safeguards existing in Germany su⁄ce, in the
¢nal instance, to protect the victims of such data-mining. But, in addition, we
have mentioned the need for a monitoring of law enforcement o⁄cers which
includes the collection of data on the impact of their activities on racial or ethnic
minorities.
78
In the context of Europe today, any further reinforcement of the
protection of personal data, especially if it leads to a form of ‘sanctuarization’ of
sensitive data relating to race or ethnicity, could discourage such monitoring: far
from ensuring a reinforced protection from discrimination, it could constitute a
step backwards in such protection, by depriving victims from having access to
statistical data which may in many instances be crucial to proving the existence
of discrimination, whether direct or indirect.
79
It is withthis concern in mind that
75 Kett-Straub, ibid citing the Constitutional Court’s judgment, at 971, n 24.
76 ibid at 973; Bignami, n 74 above, 32.
77 Kett-Straub, ibid 968.
78 See above, text corresponding to n 45^46.
79 In the case of Hoogendijk vthe Netherlands the European Court of Human Rights recognized that,
unless the victim is allowed to bring forward statistics establishing a presumption of discrimina-
tion and shifting the burden of proof on the shoulders of the defendant, discrimination may in
certain instances be di⁄cult or impossible to prove:‘[W]here an applicant is able to show, on the
basis of undisputed o⁄cial statistics, the existence of a prima facie indication that a speci¢c rule ^
although formulated in a neutral manner ^ in fact a¡ects a clearly higher percentage of women
than men, it is for the respondent Government to show that this is the result of objective factors
unrelated to anydiscrimination on grounds of sex. If the onus of demonstrating that a di¡erence
in impact for men and women is not in practice discriminatory does not shift to the respondent
Government, it will be in practice extremelydi⁄cult for applicants to prove indirect discrimina-
tion’ (Hoogendijk vthe Netherlands (dec) no 58461/00 6 January 2005). This in turn in£uenced the
position of the Court in the case of D. H. a n d Oth e r s vCzech Republic, referred to above n 45.
Ethnic Pro¢ling: A Rising Challenge for European Human Rights Law
376 r2008 The Authors. Journal Compilation r2008 The Modern Law Review Limited.
(2008) 71(3) 358^384
we examine, in the next section, how the ¢ght against ethnic pro¢ling in Europe
can be strengthened further.
STRENGTHENING THE MEANS TO COUNTER ETHNIC PROFILING
The discussion in the last section reveals that the use of ethnic pro¢ling strategies
by law enforcement o⁄cers is in potential con£ict with several European human
rights norms. But it also emerges that the present European legal framework pre-
sents certain gaps and limitations, and only partially responds to the challenges
raised by these expanding practices. Despite the encouraging statement made by
the European Court of Human Rights in theTimishev case, we still lack a clear
and speci¢c prohibition of usingethnicity or religion as a proxy for propensity to
commitcrimes, either in general or in the speci¢c area of counter-terrorism.With
respect to ethnic pro¢ling carried out by automatic means, typically through data
mining operations, while the current framework is in general satisfactory, its
implications in the context of law enforcement activities could be made more
explicit, especially since the aforementioned Framework Decision on the protec-
tion of personal data processed in the framework of police and judicial coopera-
tion in criminal matters when it will be adopted, will still be insu⁄cient in
protecting people belonging to ethnic, religious or national groups, in the face
of counter-terrorism strategies which, all too often, are tempted to rely on ethnic,
religious or national stereotyping: this, indeed, could undermine the protection
currently provided under the Council of Europe standards whose level of protec-
tion is higher.
80
In addition, in order to counter informal ethnic pro¢ling practices,
carried out by o⁄cers acting on the ground, which probably represents the most
frequent occurrence of ethnic pro¢ling as de¢ned above, it is essential that states
de¢ne with the greatest clarity possible the conditions under which law enforce-
ment authorities may use their powers in areas such as identity checks and stop-
and-search procedures. Yet, the problem with informal types of ethnic pro¢ling,
by contrast with formalised pro¢ling programmes, is not merely one of having
adequate legislation prohibiting it: because it takes the form ofa de facto practice, it
may be extremely di⁄cult to detect and prove that law enforcement o⁄cers are in
fact targeting certain people on account of their ethnic or religious background.
Fromthis perspective, a critical aspect of the development ofa strategy e¡ectively
tackling this phenomenon lies with the ability of public authorities to have su⁄-
cient information about how law enforcement o⁄cers use their powers on the
80 As mentioned above (text corresponding to n 69^70), the draft EU Framework Decision allows
for the processing of sensitive data, including data relating to the race, ethnic origin, or religion of
individuals, for the prevention of criminal o¡ences in general, without this having to relate to spe-
ci¢c inquiries into o¡ences which have been committed. In contrast, the Basic Principles regulat-
ing the use of personal data in the police sector, adopted by the Committee of Ministers of the
Council of Europe in 1987, only allow such processing of sensitive data in the police sector for
speci¢c inquiries into past o¡ences. The ‘reasonable suspicion’ standard recommended in para 3
of General Policy Recommendation n111 of the European Commission against Racism and
Intolerance constitutes a mere transposition, in the exercise of the the powers of the police relating
to control, surveillance or investigation activities, of this more general principle.
Olivier De Schutter and Julie Ringelheim
377
r2008 The Authors. Journal Compilation r2008 The Modern Law Review Limited.
(2008) 71(3) 358^384
ground vis-a
'-vis people belonging to di¡erent ethnic groups. Furthermore, in
order to facilitate the bringing of judicialcomplaints by victims, certain measures
can be envisaged to facilitate the task of the complainant inproving that he or she
has been subject to ethnic pro¢ling.
Monitoring law enforcement o⁄cers’ behaviour
That ethnic pro¢ling exists and is a serious problem in various EU countries is
attested by many individual testimonies and several NGO’s reports.
81
However,
in order to adequately address the issue, national authorities cannot rely on anec-
dotal evidence; they need to have information as precise and comprehensive as
possible about how police and other law enforcement o⁄cers behave with regard
to people from di¡erent ethnic backgrounds.This is essential in enabling them to
detect whether and when ethnic pro¢ling is practiced in the ¢rst place.The Com-
mittee on the Elimination of Racial Discrimination thus encourages states parties
to the Convention on the Elimination of All Forms of Racial Discrimination to
adopt a set of factual indicators facilitating the identi¢cation of racial discrimina-
tion in the administration and functioning of the criminal justice system. Its Gen-
eral Recommendation n131 (2005) invites them, in particular, to embark on
‘regular and public collection of information from police, judicial and prison
authorities and immigration services, while respecting standards of con¢dential-
ity, anonymity and protection of personal data’.
82
Thus, where the European
Commission against Racism and Intolerance recommends in its General Policy
Recommendation n111 that data be broken down by grounds such as national or
ethnic origin, language, religion and nationality in order to improve the moni-
toring of police activities and to detect instances of what it refers to as ‘racial pro-
¢ling’, it is hardly on unchartered territory.
83
Indeed, this hasbeen the response of many municipalities and states in the Uni-
ted States to the allegations that their lawenforcement authorities were dispropor-
tionately targeting Blacks and Hispanics in tra⁄c or pedestrian stops: they have
set up systems to collect and analyse statistical data on motorists who are stopped,
detained, and searched bypolice patrols, including ontheir racial and ethnic back-
ground, with a view to monitoring the o⁄cers’ behaviour.
84
In Europe, only the
81 See, in particular, the work done by the Open Society Justice Initiative, including the reportsJus-
tice Initiatives: Ethnic Pro¢ling by Police in Europe (NewYork: Open Society Institute, 2005); Ethnic
Pro¢ling in the Moscow Metro (NewYork: Open Society Institute, 20 06); ‘I Can Stop and SearchWho-
ever I Want’. PoliceStops of Ethnic Minorities in Bulgaria, Hungary, and Spain (NewYork: Open Society
Institute, 2007).
82 General recommendation No 31 on the prevention of racial discrimi nation in the administration
and functioning of the criminal justice system (2005) para 1(2), available on http://www2.
ohchr.org/english/bodies/cerd/comments.htm (last vi sited 18 February 2008).
83 See above, text corresponding to nn 45^46.
84 See F.Weatherspoon,‘Ending Racial Pro¢ling of African-Americans in the Selective Enforce-
ment of Laws: in Search of Viable Remedies’ (2004) 65 University of Pittsburgh Law Review 721,
743.; D. A. Ramirez, J. McDevitt, A. Farrell, A Resource Guide on Racial Pro¢ling Data Collection
Systems ^ Promising Practices and Lessons Learned November 2000 Document prepared with the
¢nancial support of the US Department of Justice, available at http://www.ncjrs.gov/txt¢les1/
bja/184768.txt (last vis ited 18 February 20 08) 2.
Ethnic Pro¢ling: A Rising Challenge for European Human Rights Law
378 r2008 The Authors. Journal Compilation r2008 The Modern Law Review Limited.
(2008) 71(3) 358^384
United Kingdom has implemented a similar programme. The Criminal Justice
Act passed in 1991 contains a section 95 which requires the Home Secretary to
publish each year such information as he considers expedient for the purpose of
‘a) enabling persons engaged in the administration of criminal justice to become
aware of the ¢nancial implications of their decisions, or b) facilitating the perfor-
mance by such person of their duty to avoid discriminating against any persons
on the grounds of race or sex or any other improper ground.’ In order to ful¢l this
provision, it was decided a few years later to introduce mandatory ethnic moni-
toring in police services from April 1996.
85
According to the system established by
the Home O⁄ce, each time a search is conducted on a person, the o⁄cer must
complete a form and record various information, including the self-de¢ned ethnic
origin of the person searched,
86
but also his or her name; the object of and the
grounds for the search and its outcome; the date, time and place; and the identities
of the o⁄cers involved.
87
This obligation is valid unless there are exceptional cir-
cumstances that would make this wholly impracticable, and the person has the right
to receive a copy of the search record immediately or at the police station within a
year.
88
The data thereby recorded are then used by the Home O⁄ce to produce and
publish statistics about stops and searches, broken down by ethnic categories.
89
Rather than, or in addition to, requiring o⁄cers themselves to record data on
all the persons against whom they use their powers, the promotion of empirical
scholarly research based on the observation of a series of police encounters with
citizens and interviews with citizens or o⁄cers, can also be envisaged as an alter-
native means to gauge the existence and extent of discriminatory practices in the
police. Systematic and regular data collection, however, presents the advantage of
providing state authorities with an overall picture of police o⁄cers’ conduct in the
whole country and of enabling them to follow the evolution of the situation in
the long run; in addition, having the law enforcement o⁄cers themselves collect
the information may encourage raising awareness within the police force of the
risks of discrimination, and lead the o⁄cers concerned to realize that theyare act-
ing on the basis of stereotypes which hitherto may have been unconscious rather
than deliberate.The main purpose of such data collection e¡orts, however, is to
highlight potential police misconduct and deter it. Having statistics on the ethnic
85 M. FitzGerald and R. Sibbitt,‘Ethnic Monitoring in Police Forces: A Beginning’ Home O⁄ce
Research Study (1997) 173, 9^13; P. Hillyard, ‘New Labour and Ethnic Monitoring in Policing’
(1999) 70 Radical Statistics, available at: http://www.radstats.org.uk/no070/article6.htm (last visited
18 February 2008).
86 Original ly, it had bee n decided that o⁄cers would record the person’s ethnic origin on the basis of
their own determination. However, following a recommendation to this e¡ect of the McPherson
Report (SirWilliam McPherson,The Stephen Lawrence Inquiry (London, England:The Stationary
O⁄ce, 1999)), the system was changed in favour of the self-de¢ned ethnic identity (Resource
Guide, n 74 above, 39^40). Nonetheless, o⁄cers must continue to include their own perception
of the ethnic background of the persons stopped and searched (Police and Criminal Evidence Act
1984,CodeA,note18).
87 Police and Crimi nal Evidence Act 1984, Code A, para 4.3.
88 Police and Crinimal Evidence Act 1984, Code A, para 4.1. and 4.2; FitzGerald and Sibbitt n 85
above, 43^44.
89 See Statistics and Raceand the CriminalJustice System (2005) Home O⁄ce publication under section
95 of the Criminal Justice Act 1991, available at http://www.homeo⁄ce.gov.uk/rds/index.html
(last visited 18 February 2008).
Olivier De Schutter and Julie Ringelheim
379
r2008 The Authors. Journal Compilation r2008 The Modern Law Review Limited.
(2008) 71(3) 358^384
background of the persons stopped, searched, or arrested makes it possible to
detect if people of a speci¢c ethnic origin are disproportionately a¡ected by these
decisions. These statistics, therefore, permit the authorities to have some degree of
control over how police o⁄cers use their margin of discretion when performing
such duties.When a disproportion is observed, further investigation is needed to
verify if it can be justi¢ed by objective and legitimate factors. Absent such justi¢-
cation, it should be taken as a sign that police forces are unfairly targeting the
members of certain ethnic groups. Furthermore, insofar as the identity of o⁄cers
conducting the search or arrest is recorded, public authorities can determine if
discriminatory behaviours are limited to some individual o⁄cers, or whether
they re£ect a general trend in the police institution ^ what the Stephen Lawrence
inquiry in the United Kingdom referred to as ‘institutional racism’.
90
More gen-
erally, the information on police attitudes towards ethnic minorities provided by
these data collection systems will help the authorities in de¢ning adequate
responses to ethnic pro¢ling practices where they are observed. Such responses
may include, in particular, individual sanctions, the formulation of appropriate
internal procedures or training programmes to educate o⁄cers about prejudice
and stereotypes that may in£uence their decisions and the promotion of more
respectful police-citizens encounters.
91
To be sure, in the European context, the processing of personal data for the
purpose of monitoring law enforcement o⁄cers’ behaviour would have to com-
ply with the norms relating to the protection of personal data, which were out-
lined in the previous section. Yet, it should be emphasised that these norms are
only concerned with‘personal data’, namely‘any information relating to an iden-
ti¢ed or identi¢able individual’.
92
Hence, no personal data is involved where
information is collected on an anonymous basis or once the information collected
is made anonymous in order to be used in statistics, since such data cannot be
traced to any speci¢c person. In the United States, in some of the jurisdictions
where a data collection system was put in place, it was decided that the identity
of the citizens stopped and whose racial or ethnic a⁄liation is recorded, would
remain anonymous.
93
In most other places however, including in the United
Kingdom, the authorities opted for the inclusion, in the initial collection, of data
permitting the identi¢cation of the person concerned, in order to render it possi-
ble to verify the information and/or to determine if the person searched or
arrested was eventually charged with any o¡ence. In European states, the legal
requirements of the 1981 Council of Europe Convention pertaining to the
90 In the Stephen Lawrence Inquiry, ‘institutional racism’ was de¢ned as ‘the collective failure of an
organisation to provide an appropriate and professional service to people because of their colour,
culture, or ethnic origin. It can be seen or detected in processes, attitudes and behaviour which
amount to discrimination through unwitting prejudice, ignorance, thoughtlessness and racist
stereotyping which disadvantage minority ethnic people. It persists because of the failure of the
organisation openly and adequately to recognise and address its existence and causes by policy,
example and leadership’ (Report of an Inquiry by SirWilliam Mcpherson of Cluny, February 1999, cm
4262-I, para 6.34).
91 Weatherspoon, n 84 above, 745^746; Resource Guide, n 74 above, 13.
92 The Council of Europe Convention for the Protection of Individuals with Regard toAutomatic
Processing of Personal Data (1981), Article 2.
93 See the experience of the city of San Jose, described in Resource Guide, n 74 above, 21.
Ethnic Pro¢ling: A Rising Challenge for European Human Rights Law
380 r2008 The Authors. Journal Compilation r2008 The Modern Law Review Limited.
(2008) 71(3) 358^384
automatic processing of personal data and, more speci¢cally, of sensitive data, are
therefore applicable. But as we have seen above, whilst these rules restrict the cir-
cumstances in which sensitive data, such as information on one’s ethnic or racial
origins, can be processed, theydo not impose an absolute prohibition on the pro-
cessing of such data.
94
Combating discriminatory behaviour within the police
appears as a legitimate public interest for the pursuance of which such treatment
could be allowed, subject to adequate safeguards. In addition, given that the data
collected in the frameworkof ethnic monitoring programmes are used to consti-
tute statistics, the principles enumerated in the Recommendation No R (97) 18 of
the Committee of Ministers of the Council of Europe on the protection of per-
sonal data collected and processed for statistical purposes
95
also should be taken
into account. This Recommendation provides in particular that the data collected
and processed shall be made anonymous as soon as they are no longer necessary in
an identi¢able form.
96
It also states that where personal data have been collected
and processed for statistical purposes, they shall serve only those purposes, and
shall not be used to take a decision in respect of the data subject, nor to supple-
ment or correct ¢les containing personal data which are processed for non-statis-
tical purposes.
97
In addition, in order for the processing of personal data for
statistical purposes to remain proportionate, the principle of ¢nality should be
strictly observed: only those personal data shall be collected and processed which
are necessary for the statistical purposes to be achieved.
98
These are important
safeguards, but they are safeguards, again, which do not impose insuperable
obstacles to an improved monitoring of the practices of law enforcement autho-
rities in order to identify patterns of discrimination.
Proving ethnic pro¢ling in courts
Like other forms of discrimination, that resulting from informal ethnic pro¢ling
is extremely di⁄cult to prove in a judicial setting. It is especially so when the vic-
94 In its resolution on Non-discrimination and equal opportunities forall ^ A framework strategy adopted on 8
May 2006 (2005/2191(INI), EP doc A6- 0189/2006 (rappT. Zdanoka)), the European Parliament
called for a clari¢cation of the requirements of dataprotection legislation on this issue, and asked
in particular the Member States to‘develop their statistics tools with a view to ensuring that data
relating to employment, housing, education and income are available foreach of the categories of
individual which arelikely to su¡er discrimination based on one of the criteria listed in Article 13
of the EC Treaty’ (para 20). Following a suggestion of the EU Network of independent experts
on fundamental rights (see EU Network of Independent Experts on Fundamental Rights,The-
matic Comment n13: the rights of minorities in the Union (April 2005), available at: http://ec.
europa.eu/justice_home/cfr_cdf/index_en.htm (last visited 18 February 20 08)), the European Parlia-
ment called for the Working Party established under Article 29 of Directive 95/46/CE of the Eur-
opean Parliament and the Council of 24 October 1995 on the protection of individuals with
regard to the processing of personal data and on the free movement of such data to deliver an
opinion on the question of how the use of statistical data for the purposes of combating discrimi-
nation could be reconciled with the requirements of data protection legislation.
95 Adopted by the Committee of Ministers on 30 September 1997 at the 602nd meeting of the Min-
isters’ Deputies.
96 Para 3.3.
97 Para 4.1.
98 Para 4.7.
Olivier De Schutter and Julie Ringelheim
381
r2008 The Authors. Journal Compilation r2008 The Modern Law Review Limited.
(2008) 71(3) 358^384
tim is required to put forward elements demonstrating, beyond reasonable doubt,
that a discriminatory motive has been underlying a particular behaviour. Indeed,
as highlighted by Canadiancourts:‘As with other systemic practices, racial pro¢l-
ing can be conscious or unconscious, intentional or unintentional.’
99
Hence,
‘[f]ailing an admission on the part of the o⁄cers, which is unlikely, the proof of
racial pro¢ling will most often be indirect.’ The Canadian courts thus consider
that ‘if racial pro¢ling is to be proven it must be done by inference drawn from
circumstantial evidence.’
100
In the view of the Canadian judges, this reality needs
to be taken into account when considering the evidence brought by a person
claiming to be the victim of ethnic pro¢ling. Accordingly, they have come to
allow a shifting of the burden of proof in such contexts: a criminal defendant,
arrested following a stop and search procedure alleged to constitute a form of
ethnic pro¢ling, mayestablish a presumption that such pro¢ling has occurred by
proving that ‘it is more probable than not’that there was no articulable cause for
the stop other than the fact of his or her racial or ethnic origin.
101
This approach could be a source of inspiration for European jurisprudence or
legislation.The EUmember states are already familiar with the notion of shifting
the burden of proof in antidiscrimination cases. Under the 2000 Equality direc-
tives, they are under the obligation to ensure that when persons claiming to be the
victims of discrimination prohibited by these directives, establish before a court
or other competent authority, facts fromwhich it may be presumed that there has
been direct or indirect discrimination, it shall be for the respondent to prove that
there has been no breach of the principle of equal treatment.
102
It is true that in the
context of criminal proceedings, the presumption of innocence, as enshrined in
Article 6(2) of the European Convention on Human Rights, may impose limits
to such mechanism. These directives therefore do not impose the possibility of
shifting the burden of proof in such proceedings. However, the European Court
of Human Rights has made clear that the Convention does not prohibit pre-
sumptions of fact or law in principle, but rather requires states to con¢ne them
within reasonable limits, which maintain the rights of defence.Thus, the criminal
law of a state party can establish a presumption of guilt in particular contexts,
insofar as such presumption is not absolute and does not lead to depriving the
defendant of every possibility of defence, for instance by establishing a case of force
majeure or proving unavoidable error.
103
A shift of the burden of proof would sig-
ni¢cantly facilitate the task of the victims in proving that they have been subject
to ethnic pro¢ling. In addition, where statistics on stops, searches, or arrestsby the
police, broken down by ethnic background, are available, allowing the complai-
nants to produce such statistics (or other types of reliable and relevant studies) in
99 The Queen vCampbell, Alexer, Court of Quebec (Criminal Division) (n1500-01-004657-042-001)
(Judgment of 27 January 2005 by the HonourableWestmoreland-Traore
Ł)at[34].
100 ibid at [35] (quoting from RvBrown,173 CCC (3d) 23 at [44].
101 ibid at [25].
102 Directive 2000/43/EC, Article 8(1); Directive 2000/78/EC, article 10(1).
103 Eur Ct HR, Salabiaku vFrance (Appl No 10519/83) Judgment of 7 October 1988 y28. See also Eur
Ct HR (2nd section), Radio-France and Others vFrance (Appl No 53984/00) Judgment of 30 March
2004 y24.
Ethnic Pro¢ling: A Rising Challenge for European Human Rights Law
382 r2008 The Authors. Journal Compilation r2008 The Modern Law Review Limited.
(2008) 71(3) 358^384
courts, could in many cases greatly help them establishing a presumption that they
have been the victim of a discriminatory practice.
CONCLUSION
We have seen that ethnic pro¢ling may take two forms: it can be formalized, and
consist in the processing of data relating to race or ethnicity, religion, or national-
ity; and it can be informal, as in stop and search procedures which, because they
are insu⁄ciently regulated, are conducted in ways tainted by prejudice or stereo-
typing, conscious or unconscious.Whatever its form, ethnic pro¢ling is not only
ine⁄cient. It is also counterproductive. It alienates minority communities at a
time when social cohesion is needed more than ever, and when it is becoming
widely recognized that intelligence ^ and, therefore, the maintenance of good
relationships between law enforcement authorities and all communities ^ is cru-
cial to combating crime e¡ectively. And it reinforces in the public the very stereo-
types which it stems from in the ¢rst place. In order to combat ethnic pro¢ling,
we must ¢rst have regard to the two forms it can take.
As regards the ¢rst of those two forms of ethnic pro¢ling ^ which consists in
the processing of data relating to race or ethnicity, religionor nationality, we sub-
mit that the existing regime of data protection in Europe is in principle adequate
to the task it is asked to perform: to protect individuals from the improper use of
such criteria in decisions relating to lawenforcement. Of course, at the level of the
EU, the Framework Decision on the protection of personal data processed in the
framework of police and judicial cooperation in criminal matters still has not
been adopted, and when it will be, it will still present a number of gaps we have
highlighted above.
104
But all the EU Member States are parties to the 1981Coun-
cil of Europe Convention for the Protection of Individuals with regard toAuto-
matic Processing of Personal Data,
105
and this instrument should be implemented
in conformity with the Recommendation regulating the use of personal data in
the police sector which the Council of Europe’s Committee of Ministers
addressed to the Member States of the Council of Europe in 1987. This recom-
mendation explicitly restricts the collection of data relating to race or ethnicity
by law enforcement agencies to situations where it is ‘absolutely necessary for the
purposes of a particular inquiry’, thus excluding such processing of sensitive data
in the police sector for general purposes or in the context of proactive policing.
106
Provided this is properly implemented in the national laws of the Council of
Europe member States, this should ensure an adequate safeguard against the
kind of data mining which took place in Germany in 2001^2003, before being
condemned by the German Federal Constitutional Court.
The danger is not so much, then, that data protection legislation is insu⁄-
ciently protective. It is, rather, that is could be interpreted too rigidly in certain
cases, and turn into an obstacle to an improved monitoring of the behaviour of
104 See above nn 70^72 and corresponding text.
105 Se e a b o ve n 53.
106 Se e above n 61.
Olivier De Schutter and Julie Ringelheim
383
r2008 The Authors. Journal Compilation r2008 The Modern Law Review Limited.
(2008) 71(3) 358^384
law enforcement o⁄cers, in order to prevent and e¡ectively prohibit the second
form of ethnic pro¢ling we have distinguished: the informal, often unconscious,
use of race or ethnicity in the exercise of their powers bylaw enforcement autho-
rities. The rise of ethnic pro¢ling in policing requires that we document more
systematically the impact on identi¢ed minorities of the practices of law enforce-
ment o⁄cers, especially in identity stops. Contrary to a widely held assumption,
data protection legislation in Europe ^ the principles of which are laid down in
the 1981 Council of Europe Convention for the Protection of Individuals with
regard toAutomatic Processing of Personal Data ^ does not prohibit such mon-
itoring. Only if we can adequately document a practice can we then re£ect on the
requirements of the principle of non-discrimination. Where it appears that law
enforcement authorities are disproportionately targeting certain minority groups
de¢ned by their race or ethnicity, religion or nationality, this should be
denounced as ethnic stereotyping, and remedial measures ^ including, in the most
egregious cases, disciplinary or other sanctions ^ should be taken. It should not
matter, in our view, that a statistically signi¢cant relationship can be shown to
exist between certain criminal behaviours and membership in such targeted
groups: ethnic pro¢ling is not simply a misuse of the resources of law enforce-
ment agencies, leading to the perverse e¡ects mentioned above; because it consists
in‘the invidious use of race and ethnicityas markers of suspicion’,
107
it is a form of
discrimination and should be prohibited as such, not only because it does not
contribute to e¡ective crime prevention or detection ^ but simply because race
or ethnicity, religion or nationality should not be used in our societies as criteria
for decision-making in law enforcement.
107 C. Stone,‘Preparing a Fresh Assault on Ethnic Pro¢ling’Justice Initiatives June 2005,1.
Ethnic Pro¢ling: A Rising Challenge for European Human Rights Law
384 r2008 The Authors. Journal Compilation r2008 The Modern Law Review Limited.
(2008) 71(3) 358^384
