What do we achieve by systematically recording
and monitoring discriminatory acts?
What do we already know, what areas do we lack
knowledge about and what are the conditions for our being able to continuously
follow the development of discrimination?
Transcript of a paper given at the conference
on the Norwegian central government's efforts to combat ethnic discrimination,
3 October 2000
Senior
Research Fellow Lars Østby
Statistics
Norway
The
title of my paper is extremely long, and some people may well find the paper as
long and boring as its title might appear. However, I would like to stress
right from the very start how important it is to measure adequately the phenomenon we want to combat. Of course,
fighting discrimination is the main issue - I do not have to tell you that.
Nevertheless, it is critical also to establish surveys that enable us to tailor
our efforts to overcome the phenomenon, assess the results achieved, and
evaluate their success. The issue of monitoring must therefore be addressed in
order to build a knowledge base on which the best possible decisions can be
made. Similarly, a lack of good statistics will yield uncertainty about whether
implemented measures are effective and which measures are having the intended
effect.
I
was asked to present this paper against the background of the criticism that
the European Commission against Racism and Intolerance ECRI (cf. Council of
Europe 1998) and the United Nations' Committee on the Elimination of Racial
Discrimination CERD (cf. the observations from its 57th meeting:
CERD/C/57Misc.44/Rev.3 23 August 2000) have raised against Norway for its
inadequate monitoring of racist incidents and acts. In cases such as this, the
Norwegian Statistics Act indicates that Statistics Norway shall be responsible
for implementing the necessary measures. Asking me to make this presentation is
more the choice of the Directorate of Immigration than of Statistics Norway, as
I do not have any special expertise in this field. I am therefore extremely
grateful to Kate Chapman who so willingly helped me with her comments and by
providing me with a variety of publications.
My main ballast is probably the general statistical competence I have
gained through analysing demography and social statistics (which include statistics
on immigrants) for three decades.
Much
of what I will say today is coloured by my personal experiences and views, and
another demographer would doubtless have put things another way - and possibly
also drawn different conclusions. Nor is this paper the result of an internal
decision-making process in my organisation. Nevertheless, there are a number of
ground rules for good statistical practice that are not affected by personal
preferences and styles, and this makes me believe that the conclusions I have
drawn cannot be dismissed as only one person's individual point of view.
Against this background, I would therefore like to spend a few minutes
discussing terms and concepts and to explain which of the opinions voiced here
are my own personal views and which are firmly anchored in accepted statistical
theory.
On a
personal level, the background for my professional commitment to the field is
based on my deliberations on the role of the statistician in fighting racism.
These deliberations were triggered some ten years ago when my young neighbour -
a Norwegian citizen, born in Norway to parents born in Pakistan - told me about
a recent trip to a shopping centre in Oslo. It was a winter's day, and she was
walking towards the entrance behind an older gentleman, who suddenly stopped in
the sliding doors, so that she was able to enter before the doors closed again.
However, it turned out that his intentions were far from courteous. As she
walked past him, he spat at the ground besides her feet and cursed with great
feeling. This is an obvious example of what some people would call a
discriminatory act and the sort of incident that we would love to know the
scope and development of in order to allow us to study the trends in
discrimination in the community.
My
thoughts on the development of a monitoring system that I am going to share
with you are intended to promote discussion that might help us one step farther
on our way towards something better. I have not developed a complete system -
or even a fully matured point of view on the matter. That would take more
months than I have had days to prepare this paper. An effective monitoring
system can only be developed through a collaboration between statisticians and
the various users, representing the government, NGOs, and the possible victims
for discrimination, with mutual respect for each other's differences!
Developing this kind of system is a major job, which requires expertise, skill,
resources, co-ordination and everything else that most jobs require, but it is not impossible!
I
would also like to confess my fear of seeming negative and arrogant in this
forum, but still I must ask that you listen to what I have to say. In the phase
that this work is now in, which I regard as relatively early in the process
towards developing statistics on discriminatory acts, I believe that it is
important that the various crossroads we come to are well signposted. We need
somebody to tell us which roads are dead ends and where the other possible
choices will lead us. Some paths clearly lead to the desired ends, without
there necessarily being good statistics lying there waiting for us. It may
sound as if I am suggesting that these paths ought not be taken, and I
certainly do not want to be perceived as underestimating the value of these
goals. Rather, my intention is merely to point out that it is not there that we will find the statistical
foundation that is a crucial element in this important work.
The objectives and contents of this paper
This
field is not so very unlike many other fields where, despite numerous problems,
we still have satisfactory statistics. We should therefore be looking for tools
that are commonly used to generate statistics. The purpose of monitoring
discriminatory acts and incidents is to:
·
provide an image of the underlying reality that is not
systematically biased
·
as a minimum, describe reasonably correctly the trends
of the development and give a level for the scope of the acts, or at
least for acts above and beyond a certain (and fixed) degree of gravity
·
provide relevant distributions of the acts and incidents, in terms
of both the victims and the perpetrators. This should also include
regional distribution.
It is not enough simply to count incidents; we must
also compare the number of incidents with the number of people that could
have been victims (or perpetrators). In other words, we need to define our
population at risk.
If we and others manage to achieve this, we will then
have a basis for comparison. For a number of
reasons, I believe that comparability is extremely important, and we must
be able to compare data over time, between nations and between groups
within the country. It is only once we have established comparable data
than we can say that our job is done.
I have included all these points in the checklist
below, which I call "What are good statistics?":
|
What are good statistics?
Good statistics:
Reflect the underlying reality without
systematic distortions - representativity
Provide a sufficiently correct description of
the development trends
Give a level for the scope (given criteria)
Provide relevant distributions:
Between regions
Between groups
Form a foundation for comparison
(the same definitions must be used throughout)
Must make it possible to relate the observed
incidents or acts to a common risk population
Note: Neither self-recruitment nor conversations with key
informants satisfies these requirements.
|
The structure of this presentation will be as follows:
Delimitation and definitions of key terms and concepts
The criticism of Norway
The European Commission against Racism
and Intolerance (ECRI)
The United Nations' Committee on the Elimination of
Racial Discrimination (CERD)
What is being done in other countries?
Does that fulfil all the requirements?
What are we doing in Norway?
What should we be doing?
Suggested solutions
Terms and concepts
Before I broach the actual points on which Norway has
been criticised, I would like to say a few words about some of the key
terms and concepts I will be using and about the need to continue the
conceptual development. The work performed to date demonstrates that this
field has exactly the same need for clarification of the terms and
concepts as any other area of social research.
It is not always easy to find a single Norwegian
translation for the English term monitoring and different people interpret it
differently. I will be using the term here to refer to the elements
related to the creation a statistical description of the field, and, as
the Centre for Combating Ethnic Discrimination (SMED 2000) says, following
the development over time with a special focus on process change (page
40).
Discrimination is a far from unproblematic term. The Centre for
Combating Ethnic Discrimination (SMED 2000, page 14) reports that the
Norwegian Ministry of Foreign Affairs translates the English word
"discrimination" with the Norwegian "forskjellsbehandling" (literally:
different treatment). Rogstad (1996 and 1998) discusses the semantic
meaning and operationalisation of the term and addresses also a number of
ethical aspects. The main distinction is between whether the focus is on
the intentions of an act or whether discrimination can be used to refer to
any process that creates differences. We can also query whether
inequalities in the outcome can be related to discrimination if there were
no differences in the individuals' chances to begin with. We can scarcely
regard differences that are the result of different preferences as
discrimination, but the issue becomes a little more complex if the
differences in preferences are based on differences in the frame of
reference for the formation of preferences between different groups.
It is also important to consider whether we are going
to include only acts that have a discriminatory motive, or whether all
acts that yield a result that can be interpreted as discriminatory shall
be included. The issue of the other party's motive will be
particularly tricky in connection with gathering data from individuals
that perceive themselves as victims of discrimination. The legislation
(Sections 135a and 349a of the Penal Code) also seems to presuppose that
there must be a malicious intention. We cannot have an in-depth discussion
of this question, but it is crucial that we use a uniform conceptual
apparatus and standard definitions as the basis for a joint understanding
of the reality. It is often the case that the two parties in a
confrontation have completely different perceptions regarding whether the
incident in question contained elements of discrimination. It is certainly
not the case that a single party's opinion about discrimination is grounds
enough for an incident to be recorded as an instance of
discrimination.
There are a number of central ethical aspects related
to compiling statistics and performing analyses of immigrants or ethnic
minorities. I would like to mention two aspects here, because they are
inextricably intertwined with our topic.
Every now and again I am asked whether Statistics
Norway cannot simply allow immigrants to be included in the population in
general, as sooner or later they must be treated like the people that are, not
merely as members of a group (with immigrant background). My opinion is
that as long as immigrants are discriminated against or receive different
treatment of any kind, it is necessary to describe the results of this
discrimination or treatment. We will gladly stop keeping statistics on
immigrants when there are no significant differences. However, in my view,
stopping this work today would not serve to protect immigrants, rather it
would protect society against insight into the discrimination it is
actually exercising.
As an extension of this, it is also possible to query
which groups it is worth analysing. The available registers suggest a
whole range of possibilities, and there are also a number of theoretical
possibilities not available to us. The concept most commonly used in
immigration statistics in Europe is probably citizenship, but from the
perspective of integration and discrimination, changing one's nationality
does not make a huge difference. Many statistical institutes also use
information about the country in which a person was born, but the story of
my neighbour that I mentioned in my introduction illustrates that where a
person is born does not affect whether they are subject to discriminatory
acts or not. In some countries where there is a lot of immigration, with
Great Britain as a prime European example, people are asked regularly
which ethnic group they think they belong to. Defining one's own
ethnic group entails changes over time; for example, there are an
increasing number of immigrants that call themselves "Canadian", without
this appearing to be an obvious ethnic group. By asking people about their
ethnicity, we would also be able to operate with national minorities that
are otherwise ignored by a system based on immigration. However, whether
the national minorities themselves would think it was all right to be
asked about their ethnicity is a horse of a different colour. These
aspects are discussed in some detail by Østby (2002).
Norwegian population statistics are fully
registerbased. We have a Central Population Register (CPR) including
everyone with a legal permit to stay in Norway, and with a PIN-code that
can be used when linking information between different registers. CPR
contains vital events and migration, and can be arranged for life course
analyses. Based on CPR-information, Statistics Norway has for the last few
years, unless there have been good grounds to do otherwise, defined the
immigrant population as everyone who has two parents born outside Norway.
This gives a much more detailed picture than if we asked about ethnicity,
but it does not differentiate between Kurds and other people from Turkey,
for example.
One means of revealing discrimination that is used in
some places, but which most definitely has its questionable ethical
aspects, is to submit fake applications for jobs or housing and to analyse
differences in how these applications are processed in terms of ethnic
origin and other factors. These kinds of experiments can yield very
interesting results, but they involve analysis techniques that Statistics
Norway does not normally employ. In this way, we avoid having to form an
opinion about the moral dilemma of sailing under a false flag for a good
cause.
The criticism of Norway
The European Commission against Racism and Intolerance (ECRI)
I regard the criticism from the United Nations'
Committee on the Elimination of Racial Discrimination (CERD) and that from
the European Commission against Racism and Intolerance (ECRI) as
fundamentally the same, although ECRI's criticism was obviously much more
detailed. What is slightly more worrying is that we were criticised for
basically the same things in 1998 and 2000. For example, in section 19 of
the report for 1998 (Council of Europe 1998) on Statistics, it is stated:
As yet, no systematic registration of the main
manifestations of racial or xenophobic violence or harassment has been
carried out in Norway. .... Norway is encouraged to continue with its
efforts to put in place as rapid as possible a reliable and systematic
system of data collection in the field of racial violence and
harassment.
In the 2000 report (Council of Europe 2000), it is
stated (translated back from Norwegian):
K. Monitoring of the situation in the country:
42. Sensitivity about the collection of data
pertaining to ethnic origin has made the monitoring of the situation of
the different minority groups living in Norway problematic. In this
respect, ECRI draws attention to its general policy recommendation No. 1,
in which it highlights the importance of data collection to assist in
assessing and evaluating the situation and experiences of vulnerable
groups, and encourages the Norwegian authorities to further consider ways
in which such data might be collected while respecting the right to
privacy, data protection and free and informed consent of the persons in
question.
43. ECRI notes that the Directorate of Immigration
will present reports on the type and extent of racial discrimination in
1999-2000 and that the Centre for Combating Ethnic Discrimination
published its first such report in November 1999. One of the main
challenges in the future is to develop a registration system that will
enable the government to monitor racially motivated incidents.
44. ECRI also feels that the monitoring and regular
evaluation of the many projects and initiatives which have been set
underway in recent years should be given priority, in order to ascertain
their effectiveness and to disseminate good practices.
We can safely say that Norway has implemented the
general policy recommendations to a large extent, but there are grounds to
query whether these recommendations will actually help us attain good
statistics on racist and discriminatory incidents. Section 42 above and
the general recommendations are actually advising us to base our
monitoring on self-reported ethnicity. However, I believe that this
entails undeniable statistical problems. In order to be able to analyse
the various different groups' risk, we would need a full overview of all
the ethnic groups in the country - not only of those that classify
themselves as such. Self-reported ethnicity offers a number of advantages
in connection with recording incidents, and many people have strong
objections to establishing any form of register of such information. The
clearest advice in the general policy recommendation is related to matters
that are embraced by the legislation. The statistics on incidents of this
nature that are reported to the police would appear to be excellent
compared to those in many other countries, since we collect personal
information about both the victim and the perpetrator (cf. Gundersen 2000
and Gundersen et. al. 2000).
ECRI also has a special policy recommendation
regarding surveys that focus on perceptions of racism from the point of
view of potential victims (ECRI general policy recommendation No. 4). The
recommendations here are targeted at the acts in their own right and at
attitudes. The conclusion reads:
...recommends to the governments of member States to
take steps to ensure that national surveys on the experience and
perception of racism and discrimination from the point of view of
potential victims are organised, drawing inspiration from the guidelines
set out in the Appendix to this recommendation.
The Appendix referred to is too extensive to be
presented here. However, before the recommendations are translated into
concrete surveys, it seems reasonable that their implementation should be
discussed by the various actors in the field and statisticians. It lays
down guidelines that, as far as I am aware, are not observed in any
country and therefore do not provide grounds for comparative analyses.
Important aspects of the recommendations concur with current Norwegian
practice (Crime Statistics, Surveys of Living Conditions, demographic
statistics). Needless to say, the objective of being able to perform
comparative analyses is important, and ECRI also emphasises the importance
of establishing goals that provide grounds for comparisons over time.
Comparability requires uniform definitions of terms and uniform standards,
and it is surprising to many people how little it takes to render
otherwise good data incomparable.
The
United Nations' Committee on the Elimination of Racial Discrimination
(CERD)
In my eyes, the criticism that the United Nations'
Committee on the Elimination of Racial Discrimination levelled at Norway
(CERD 2000) does not seem quite fair. In section 12, they say:
The Committee notes that there has been little
progress in monitoring racial discrimination through record keeping of racist incidents,
indictments, sentences and compensation. The Committee maintains its
request for information on court decisions on racial discrimination and
recommends the State party to review its procedures for monitoring racist
incidents in order to increase their effectiveness.
I take this as criticism primarily of the way in which
statistics on reported offences are generated. I was completely unaware
that we had poor statistics on racist incidents, indictments, sentences
and compensation. In my opinion, the real problem is associated more with
incidents of a racist and discriminatory nature that are not covered by
the legislation, than with the registration of incidents embraced by the
legislation. The registration of these very few incidents is already there
for many years.
Statistics Norway is not entitled to have an opinion
about whether the legislation is pertinent, and personally, I lean towards
the view that the mesh of the net used to capture racist acts is a little
on the large size. An important objection that I would have expected CERD
to mention is the degree to which the legislation actually captures the
kinds of incidents that it was originally intended to prohibit. There are
probably many causes for the discrepancies, but this is not a problem that
can be resolved by means of statistics, and for this reason, we
statisticians should probably refrain from giving advice.
What is being done in other countries?
There is a great deal of work being done in this area
across the whole of Europe and in the USA and Canada - and probably also
in many other countries too. Below I will be summarising a small part of
the work that is being done. I do not have primary knowledge about these
projects gained from having studied other countries' systems directly;
instead I have borrowed from printed sources, either published by the
nations themselves or from the European Commission against Racism and
Intolerance's review of the countries. I do not doubt that there are very
many people who know a lot more about this area than me; my aim is simply
to give some examples as a basis for assessing the issue in light of the
requirements for good statistics presented above. The Council of Europe
(1998), ECRI's volume III, reviews six countries. None of the countries
manage to avoid criticism for their statistical methods, but the criticism
appears to be tailored according to what was and could be expected of the
particular nation in question. In other words, the threshold varies from
country to country, entailing that countries that have relatively poor
routines with regard to gathering and keeping statistics are judged milder
than countries that have already established good routines and keep
generally high statistical standards.
Bulgaria was given a cautious reminder that it is lacking
complete statistics over its (many) minorities, but otherwise avoided
criticism in this area.
France had its lists of "racist acts" approved, although, as
far as I can make out, the number of acts defined has only two digits. And
without statistics on ethnic groups or immigrants, which it is pointed out
that France is lacking, these kinds of figures do not serve any purpose
other than to act as an apology for blatant under-registration.
Italy has considerable deficiencies in its basic
statistics, which was commented upon, but which should not cause the
nation undue problems.
The Netherlands is probably subject to the same type of expectations
regarding good statistics as Norway. They received much harsher criticism
than we did, because the general statistical foundation there is data from
sample-based surveys with very poor data about immigrants. Racist
incidents are not mentioned in connection with the Netherlands, either
because the basis for interpretation was obviously inadequate or perhaps
because their RADAR system (Knippers 1997) was deemed satisfactory.
When the report was written, Portugal did not have this kind of data in its
statistics system, and the standard paragraph on statistics that is found
in most reports was absent from the discussion of Portugal.
Slovakia includes self-declared ethnicity in its censuses, but
is nevertheless subject to criticism, in particular because ECRI holds
that many people do not dare state their true ethnic identity in the
censuses. In addition, they update their refugee statistics on a daily
basis, but the figures are not published. They were reminded that it is
generally a good idea to publish statistics, not just collect them.
Statistics that are not used are never good.
Having read ECRI's review of the countries, I can only
conclude that Norway has relatively little to learn from this report when
to comes to improving our own systems. In my opinion, the European
Commission against Racism and Intolerance and the United Nations'
Committee on the Elimination of Racial Discrimination have an unfounded
faith in self-declared ethnicity and in systems based on reported criminal
offences or client-registered racist incidents. I think far too little
importance is attached to the prerequisite for achieving comparative data
in these recommendations, despite the fact that comparability is ascribed
priority in other and more general contexts. In the absence of absolute
measures, comparative data is very useful, especially for policy
purposes.
The Netherlands clearly has a useful system, called RADAR (Rotterdamse Anti
Discriminatie Actie Raad [the Rotterdam anti-discrimination action
council]), which is an office charged with initiating measures, as opposed
to gathering statistics (Knippers 1997). We will learn more about that
during the Sixth Metropolis Conference. The work performed here also seems
to have served as a model for certain Norwegian bodies, and the main
sources of information are:
- complaints from victims
- reports from people and institutions that come into contact with
victims, such as:
the police, the anti-discrimination ombudsman, emergency wards,
hospitals, crisis centres, graffiti removers, etc.
It will thus be primarily the most serious incidents
that are registered by this system, but a good number of less serious
incidents will also be included by the various groups of people reporting
incidents to the system. This kind of system may well form an excellent
starting point for helping victims and implementing corrective and
preventive measures, and it will indicate the level of and trends in the
kind of discriminatory incidents and acts that this kind of apparatus
encounters. It will also be able to provide information about the nature
of discrimination in general. However, it does not seem to be equipped to
determine either the level or trends in discrimination in society as a
whole. Furthermore, it is often difficult to undertake this kind of work
equally thoroughly throughout the country, which must surely be one of the
prerequisites for improving the weaknesses pointed out by the European
Commission against Racism and Intolerance and the United Nations'
Committee on the Elimination of Racial Discrimination. Good offices will
record much more than offices that people feel there is no point in even
approaching.
Great Britain too has a well-developed system that is
run through local offices that are found in most towns and regions. Again,
I hope for more information during the Metropolis Conference. Needless to
say, there is variance in the size of and expertise in the offices in
different parts of the country (Lunde 1994). There are many very good
reasons to consider this type of work when creating a national statistical
system for recording racist incidents and acts - not least since these
kinds of local offices have good contact with a sample of victims.
However, this sample population can vary quite considerably. We can only
hope that the tendencies towards co-ordinated registration of incidents
will continue and be further developed so as to avoid double counting and
in order to ensure that all incidents over a defined degree of gravity are
recorded. In this kind of project, it seems only logical to bring in
statistical experts from the fields in question to establish uniform and
operational definitions, for quality assurance and to relate this work to
other statistics.
The British have drawn the conclusion that the better
a local group is at defending a discriminated person's case, the more
discriminatory acts are recorded in the area. Areas with poorly developed
support groups scarcely record any discriminatory acts at all - without
this necessarily indicating that there is no discrimination. Grattet
(2000) reached a similar conclusion for so-called Hate Crimes in the USA.
Denmark is now planning to implement a system similar
to the one in the Netherlands and Britain in collaboration with various
voluntary organisations. This comes after the European Commission against
Racism and Intolerance explicitly recommended this course of action in its
report on Denmark for 2000, and repeated it in 2001.
Do any
of these systems fulfil the requirements?
Of course, I am in no position to draw categorical
conclusions about whether other people's work is good enough. There are
many countries that have extensive, well-constructed systems to record
discriminatory acts, both as a ground to help individuals and in order to
influence attitudes in the community. However, if we consider the systems
and outcomes in relation to a defined goal of being able to describe
levels and trends in a way that covers all the developments of
discrimination in society and provide a foundation for comparative
analyses, very few (if any) of the countries meet the mark. The situation
is thus not so clear cut that we can simply import an existing system and
use it for all or most of our own good intentions. The job ahead is
demanding, but - and here I am repeating what I said on page 2 - it is not impossible!
What is being done in Norway?
I would also like to say a few words about some of the
important monitoring schemes that are underway in Norway today.
Organisations performing this kind of work include the Centre for
Combating Ethnic Discrimination (SMED), the Organisation against Public
Discrimination (OMOD), the Anti Racist Centre (ARS) and the Directorate of
Immigration (UDI), whose report is the raison d'être
for this meeting. There are also numerous local initiatives, of which one
employs a system that appears to comes close to the kind of system that
the European Commission against Racism and Intolerance wants, namely the
Contact Forum between immigrants and the police in Stavanger that was
started in 1995 (Kontaktforum 1995).
Working group on systematic collection of data
As early as in 1992, a working group was established
to compile a report on systematic collection of data on racial violence
and harassment. The working group consisted of representatives from the
immigration authorities, the internal affairs authorities (Ministry of
Justice), voluntary organisations and Statistics Norway (see the final
report of the Working group (1992): Recommendations and proposals).
In the press release about the final report, it was
stated:
We need better documentation and registration of
violence and other forms of harassment that is due to skin colour,
religion or national or ethnic origin. In Norway, there is no real
systematised knowledge about the situation in this area. The working group
therefore suggests improved police records of discriminatory incidents and
research projects as means to detect various previously hidden
figures.
Little has been done about some of the matters that
were regarded as requiring follow up according to the group. However,
there have been marked improvements in respect of statistics and better
police records. The central
police registration system STRASAK has been improved and can now
produce detailed statistics about the victims of crimes as well as the
perpetrators (Gundersen 2000). To date, this has only been done for 1998,
and there is uncertainty about whether this work is to continue or not.
The new data that has been collected allows for far more analyses than
have previously been performed. However, this system only covers cases
that are officially reported to the police, meaning that it does not
record perhaps the majority of discriminatory acts. Nevertheless, it might
be interesting to see whether the United Nations' Committee on the
Elimination of Racial Discrimination will accept that the data they have
requested now exists - at least for one year.
The working group also suggested that the police are
instructed to keep their own records of racist elements in all relevant
offences. I have not been able to ascertain whether or to what extent this
has been implemented. The aim of uncovering the "dark figures" that were
previously hidden can only be achieved by looking beyond the Crime
Statistics, because no matter how good these statistics are on their own
premises, a key problem will always be the incidents that are not recorded
in the statistics, either because of underreporting to the system or
because the legislation does not catch all the acts that might be
relevant.
Papendorf 1998
The work of the working group was followed up by a
report published by the Department of the Sociology of Law at the
University of Oslo (Papendorf 1998). This report too seems to attach a
great deal of importance to the trends demonstrated by the Crime
Statistics. An advantage of this type of source is that the criteria for
inclusion are fixed and verifiable, but there is still a considerable
degree of under-registration of acts that might have been covered by the
legislation.
Papendorf has an interesting discussion of the
methodological problems, in connection with which (on page 8) he quotes
article 2 of the International Convention on the Elimination of All Forms
of Racial Discrimination:
In this Convention, the term "racial discrimination"
shall mean any distinction, exclusion, restriction or preference based on
race, colour, descent, or national or ethnic origin which has the purpose
or effect of nullifying or impairing the recognition, enjoyment or
exercise, on an equal footing, of human rights and fundamental freedoms in
the political, economic, social, cultural or any other field of public
life
It is thus the results of discrimination that are critical, not
whether the grounds that are given can be regarded as acceptable. This
paves the way for interesting perspectives on ways to measure racial
discrimination. The problem with most of the systems we have reviewed thus
far is that the intentions, motives, justifications and perceptions have
been more important than the outcome.
The
Council of Europe's "Racial violence and harassment in Europe"
There is a whole series of publications that deals
with racist incidents in individual countries. There are also a number of
reports on Norway. The Directorate of Immigration published its
"Racially-motivated violence and harassment in Europe" in 1999, which is a
translation of the Council of Europe's publication from 1992. The
underlying view in this document is that one important source of data is
the relevant grouping of crimes. The other main source consists in
different forms of registration of incidents, whereby the victim reports
the incident to a registration body, perhaps via someone in a treatment
centre or support agency. This can scarcely meet the requirements we have
laid down for good statistics, but still, we must not forget that an
important aspect in the battle against incidents of this nature is to
record them.
The Directorate of Immigration's report "The Nature
and scope of racism and discrimination in Norway
1999-2000"
This report (UDI 2000) is based primarily on acts that
have a discriminatory effect. The goal of the report was to describe the
reality as fully as possible, but this goal still seems to be a long way
off in relation to the somewhat ambitious title of the report. It covers
29 municipalities, which have 75 % of the relevant immigrant population,
and collects information from the local authorities and from
representatives for immigrants. This probably provides the Directorate of
Immigration with crucial and useful information for its own work, but it
is not satisfactory as a basis on which to describe the nature and
especially the scope of racism and discrimination in Norway. Perhaps the
report should simply be given a different title.
The report contains key informants' subjective
perceptions of the nature and extent of discrimination, with the explicit
ambition of ascertaining whether the municipalities
believe that discrimination explains the immigrants' problems (page 16).
The report presents very many very interesting observations and is
probably useful for the Directorate of Immigration in its contacts with
the municipalities, but it does not appear to be the solution for the
problem areas that the European Commission against Racism and Intolerance
and the United Nations' Committee on the Elimination of Racial
Discrimination have pointed out in their criticisms of Norway. However, if
this information was collected repeatedly over several years using the
same sample municipalities, it may well serve to reveal a number of
trends.
Moving Towards Better Protection 2000
The Centre for Combating Ethnic Discrimination (SMED
2000) has published a report of a slightly different type. It is called
"Moving Towards Better Protection 2000" and is their contribution to
knowledge about the nature and scope of ethnic discrimination in Norway.
To borrow some terms from social medicine, this is a report based on cases
(incidence) as opposed to frequency (prevalence). However, both types of
information are important in this work. We need both types of data to be
able to understand the processes and choose the most appropriate measures.
We need to know the actual number of incidents and we need to know the
size of the at-risk group, i.e. the people that might be
affected, in order to find out what needs to be done and what needs to be
done first.
The Centre for Combating Ethnic Discrimination was
founded against the backdrop of recommendations made by the European
Commission against Racism and Intolerance and the United Nations'
Committee on the Elimination of Racial Discrimination and is clearly doing
a massive job in terms of providing legal aid and changing attitudes. I am
sure that they keep excellent documentation of their clients' cases for
their own purposes, but I am still in doubt about the role of interested
parties when it comes to keeping statistical records. The Centre's clients
need all the support that it can give them; they need good defenders and
advocates. However, it is not always easy to combine defending interests
with the role of documentalist. Measuring a phenomenon when one also has
interests connected to the results of the measurement can be likened to
selling elastic by the metre.
In its report from 1999, the Centre for Combating
Ethnic Discrimination says regarding documentation (page 17):
According to Report no. 17 to the Storting (1996-97) on immigration
and Norway as a multicultural society, to date it has been difficult
to say very much about the scope of racism and discrimination in Norway.
Is it a growing or declining problem? Have there been changes in the
nature of discrimination? These questions have been and will always be
difficult to answer. Insufficient case law is one of the reasons that we
know so little about these issues. Another important element, according to
Report no. 17 to the Storting, is the lack of other types of documentation
and the fact that there are a number of methodological problems that make
it difficult for us to ascertain the nature and the scope of ethnic
discrimination.
The Centre for Combating Ethnic Discrimination regards
recording information about the nature and extent of ethnic discrimination
in Norway as one of its main tasks. However, the two questions posed in
the excerpt cited above cannot be answered on the basis of the information
presented in the Centre's report. The Centre for Combating Ethnic
Discrimination probably lacks the expertise and experience in keeping
statistics that a task of this nature requires. The Centre's ambition is
to document matters concerning discrimination in Norway. Many others will
probably be uncomfortable with the kind of description the Centre is
likely to produce and will try to explain it away. It is in precisely this
kind of situation that it is particularly important that descriptions are
firmly anchored in the necessary statistical competence.
The type of documentation that enables us to analyse
levels and trends requires that a large amount of information be collected
about each individual incident, meaning that the case and the people
involved can be identified. However, it is not necessary to include every
single complaint. A rational approach is to filter out unreasonable
complaints so that only incidents that according to a fairly standard
assessment are deemed to constitute discrimination are included. The
criteria must be precise enough and the information good enough to allow
different people to reach the same result when considering the case.
Neither the motive for an act nor the effect of the act can be classified
as discriminatory on the basis of the victim's judgement alone.
What
should we be doing?
In this section, I will discuss the following
rhetorical question, even if you and I all know that I believe the answer
is a resounding "yes":
Should all the interests of public, semi-public and
private actors in this field be better co-ordinated to ensure that all the
resources that are available for description of the phenomenon are used
efficiently to achieve a common goal? Should we pool our resources to
build a common house that can accommodate everybody's needs or should we
each build our own small house?
In my opinion, there is a need for co-ordination both
between the various agencies in Norway and not least between nations.
Furthermore, I think that the responsible Ministry should take the
initiative for this kind of co-ordination - both nationally and
internationally.
The work that is currently being done in this field -
by activists and government bodies alike - covers several important needs.
However, we will reap little fruit from all this good work if we only
focus on the issues related to the statistical perspective as I have been doing. In
order to compile and get the most out of this type of statistics, they
must be compiled in a co-ordinated collaboration, even on the
international level. It is also essential that the national records are
based on standard data that operate with uniform definitions and a single
joint system for identification of incidents in order to avoid double
registration.
I think that the international co-operation should be
organised under the auspices of a body that has statistical competence. Of
the various bodies that fit the bill on the European level, I would
personally prefer Eurostat (the European Commission's statistical office)
to be in charge of this work, as opposed to the Council of Europe, the
United Nations Economic Commission for Europe (ECE) or specialised
organisations, such as the European Monitoring Centre on Racism and
Xenophobia (EUMC). This view is not based on an in-depth analysis of the
individual actors' strengths and weaknesses, but I have worked
sufficiently on systems and comparative surveys to have reached the
conclusion that it is more important for a body charged with these kinds
of tasks to have expertise in statistics than insight into the subject
area. Most countries' national statistics systems are organised on similar
principles. Another advantage of putting Eurostat in charge of this work
is that they have the best contact with all the national statistics
agencies, the European Union, EFTA and the 12 countries applying for EU
membership.
The underlying principle of many of the reports that I
have reviewed here, i.e. the principle of self-reporting by people that
believe they have been subject to discrimination, is problematic as a
foundation for compiling statistics, especially if they are to be used in
a comparative perspective.
The objective of the Ministry of Local Government and
Regional Development's plan of action against racism and discrimination
(KRD 1998), under the section titled "Measures - Legal aid and monitoring"
is: "...
to gain better knowledge about the nature and scope of
discrimination and racially motivated acts." The status report from spring
2000 (KRD 2000) mentions two measures that are relevant in our
context:
1.2.1. The Directorate of Immigration is going to
develop a system for reporting based primarily on the Directorate's own
sources and in close contact with the Centre for Combating Ethnic
Discrimination. These reports shall provide a picture of the extent of and
developments in racism and discrimination in the municipalities and what
the municipalities are doing to counteract discrimination. The findings
will be reported to the Ministry of Local Government and Regional
Development annually.
Status: The first report (available).
1.2.2. The Ministry of Justice is going to assess on
an ongoing basis ways to improve statistical work with a view to providing
the best possible documentation and statistics about racially motivated
crime.
Status: To date, the Ministry of Justice has not
assessed improvements in statistical work.
This constitutes fairly modest progress, since I
believe I have demonstrated that both the reports that have been submitted
under 1.2.1 - despite some obvious strong points and their new
descriptions of the situation - make only limited contributions to
achieving the cited objectives. Beyond these two points, the objective of
monitoring the situation seems to be absent in the other areas covered by
the plan. If the system I proposed earlier for co-ordinated activity to
establish a better description of discrimination in Norway is implemented,
this would also serve to correct a significant deficiency in the
government's plan of action.
Suggested solution
Having reviewed a number of Norwegian and foreign
systems, I can only conclude than none of them appear to constitute a
suitable template for a system to meet our needs. I would therefore like
to propose sketch of a system that can be discussed in more detail to see
if it offers some possibilities that can be developed to incorporate the
considerations voiced by several governmental and non-governmental
organisations, and also meet the statistical requirements. My starting
point is article 2 of the United Nations' International Convention on the
Elimination of All Forms of Racial Discrimination, which focuses on the outcome of
discrimination rather than the grounds for it. As discussed above, we cannot say
outright that differences in the outcome are the result of any form of
active or conscious discrimination, but in some contexts, the result is
more important than the grounds. It is also important to chart differences
in outcomes that are not the result of active discrimination, and this
charting will then provide grounds for assessing the gravity of the
incident and the need for corrective and/or preventive measures. The fact
that I attach greater significance to the results than the victim's
perception of the background places quite a distance between me and the
main lines in most of the works I have referred to. Nevertheless, I cannot
see any other relatively simple method that will be able to provide us
with reliable analyses of the level of discrimination or trends in its
development. I would like to reiterate my earlier suggestion that the
monitoring that the Centre for Combating Ethnic Discrimination, the Anti-Racist Centre, etc.
performs be co-ordinated internally and also be co-ordinated with the
information stored in the police computer systems.
I would now like to present for your assessment a
model for the statistical work that is capable of amending some of the
points on which Norway has been criticised and that meets many of the
needs that the country has for knowledge in this area. It is scarcely only
our neighbours that rue the lack of reliable information. My proposal is
intended as a supplement to the work that is already being done
and can be achieved by means of client-registered incidents. I am
suggesting two types of survey, both of which are intended to be representative, in the sense that they provide a
picture of the underlying reality without any known biases.
1) Analyses of statistical data in registers
2) Questions in a Survey of Living Conditions for
immigrants
The main perspective in both types of survey is the
"effects of discrimination"; with survey type 2, we also get an account of
experienced incidents and an assessment of them.
1) The main basis for these kinds of analyses is
registers that are established for purposes other than statistics, but
that the Statistics Act provides can also be used for statistical
purposes. Data about identifiable individuals can be used in analyses, but
of course must not be used in such a way that the individuals can be
identified. The most important register in this context is the Central
Population Register (DSF), which contains information about every single
person that has the right to be registered as living in Norway and
ascribes a personal identity number to all residents. This personal
identity number is the key to all the other registers based on individuals
from which we might want to gather information. If a register does not
contain people's personal identity numbers, the data it contains cannot in
any reasonable way be linked to other information about the individuals.
Among the most interesting registers for our purposes
are registers about:
·
Income and wealth, including sources of income and tax
information
·
Education - ongoing education and the population's
highest completed education
·
Employer / employee - participation in the labour
market and working conditions, unemployment
·
The police register of criminal cases.
With the approval of the Data Inspectorate, these
registers can be linked up and used together for statistics and analysis.
This will constitute the most important foundation for statistics on
immigrants, which by now has come on an acceptable level. The most
lamentable deficiency is perhaps our lack of information about immigrants'
language skills.
2) Questions can be asked in interviews with a
representative sample of (groups of) immigrants. Non-response is a
recurrent problem with all surveys of this nature. It is especially
important to interview immigrants in their own language to ensure that we
do not only get to hear the views and experiences of immigrants that have
a reasonable command of Norwegian. Of course, this will make any survey of
this nature labour-intensive and costly, but there is no simpler method
that will yield good enough data. Statistics Norway carried out an
interview survey among immigrants on behalf of the Ministry of Local
Government and Regional Development in 1996 (Blom 1998). The survey was
based on the questionnaire used in the annual, general Survey of Living
Conditions. (The number of immigrants included in the general survey is
far too low to allow the data to be used for our current purposes.) This
survey also included a handful of questions about discrimination, which
served our purpose. If similar surveys are carried out again in the
future, it might be useful to focus more on experienced discrimination.
The strength of this type of survey is that they can
be made comparative: between countries, between minorities and majorities,
over time, and between different subgroups within both the majority and
the minority populations. This type of survey also allows us to chart
attitudes (an area where data has previously been lacking), the
minorities' relationship to the majority population, contact patterns,
participation in the community, etc. Another considerable area of weakness
that this method cannot help correct is insight into immigrants' language
skills. Self-evaluated command of languages is probably the only
information we will be able to collect by means of this method, which is
scarcely good enough for all our purposes.
Allow me also to mention an example of analyses in
which these two types of data can be used together. An interesting
question is whether the housing market constitutes a problem only when
access is limited as a result of discrimination, or is a tight housing
market a problem in its own right? I would imagine that a tight and/or
expensive housing market would affect immigrants more than others and that
discrimination in this market is an additional, separate problem. We ought
to be able to establish a basis on which we can include the entire problem
area in the analysis and illustrate immigrants' overall, general problems
in the housing market and give an estimate of how large a percentage of
these problems are due to discrimination. We must distinguish between
those problems that affect immigrants as "disadvantaged" actors in a tight
housing market in Oslo and those problems that affect them because they
are immigrants. If ten people voice an interest in renting a house, the
nine that do not get the house are not necessarily victims of
discrimination, but that does not mean that there was not any
discrimination either. Similar observations can be applied in respect of
the labour market, but here one would imagine that it would be easier to
monitor employers' employment policy than property owners' preferences.
Statistics based on registers can only tell us something about the outcome of the
processes in the housing and labour markets, but this information may well
be significant enough in its own right.
Conclusion
Clear complaints have been expressed by both the
United Nations and the Council of Europe about the statistics Norway has
on racist acts and incidents, and there are also people within Norway that
claim that the status in this area is inadequate. My opinion is that the
critics are right to a large extent, but that the advice they are giving
will not help us achieve the goals - neither the goals that they have set nor the slightly more general goals
that I have tried to relate to in this paper.
This is an area marked by little knowledge and a great deal of strongly
held (positive and negative) opinions. The situation was roughly the same
for all the immigration statistics some ten years ago, but a concentrated
and long-term effort may give better results.
Our critics have not made any particularly useful
contributions that will make the situation fundamentally better - at least
from the perspective of purely statistical interests. I believe that the
proposals I put forward at the end of my presentation represent a
realistic system that will allow us to establish a monitoring system that
is perhaps not 100 % perfect, but that will cover essential factors that a
monitoring system of this nature should cover. It will serve to counteract
some of the criticism that has been levelled at us, but not all of it.
However, that
would be neither right nor possible.
The statistics that can be established by means of the
suggested systems are intended to constitute one step towards building a
basis on which policy can be designed and especially to allow us to check
the impact of policy measures. Quantified personal opinions based on a
body of data whose representativity is unknown is often an excellent basis
for the formulation of hypotheses, but cannot be used to test these
hypotheses empirically.
References
Arbeidsgruppe for systematisk datainnsamling om
rasistisk vold og trakassering (1992): Forslag og anbefalinger.
Avgitt 23. desember 1992. [The Working Group on systematic collection of data on
racial violence and harassment: Proposals and recommendations. Report
submitted on 23 December 1992.]
Blom, S. (1998): Levekår blant ikke-vestlige
innvandrere i Norge. [Living conditions among non-western immigrants in
Norway]. Rapporter 98/16, Statistisk sentralbyrå,
Oslo-Kongsvinger.
CERD (2000): The United Nations' Committee on the
Elimination of Racial Discrimination, observations from the 57th meeting:
CERD/C/57/Misc.44/Rev.3 23 August 2000)
Council of Europe (1998): ECRI's country-by-country
approach: Report on Norway. CRI (98) 24
Council of Europe (1998): ECRI's country-by-country
approach: Volume III. CPI (98) 54
ECRI General policy recommendation No. 4: National
surveys on the experience and perception of discrimination and racism from
the point of view of potential victims.
www.ecri.coe.int/en/02/02/07/e02020701.htm
ECRI (2000) Report on Denmark.
www.ecri.coe.int/en/02/02/05/e02020583.htm
Grattet, R. (2000): Hate Crimes: Better Data or
Increasing Frequency? Population Today, Volume 28, Number 5 July.
Population Reference Bureau
Gundersen, F. (2000): Mer vold enn ventet [More violence than expected].
Samfunnsspeilet 3/2000, 3-15.
Gundersen. F. et. al. (2000): Innvandrere og nordmenn
som og gjerningsmenn. [Immigrants and Norwegians as victims and
perpetrators]. Rapporter 2000/18, Statistisk sentralbyrå,
Oslo-Kongsvinger.
Lunde, H. (1994): Det engelske svaret på rasistisk
vold og trakassering. [The British response to racist violence and
harassment]. Antirasistisk Senter, dokument 1994.
KRD (1998): Handlingsplan mot rasisme of
diskriminering (1998-2001). Lagt fram 15. juni 1998. [The Ministry of Local Government and Regional
Development: Plan of action against racism and discrimination (1998-2001).
Submitted on 15 June 1998.]
KRD (2000): Handlingsplan mot rasisme og
diskriminering. Statusrapport fra våren 2000. [The Ministry of Local Government and Regional
Development: Plan of action against racism and discrimination. Status
report from spring 2000.]
Knippers, K. (1997): Recording and monitoring
discrimination and racism: Experiences from Radar, Rotterdam. Paper
prepared for the Seminar "Interfere Racism-Prevent Discrimination" 3-4
April 1997, Helsinki
Kontaktforum mellom innvandrere og politiet (1995):
Rasistisk trakassering og vold - "offerets
perspektiv". Stavanger. [Contact forum between immigrants and the police: Racist
harassment and violence - "The victim's perspective".]
Papendorf, K. (1998): Bistand i forbindelse med
diskriminering. Kartlegging av problemet og forslag til tiltak.
[Aid in connection with discrimination. Charting the
problem and proposed measures]. Institutt for rettssosiologi, Oslo.
Rogstad, J. (1996): Etikk og metode - noen problemer i
studiet av diskriminering. [Ethics and methods - some of the problems in the study
of discrimination]. Sosiologisk tidsskrift 3, 197-218.
Rogstad, J. (1998): Diskriminering - hva er det og
hvordan avdekke det? [Discrimination - what is it and how to detect
it?] Søkelys på arbeidsmarkedet, årgang 16, 125-132.
SMED (2000): Underveis mot et bedre vern 2000.
Oktober. [The Centre for Combating Ethnic Discrimination: Moving
Towards Better Protection 2000. Oktober.]
UDI (1999): Rasistisk motivert vold og trakassering i
Europa. Integreringsavdelingen. [The Directorate of Immigration: Racially motivated
violence and harassment in Europe. The Integration Department.]
UDI (2000): Art og omfang av rasisme og diskriminering
i Norge i 1999-2000. September. [The Directorate of Immigration: The nature and scope
of racism and discrimination in Norway in 1999-2000.
September.]
Working Group (1992): Arbeidsgruppe for systematisk
datainnsamling om rasistisk vold og trakassering (1992): Forslag og
anbefalinger. Avgitt 23. desember 1992. [The Working Group on systematic collection of data on
racial violence and harassment: Proposals and recommendations. Report
submitted on 23 December 1992.]
Østby, L. (2002): Why Analyse Immigrants? Ethical and
Empirical Aspects. Yearbook of Population Research in Finland 38 (2002),
pp.127-145
