Measurement and indicators of

integration

Community Relations
Measurement and indicators of
integration

Community Relations

Directorate of Social and Economic Affairs

Council of Europe Publishing

 Introduction

Measuring social behaviour and social phenomena always is a very challenging task.
This is especially true when it comes to evaluating the integration of migrants into
their host societies, because it means in fact evaluating two social processes: One
cannot look at the migrants alone, but also has to take the members of the host
society into consideration.

As with any undertaking in social sciences, the evaluation or measurement of

integration has to begin with a definition of the basis terms. In the papers submitted
by the experts as well as in the discussion during the meeting it became clear that
clear-cut definition of:

a. who belongs to the target group of integration policies and

b. what exactly is meant by the term "integration"

are of great importance. Without common standards as to what is meant by "migrant"

and by "integration" all attempts to measure migrants' integration in different
countries are likely to be of little meaning.

Defining basic terms can at times be complicated but finding common definitions of
the crucial terms "migrant" and "integration" is made even more difficult, because
national integration policies heavily influence the understanding of "migrant" and
"integration".

1. Concepts of what integration means and is to achieve differ.

These differences are reflected in the national policy goals and
range from next-to-assimilation to multiculturalism.

2. National integration policies define different target groups for

their integration measures: Not everybody who has come from
abroad is a member of the target group (e.g. not formally
recognized refugees), not everybody who is part of the target
group has ever migrated from abroad (e.g. the second
generation).

5
In addition to this, the national focal points for fields of integration differ according to
the situation of migrants in a given country. This has a heavy bearing on the section
of indicators or integration. The selection of indicators and the problems of
measuring social behaviour were dealt with in an in-depth paper that the consultants
presented to the Specialist Group on February 21st 1995.

From 4 to 6 October 1995, almost sixty experts from more than twenty different
countries were invited by the Council of Europe to gather in Strasbourg in order to
evaluate the prospects of better measuring migrants' integration in Europe by using
indicators or integration. Their task was not an easy one, but despite the difficulties
described, the participants of the meeting have come to a set of common
conclusions on indicators of migrants' integration.

These conclusions are, on the one hand, based on the sixteen different presentation
documented in this volume which deal with national concepts of integration as well
as with attempts to measure integration in various countries. On the other hand, they
also are the result of the many animated and knowledgeable discussions which,
unfortunately, cannot be included as well.

The meeting in Strasbourg has shown how much there has already been done in the
field of identifying indicators of integration for measuring migrants' integration.
Although it has just as well proven the need for more steps to be taken, it has been a
very pleasant and fruitful experience.

1. Whose integration is to be examined?

One of the major obstacles for finding common ground when discussing the
integration of migrants is to define the fundamental terms. Neither "integration",
"migrant" or "migration" are clear-cut terms.

International migration, though easily defined as the medium to long-term movement

of persons from one country to another involving a change of residence, describes a
large variety of phenomena making topologies or classifications necessary. Some of
these classifications refer, for instance, to the migrant's (economic or political)
reasons to leave his country of origin. Other attempts to categorize migratory
movements or migrating persons were made along the lines of regular versus

6
irregular, voluntary versus forced or permanent versus short-term migration.

The term "migrant", too, is a broad one: Migrant workers, family members, asylum
seekers, political refugees (recognized as such after an administrative procedure) as
well as other refugees (accepted or tolerated without any formal procedure) or
irregular/illegal immigrants all fall into the category "international migrants". Other
definitions - neglecting the requirement of a change of residence and thus a change
of the social environment - also include persons participating in international student
exchanges, seasonal workers or cross-border commuters.

Each discourse on migrants' integration has to take this large spectrum into
consideration. Furthermore, it has to, as a first step, define exactly whose integration
will be examined. In this context a first problem of criteria for inclusion or exclusion
arises. Only the criterion of a change of residence and social environment - thus
leaving tourists and commuters out of consideration - may be suitable for
compromise. But all other criteria are likely to cause dispute: Does it make sense to
look exclusively at regular migrants, because they - from the government's point of
view - "deserve" integration into the host society? Would irregular migrants have to
be excluded, because their - reluctant -host countries refuse to offer any integration
policy measures to them? Or will they have to be included, because they themselves
intend to stay for good? How about recognized refugees? Aren't they expected to
return to their home countries once the political conditions allow for repatriation?
Even more difficult is the case of war refugees or refugees who aren't - for what ever
reason - recognized as such, but who, nevertheless, are allowed to stay, because
sending them back is considered inhuman or impossible.

Defining the "objects" of integration measures and their analysis becomes even more
complicated when one takes into account that some countries - with good reason to
do so -extend their integration policies to persons who are not migrants in the original
sense of the word, because they never moved from one country to another.
Especially children and grandchildren of migrants fall into this category. Even though
members of this group may have acquired the citizenship of their parents' country of
immigration on request or when they were born, they may, nevertheless, be
confronted with difficulties due to their ethnic origins. The same is true for adult
foreigners who were naturalized. They have obtained a new passport, but they have
not changed their ethnic, cultural or religious roots. Neither did their naturalization
change their professional qualifications or their language skills. Consequently, they

7
still might - labelled as ethic minorities - be target persons for measures aimed at
improving their integration.

Another problem of selection is caused by immigrants who come from foreign

countries but who either already possess the citizenship of their new home country or
acquire it more or less automatically upon arrival. These immigrants certainly have a
legal advantage over other migrants, but they nevertheless have to adapt to a society
they are not familiar with. Thus, they, too, belong to the target groups for integration
measures. This is especially true for so-called ethnic migrants who are descendants
of emigrants settling in Eastern Europe several generations ago. Ethnic Germans or
ethnic Greeks coming from the former Soviet Union face considerable difficulties of
social, economic and cultural adaption despite their cultural roots and their legal
privilege.

2. What is integration and who is to define it?

A satisfying compromise when identifying target groups will be difficult to reach, not
only because of the various types of migrants or because researchers' individual
interests differ, but also because of different concepts of integration as reflected in
the national integration policies. The member states of the Council of Europe all
have their own migratory traditions and migration policies shaping a country's
fundamental assumptions on the integration of foreigners. Some countries have long
standing immigration traditions (France, United Kingdom), whereas others have only
just recently become countries of immigration (Spain, Italy). A third group has few
problems with immigration, but faces considerable difficulties with indigenous ethnic
minorities (Romania, Slovak Republic).

And even although some groups of countries share common migratory histories -
such as the countries recruiting "guest workers" in the 1960s - all governments
pursue different approaches when designing their integration policies. Among the
more "experienced" countries of immigration, for example, the Netherlands are
characterized by a very active integration and "minorities" policy, whereas Germany's
approach is more restrictive when it comes to legal and political integration of
foreigners.

The national policies often reflect different definitions of what is meant by

8
"integration". While the term itself means "joining parts (in) to an entity" its practical
interpretation and social connotation may vary considerably: "Assimilation" as well as
"multicultural society" may be considered synonyms or descriptions of (successful)
integration. Thus, all forms of cultural or social behaviour ranging from completely
giving up one's background to preserving unaltered patterns of behaviour are
covered by the term of integration. This problem of definition, of course, has a
bearing on measuring integration, because the requirements for success in
assimilation are much more difficult to meet than requirements for multicultural co-
existence in a society which remains indifferent about other people's rites or
customs.

This, in turn, is important when it comes to comparing the integration of migrants in

different states and societies: Since different definitions of what integration means
form the basis of the national policies for improving migrants' integration, the
standard of when integration can be considered successful varies.

It is because of these differences that the principal concepts of integration and the
different national policies resulting from these concepts need to be looked at more
closely, because they form the background for evaluating migrants' integration.

3. Dimensions of integration

But whatever definition or concept of integration applied, one will agree that the
integration of migrants into their respective host societies has at least three basic
dimensions concerning the social, economic and cultural role migrants play in their
new environment. While these three dimensions will hardly be disputed by anybody
as important fields of integration, a forth dimension, namely the role migrants play in
political life, very much depends on whether the host government allows political
participation or even grants voting rights. The political dimension of integration is
often rejected as irrelevant by states disliking the idea of granting political rights to
migrants with a foreign passport.

Despite these reservations the political participation of migrants should be taken into
consideration when assessing their integration, because political life represents an
integral part of a person's life. Besides, political participation comprises more than
the right to vote or the right to be elected for political office. It also includes the right
to express one's opinions and beliefs. Political participation furthermore covers the

9
foundation of associations or the membership in political parties.

Leaving it out of consideration would thus mean ignoring important facets of

migrants' integration. It occurs rather often that another important factor is not taken
into account when the integration of migrants is discussed: the host society with its
value system as well as its prevailing attitudes concerning the presence of migrants
forming the background of migrants' integration. Integration is not a one-way-street
leaving the burden on the shoulders of the migrants alone. It is a social process
involving both sides, the migrants and their host society. Opinions, behavioral
patterns and attitudes in the host society can be decisive for the integration of
foreigners, because they can strongly influence the migrants' integration efforts.

If, for example, the host society remains hostile towards persons coming from
abroad, migrants are likely to live in their own neighbourhoods reducing their
contacts to the outside world to a minimum of unavoidable contacts. In the end,
migrants do not integrate, but live in a parallel social system of their own. If, on the
other hand, the host society is a very open one with permeable social stratifications
and a cultural life characterized by a large diversity, migrants will find it easier to
integrate, i.e. to find their place within such a society.

4. Indicators of integration

Since integration concerns complicated phenomena and refers to a very widespread

field, one cannot restrict the evaluation of a progress (or non-progress) in integration
to one single unit of measurement. There is no such thing as "one meter of
integration" or "two kilos of integration" which would make comparisons over time
and/or between two countries an easy task.

As a way out, one can only try to identify facts and phenomena giving an impression
of the current social, economic, cultural and political role migrants play in a given
society as well as at changes occurring over time. These indicators of integration can
be selected from all four dimensions of integration. But in addition to this the host
society - as was already mentioned - is very important as well, because public
opinion - though only measurable with a good deal of doubt left behind - can for
example give a hint as to the willingness to accept immigrants.
Sometimes, violent attacks on foreigners are taken as an indicator of integration. To

10
be sure, they certainly are an indicator for the xenophobic tendencies of a certain
group of people. But are these persons representative of the entire society? Or aren't
they rather a small extremist fragment with opinions contrary to the mainstream?

Although in theory identifying indicators of integration sounds simple, this soon

proves to be a very difficult task. This is partly due to a lack of relevant data.
Statistics on migrants are only very rarely available in the form, quality and
exactness desirable. They are not up to date or simply do not exist, because it would
be too difficult to gather the information. One can, for example only find out about
the migrants' housing situation by carrying out an expensive survey among them,
because the housing market is predominantly in private hands and information on the
nationality of the persons renting e.g. two-bedroom flats simply is not being gathered
anywhere - and thus is not accessible for researchers.

But even if the necessary data are available, one of the major problems when
evaluating migrants' integration will remain: the decision about the reference group or
reference data. This question touches one of the core problems in the field of
indicators. Indicators alone do not mean much. In order to become meaningful they
have to be compared over time and - more importantly - to other sets of data.
Although this is common practice, the question has to be asked whether it is really
useful to compare the migrants' characteristics to those of the indigenous population.
Is the average of the non-immigrant population really a good point of reference? Or
does this sort of comparison neglect central characteristics of the migrants and other
important factors determining the indicator in question? Just to give an example: can
the high unemployment rate among migrants in many European countries really be
considered an indicator for a lack of integration? Isn't it rather an indicator for a lack
of qualifications? In other words: is their unemployment due to few integration efforts
or to the poor skills they have?

Especially when migrants are significantly different from the indigenous population
with regard to their education or their professional qualifications it seems rather
pointless to use the average of the non-immigrant population as reference group in
order to assess economic success and integration.

Finding a telling unit of measurement becomes even more difficult when one wants
to assess the degree of cultural integration. Apart from language skills, which are
relatively easy to evaluate, it seems almost impossible to find indicators for cultural

11
integration everybody can agree with. The problem starts with the term itself: What
exactly is "culture"? Religion? Music? Cooking? It continues with the difficulties
which were already mentioned above: What exactly is meant by cultural integration?
Giving up one's folk songs or being tolerated the majority culture?

The question of political integration is only slightly less complicated. One frequently
used indicator in this context are naturalizations. Sure enough, they do indeed tell
something about an individual migrant's willingness to become an equal part of his
new home country. And even the statistical basis is more or less a reliable one,
because naturalization data are readily available. But in reality statements
concerning the development of naturalization rates over time or comparisons of
naturalization rates in various countries are of limited use when serving as indicators
of integration. Can a recent rise in naturalizations really be explained with a growing
inclination to integrate? Or is it the result of the host country's decision to tolerate
dual citizenship when naturalizing foreigners (as it was the case in the Netherlands
after 1992)? In other words: changing naturalization requirements can have an
influence on naturalization rates. But does it influence the individual's degree of
integration? Is it not possible that someone regards further step towards integration
superfluous once he has got the new passport in his hands?

Other often used indicators of political integration also leave a good deal of doubt
behind. Can the participation of migrants in clubs, parties and associations with a
majority of non-immigrants be considered a valid indicator of integration?
Membership alone will hardly be sufficient, because it can not altogether be excluded
that the purpose of participation is to promote the interests of the migrants rather
than pursuing the common goals of the association. Thus, reasons for membership
and activities within the association would have to be taken into consideration as well
- leading, of course, to almost unsurmountable problems as regards data available or
to statements with a very limited meaning.

When it comes to assessing the social integration of migrants the availability of data
is one of the major problems. Almost all relevant data are the result of small-scale
surveys, if they exist at all. Binational friendships, housing or outer appearance
(clothing, hair styles) are as difficult to evaluate as leisure activities or social status
within a certain group of persons. Therefore, statements concerning the migrants'
social integration are often limited to speculations.

12
5. Measuring integration

Bearing these problems in mind it becomes rather doubtful whether the identification
of indicators and the measurement of integration based on these indicators can be a
fruitful undertaking at all. It is questionable whether one can succeed in reliably
identifying a set of indicators really pointing at a progress in integration (or the need
for further measures) and covering all dimensions of integration at the same time in
order to supply a complete impression of the state of integration in a given country.

It may be assumed that local projects aimed at improving the integration of migrants
living in a given community will very well lead to some statements actually allowing
for some indicator-based measuring of integration. This is due to the limited scope
and numerically relatively small target groups dealt with in these projects. Local
projects can, for example, aim at improving the housing conditions of foreigners. By
comparing the situation before starting the initiative to the situation afterwards, the
researcher may conclude that the housing situation has improved and that
integration has advanced, because migrants no longer live in ghettos but in mixed
immigrant/non-immigrant districts.

But then again, these statements can only refer to the housing situation. They do not
give an overall impression of the cultural, economic, political or social integration of
the migrants concerned.

Starting from this - admittedly pessimistic - point of view it soon appears extremely
difficult to see prospects for a nation-wide measurement of the integration of all
migrants living in one country. Using indicators for evaluating the situation of
migrants in various countries is even more difficult: Will it be possible to find a
consensus on meaningful key indicators building the basis for common guidelines on
how to shape integration policies in the member states of the Council of Europe?
Can the definition of "alarm indicators" - e.g. a sharp decline in public opinion - be a
way out?

Answering (or even solving) these questions was the crucial task of the conference in
October. In order to achieve this goal it began with a set of introductions in the field
of concepts of integration and some remark on the problems of measuring integration
as well as difficulties concerning the availability of data. After dealing with this

13
theoretical background the focus of the meeting turned to practical projects aimed at
evaluating the integration of migrants in various European countries. A concluding
debate then assessed the chances for finding common ground on measuring
migrants' integration

Dr. Mandred Werth

Mr Willy Stevens
Ms Silke Delfs

14
 Indicators of integration
Mary Coussey and Elisabeth Sem Christensen

1. Assumptions

The basic assumption in a liberal democracy is that everyone who is legally settled
should have an equal chance to fully participate in the economic, social and political
life of the country, regardless of their race, colour, ethnic or national origins.

Equality has been defined as the opportunity to have the same living standards as
everyone, based on the same freedom of choice including the retention and
development of cultural and religious identity. In the context of equal opportunity, full
participation may not lead to the same results. Integration was defined by the
Council of Europe as: a common framework of legal rights; active participation in
society, on the basis of minimum standards of income, education and
accommodation; freedom of choice of religious and political beliefs, cultural and
sexual affiliation, within the framework of basic democratic rights and liberties.1 A
distinction has to be made between integration as defined above and assimilation,
which requires ethnic minorities to adopt the pre- dominant national culture.

Equal opportunity policies are the means by which national and local governments,
as well as individual organisations ensure that their practices and behaviour do not
act as barriers to the participation of minorities. Where there are such barriers, it is
necessary to establish the reasons for changing practices. As in every area of social
policy, it is necessary to have ways of evaluating the effectiveness of equal
opportunity policies.

Evaluation can be done by using a range of qualitative and quantitative measures

which show over time the extent to which immigrants participate in economic life,
and the extent of inter-action with the receiving society. The measures would vary
according to the activity; for example whether the aim is proportionality of access, or
relations between groups, or the provision of special facilities and compensatory
schemes.

1 "The Integration of Immigrants" by Rainer Bauböck, CDMG (94) 25

15
With regard to access to employment, the assumption behind the concept of full or
fair participation is that if there was no discrimination against immigrants, and they
had not additional disadvantages such as lack of fluency in the receiving country's
language and inadequate education or training, they would be evenly represented in
work and training, and receive qualifications in proportion to their numbers in the
relevant part of the population. Similar assumptions are made with regard to access
to education, housing, benefits and services.

However, it is impossible to find any society which has successfully achieved full
participation of immigrants in all areas of economic and social life. The causes of
inequality, which vary according to the background of the immigrant group, have
tended to be deep-rooted and persistent, and require the sustained application of
social policies in order to tackle the range of accumulated disadvantages from which
many immigrants suffer and change the practices of the receiving society which are
barriers to full participation.

Many of Western Europe's immigrants came from economically less developed areas
in search of unskilled work. Many had a different language, culture, and often a
different religion from that of the receiving country. Many also came from more rural
peasant societies, were less well educated or illiterate, and had no skills or
professional training, or, at least, none which were recognised in their new country.

Although the children of immigrants had fewer disadvantages than their parents, as
they were born and educated in the receiving country, they nevertheless still suffer
from racial discrimination, the effects of living in disadvantaged circumstances and
the fact of attending poorer schools. Some immigrant groups with strong cultural or
religious customs and ties with their country of origin also had low aspirations for
their children, especially for girls, which in turn affects their motivation to gain
qualifications or to get a qualified job.

There is evidence that in most of Western Europe racial discrimination and

discriminatory practices by employers and private house-owners are more extensive
and persistent against those who are visibly different from the receiving population.

There are also significant differences between immigrant groups which need to be
allowed for devising measures. For example, those with high level qualifications,

16
professional or entrepreneurial skills, are better able to overcome any disadvantages
of newness, or of cultural or linguistic differences. For example, in the United
Kingdom, some sections of the population of Indian origin are entering higher
education and universities in proportion to their number in the relevant age2 group.
Research has shown, however, that they continue to experience the same degree of
racial discrimination as less well educated immigrants.

2. Principles in developing indicators for integration

As indicated, a range of indicators is needed for the analysis of a wide variety of

different dimensions. The data should be capable of being aggregated and
disaggregated, to provide a hierarchy of measures, and to enable remedial policies
to be applied at the level of the institution. Central and local governments need
indicators at different levels, both to measure progress and to target resources
effectively.

The data should identify the gaps between the distribution among particular groups
and the majority, with the change in this distribution overtime. It has to be cross-
referenced with other characteristics, such as age, sex, educational attainment,
income etc. in view to compare like with like.

The aim of equal opportunity and integration policies would be to achieve equal
outcomes in a relevant set of positions, through combined actions to remove
disadvantages and to change practices and behaviour. Equal opportunity to
participate is the first step, but will not secure any result, if one of the competing
parties lacks a required attribute. Similarly, there is the question of individual choice,
which even under ideal conditions, may lead to different patterns for some groups.

It may be that indicators are easier to analyse at the level of the specific organisation
as the number of variables is likely to be reduced.

There is also a need for qualitative data to assess changes in the perception and
experiences of both the majority and minority populations. Such data tend to be more

2 The British Race Relation Act 1976 uses disproportionate exclusion to

define indirect discrimination; and fair participation is used in the Fair Employment
Act in Northern Ireland.

17
useful at the level of an institution or sector. For example, it is useful to measure the
extent to which government codes are being applied, or how and whether behaviour
has changed after awareness training in polices forces. Surveys and qualitative
studies, analysis or newspaper cuttings can provide the necessary data.

3. Areas to be measured

Although indicators will vary from country to country according to the make-up of the
population and the legal and policy framework, it is possible to identify some key
economic, social and political areas. In the following examples, the measure is a
comparison of the proportional distribution of immigrants and the majority, as a trend
over time. Supplementary data may be needed to assess whether the gap in
distribution is due to discrimination, structural segregation or voluntary choice.

Indicators which measure the degree of change in the majority society will mainly be
related to attitudes towards acceptance of the particular immigrant group.

Some of the main indicators are:

Access to the labour market

- Employment/unemployment rates and duration, occupation and level,

proportions in dangerous/dirty jobs, all by age sex;

- proportions in key professions such as architects, lawyers, teachers,

engineer, doctors, and in managerial and governmental posts;

- proportions in vocational training, professional training by type of training,

sex and age;

- proportions gaining vocational and professional qualifications, and entering

employment or further training;

- relative earnings, hours worked, self-employment.

The above would give a measure of the distribution of immigrants in sectors and

18
occupations, and allow comparison over time of changes. If policies against
discrimination are effective, immigrants would be expected to become more
dispersed across sectors, and, more important, to gain access to higher levels and
professions. Thus unemployment rates of the higher qualified are a particularly useful
indicator, but need to be related to rates of attainment of specified qualifications to
give the true picture (for example, in looking at whether there is equal opportunity in
the teaching profession, data on the relative proportions of immigrants and the
majority entering teacher training institutions are necessary).

Housing and social services:

- concentration and segregation in districts/quality of housing/overcrowding;

- proportions in public, rented and in self-owned housing.

Data on housing needs to be related to other social indicators, specific to the country
concerned, as public housing allocation policies vary.

- proportions claiming social security benefits; child benefits/maternity

benefits/state pensions.

These data have to be related to the eligible proportions of each group.

Education

- Distribution in types of schools, relative to areas of residence; participation

in pre-school education/results school-leaving certificates and higher
eduction. Take-up of adult language training.

In order to allow comparison at this level, and to help check against discriminatory
practices, the above indicators should be collected by individual institutions.

Participation in political processes and in decision-making

- Voting registration and participation in local and national elections, ethnic

polarisation or integration of votes;

19
- Proportions nominated as candidates and comparative success rates;
proportions of deputies and immigrant constituencies;

- Participation in key institutions and organisations; proportions on boards or

ruling bodies (e.g. trades unions, school boards, works councils).

Mortality, fertility, and demographic changes

- Data on the proportions of immigrants affected by major illnesses and on

causes of death gives a basis for assessing whether mainstream
healthcare needs to be adapted to particular unmet needs.

- Data on birth rates and inter-ethnic marriage give an indication on

increasing social interaction.

20
Judicial indicators

- Comparative data on arrest, conviction and acquittal rates can show social
exclusion patterns for immigrants. However, such data have to be carefully
weighted to allow for other variables.

- The data may also indicate discriminatory patterns in police or judiciary

behaviour.

- Data on racially violent crimes, racial harassment and similar racial

incidents is essential to assess overt hostility and the effectiveness of
prevention.

- Data on complaints of discrimination and convictions will give a measure of

its scale and the effectiveness of legal instruments.

Methodology

Among a wide variety of methods to collect relevant data some of those in use are
the following:

1. Population Census

In some countries the census includes occupation, employment status, qualifications,

housing, languages spoken, ethnic origin or nationality (e.g. the 1991 census in the
United Kingdom asked for ethnicity). It constitutes a valuable source for assessing
demographic and socio-economic changes, and permits a very detailed analysis.

2. Population Surveys

Periodic surveys by households of samples of the population are useful to provide

more detailed data than a census on employment, occupation, qualifications,
earnings, job-seeking patterns and ethnic origin. Such surveys are needed to bring
the census up to date, but allow less desegregation, samples being generally small.

3. Cohort Studies

21
Cohort Studies provide longitudinal data and are used to compare immigrants' and
the majority's experiences in similar circumstances: for example, in searching for
work after school or progress in an occupation. These studies also allow comparison
of progress in institutions or a sector. Cohort studies do not allow desegregation of
data unless very large samples are used.

4. Monitoring and Evaluation Surveys

These surveys are undertaken to establish whether there are disparities in access or
in success rates for any particular groups in a specific organisation. They are useful
for measuring flows into different structures. If collected according to nationally
agreed criteria, it is possible to use external data as a benchmark and to compare
different organisations or units. An example of a national categorisation can be seen
in the USA, where all employers complete a standard form, the EE01.

5. Qualitative Surveys

Studies of experiences give indications on perceptions and attitudes of immigrants

and the majority and on changes and social distance. Analysis of the language and
content of press reporting also provide information on changes in majority attitudes.

Classification of Immigrants

There is a considerable variety among member states in the origins of immigrants

and their circumstances. For example: research into discrimination in the United
Kingdom has shown that colour and visible difference are the factors most likely to
lead to discriminatory treatment. Therefore measures used categorised by ethnic
group and colour. Nationality may cease to be useful in countries with high
naturalisation rates. In general, it seems important to define the target groups
broadly and to include the victims of social and racial discrimination, regardless of
their current citizenship or place of birth.

An index to measure legal integration

Christop Hofinger

22
1. Why a legal index?

The project carried out at the Institute of Advanced Studies, Vienna, attempts to
measure integration with an index, the "IAS"-Index of Legal obstacles of Integration
(LOI)3. But first two preliminary questions have to be answered.

Why do we want to investigate legal aspects of integration?

The destinies of immigrants are widely determined by national legal systems. Thus,
legal integration is a necessary condition for social integration. In other words, the
systematic prolongation of legal differences between citizens of a state and
immigrants reinforces social discrimination against the latter.

Usually, empirical analyses of the integration of immigrants are based on

demographic data and try to investigate the degree of social integration of
immigrants. However, statistical data are generally available only to a limited degree.
An even more serious problem is that, in many cases, statistical data are not suitable
to support comparative analyses because almost every country has its own survey
methods.

Legal aspects of integration, on the other hand, are not always considered to a
sufficient degree. Studies that take the share of settled immigrants as an indicator for
integration easily neglect the legal impacts of a settlement permit: Whereas in one
country, an unlimited residence permit may put you in a position close to a citizen of
the country, you may achieve much less security by an unlimited title in another
country that perhaps still might revoke the permit in case of unemployment.

Why did we investigate legal integration by an index?

To gain standardised numbers as a result of social science research may include, not
only in our case, hard work. However, standardised numbers have some very
convenient properties. The first and biggest advantage is that numbers help to
compare: With the "IAS"-Index, we can compare integration horizontally, finding out
the differences between several countries at time x. We can also compare

3
The researchers who developed the index are, besides the author, Rainer
Bauböck, Dilek Cinar, Ulrike Davy, Henriette Riegler and Harald Waldrauch.

23
integration vertically, investigating the differences between different points in time for
a specific country Y. Furthermore, numbers as a condensed result of research have
another important effect: They make it easier to communicate research results. It
would be very difficult to clarify the various differences in legal integration of eight
European nations if you can't summarise them in a brief history. An index is able to
capture the most important aspects of legal systems and to compound them to a
brief history of integration.

2. How did we proceed?

We confined ourselves to "legal integration". This means that we did not compare
other important aspects of integration such as social or cultural integration.
Furthermore, we analysed laws regulating integration of immigrants that already live
in a foreign country and not laws regulating the process of immigration.

Another limitation was the focus on "ordinary" working migrants and their families.
We did not look at asylum seekers; we did not look at students; we did not analyse
the situation of privileged migrants, such as EU-citizens within EU-countries. Our
focus was on people at the bottom end of the scale. However, the share of non-
privileged immigrants in European countries is generally high: In some countries this
share is relatively low because many immigrants belong to "favoured" nationalities,
like EU-citizens within EU-countries. In other countries, the huge majority of
immigrants do not have any advantage due to their origin. This is true in Austria, for
instance, where more than 80% of the immigrants do not belong to any privileged
group.

Eight European countries were included in our study:

A Austria
B Belgium
CH Switzerland
D Germany
F France
NL The Netherlands

24
S Sweden
UK United Kingdom

We consider this sample as representative for Western, Northern and Central

European countries. These countries are immigration societies (although not all of
them consider themselves as such). We did not include Southern European countries
in our study for two reasons: Firstly, they have different migration traditions being
(except for the recent past) characterised more by emigration than by immigration. A
second, rather pragmatic reason is the fact that, in the case of Southern Europe, it
was much more difficult to obtain English or German translations of the laws we
were interested in, as well as to obtain sufficient demographic data.

When a method is to be applied for the first time, the potential user has to examine
previous work following this method: We didn't find any other attempt to compare
legal systems or parts of legal systems by indices. There are only two exceptions
worth mentioning: the Freedom House-index that measures the level of democracy in
different nations, and the UNDP-index that now also includes assessment of
democracy standards and human rights. These indices have some parallel features
to our index but there are huge differences, like the fact that they do not only take
into account laws. Besides that, the structure of indices like the Freedom House
index is, compared to the, "IAS"-Index, very simple. This simplicity may have its
advantages in the case of a democracy index. A legal integration index, whatever
structure it might have in the end, can't be a simple one.

All in all, we included almost 80 different legal issues concerning the legal regulation
of integration in the countries investigated. We aggregated these issues in either
two, three or four steps into five main areas. These five main areas are:

1. residence,
2. labour market,
3. family reunion,
4. naturalisation,
5. second generation.

What we haven't done so far is to aggregate these five dimensions to an ultimate

25
level, i.e. calculating a single numerical indicator for each countries system of legal
integration. For this, we would have to take into account all other relevant legal areas
too.

How does a country get its score on the "IAS"-Index? Its value on the index is higher:

- the longer waiting periods for achieving higher levels of integration are, eg
for permanent residence or naturalisation;
- the more conditions there are which have to be fulfilled by the applicant to
achieve a better status;
- the more external constraints (like quotas) there are that are imposed on
improvements in status;
- the easier it is to lose a title;
- the fewer the rights which are accessible for immigrants and the more the
immigrants are dependent on discretionary decisions by the administration;
- the less choice there is for immigrants.

We standardised our results within a range from zero to one. A value of zero would
mean that a country imposes almost no legal obstacles upon the integration career
of immigrants. A value of one means that the integration is characterised by so many
legal restrictions that it is hard to conceive a more rigid legislation.

We tried to include all relevant and comparable legal rules regulating integration
within the five dimensions mentioned above. This distinguishes the IFS-index from
other indices that are highly selective in their choice of items.

Probably the most delicate part of an index are the weights related to the particular
items and aggregate constructs. Weights are inevitable when building an index. We
had to determine which legal regulations are considered more important and which
less important. In order to aggregate scores of non-weighted items, each has to be
given the same weight - a procedure that requires justification as well.

It is obvious that every group of experts charged with the task of deciding on weights
would come to different results. To test the impact of our particular decisions on

26
weights we used alternated and even random weights. We found out that our index
was surprisingly robust.
Our index is inevitably normative in its definition of integration standards. However,
we still hope to have developed a tool with high descriptive powers.

3. What are the results?

In our sample, Sweden and the Netherlands can be considered as the two nations
with legal systems that help immigrants most in their integration careers. Sweden is
first in three of five domains, and the Netherlands is consistently among the four
most integrative nations. Belgium and France show similar openness towards
immigrants in terms of integration.

27
3.1. Residence

Our first and maybe most important dimension is residence. The main questions in
this area were, among others: How many conditions do immigrants have to fulfil
before a limited residence permit is prolonged? How long do they have to wait before
they can acquire an unlimited permit? Do they have a right to settle after having lived
in a country for a certain time period? And how easy is it to lose the right to stay?
European countries vary considerably in their responses to these questions.

3.2. Access to the labour market

One of the narrowest bottlenecks for integration is the labour market. Even migrants
that crossed the border years ago may be confronted with restrictions in this area.
The variance among European countries is not very high. Some legal systems follow
the principle of consolidation and security of residence.

3.3. Family reunion

Family reunion is generally recognised as a human right. Therefore, each European

country makes it possible, at least to a certain extent, for migrant workers to let their
spouses, children, or maybe even parents come into the country where they live and
work longer periods. However, the countries differ considerably in the conditions that
have to be fulfilled to do so. For example, Sweden does not impose substantial
obstacles on family reunion, so that it achieves a score of 0.01. This is extremely
close to the minimum that our index is able to measure. In most other European
countries, family reunion seems to be guaranteed to a satisfying extent.

3.4. Naturalisation

There is no country in the world where citizenship is offered as a gift to newcomers.

It is obvious that some conditions have to be fulfilled before a foreigner is able to
acquire the nationality of the receiving society. However, it seems that European
countries are a bit more reluctant to grant citizenship to foreigners than other
societies, like Canada or Australia. These countries have a waiting period of three

28
years which would lead to a really low score on the index. In the nations analysed
here, the waiting period comes to at least five years. In few countries, this waiting
period is even between ten and twelve years for discretionary decisions.

3.5. Second Generation

The biggest differences among the analysed countries can be observed with regard
to the legal position of the so-called second generation. Again, we looked at the
regulations concerning labour market, residence, family reunion, and naturalisation,
but this time for non-citizens that are born in the receiving society. Four countries out
of eight have rather low ratings. Obviously, foreigners born in these countries find
themselves in a situation that comes close to the position of naturalised citizens.

4. How can we improve the index?

What is missing is the summing-up of all the presented dimensions in a single

number. Of course, in the long run, this has to be the aim of our endeavours.
However, we have to postpone the final aggregation step until we are able to include
other crucial legal areas. One of these areas is the granting of political rights like the
right to vote. Other important points would be the inclusion of social rights and,
finally, legal measures to protect immigrants from discrimination as well as positive
cultural minority rights.

A possibility to justify the weights empirically would be the linking of the weights with
demographic data or surveys. By doing this we could include the opinion of
immigrants themselves about the meaning of particular legal regulations. However,
this plan seems very difficult to realise when looking at the general availability of
empirical data in this realm.

Finally, it could be interesting to calculate the index scores for other immigrant
groups. We plan, for instance, to compare for the EU-guidelines concerning
integration of EU-citizens with the regulations for non-privileged immigrants already
analysed.

5. What can we use the index for?

29
Even after the inclusion of other legal domains into the "IAS"-Index we have to keep
in mind that we are dealing with a legal index. This means we only compare laws.
We don't compare the public attitude towards immigrants; we don't compare the
educational systems in terms of their ability to include foreign pupils or students; we
don't compare the economic positions of immigrants either. All this matters for
immigrants but, if we want to measure it, we have to develop a new type of index or
to use completely different methods.

We think that we are able to use our index as an evaluation instrument for national
integration policies. If a government changes the legal base of integration, we can
measure that change.

At the moment, we see much political will to improve the situation of immigrants. Our
index has two functions in this process: first, it induced a broad range of reactions in
the general public and increased the pressure for a change; and second, as soon as
amendments are discussed, it helps to assess their overall impact on the legal
integration of immigrants.

30
 Concepts of migrants' integration:
a comparison of national policies
Philip Muus

1. Introduction

It is not easy to conceive of a Europe united in immigration and integration policies. I

define immigration policy here as the policy related to the admission of immigrants,
and integration or immigrant policy as the policy related to the incorporation of
immigrants into the receiving society.

Why is it so difficult to reach a common immigration and integration policy in

Europe? And why should we start with questions about immigration policy, when
talking about concepts of integration? Is integration not difficult enough a subject to
be treated in its own right? The answer on the last question is without any doubt
affirmative. But since the concern here is the integration of immigrants, the linkage
between concepts and policies of immigration (admission) and integration deserve
some attention. Most of the immigrants enter the European nation-states under
national Alien Laws, while only EU -and Nordic citizens are free to move in the
respective free movement areas. In the process of Europe's integration, the member
states of the EU encounter serious problems when it comes to define common
immigration (admission) policies with regard to third country nationals. It is only state
sovereignty that is at stake. European nation-states have gone through different
ways of nation building and some of them have been important colonial powers. As a
consequence, their views and ideologies might differ about the nation, about who is
a citizen and who is to be considered an alien, and moreover who might naturalize
under which conditions.

Before we start to think about concepts and policies of integration of immigrants in

the European context, we should remind ourselves of the following: The composition
of immigrant populations is different in each of the European nation-states, according
to country of origin, migration history and legal status. These immigrants have been
admitted under nation-specific alien laws. But not only the 'mix of immigrants' is
different in each European receiving country. The socio-economic, cultural and

31
political aspects of society in these receiving countries are not similar either.

In thinking about concepts of integration, and in proposing indicators measuring

integration, we have to keep in mind that we have to do with these existing
differences before we start to compare. Integration is broadly to be defined as the
incorporation of immigrants into the receiving society. We have to find ways to
compare what seems incomparable by stressing those elements in different
integration concepts that go beyond the purely normative aspects.

2. Concepts and policies of integration

There is no uniform concept of migrants' integration in the European receiving

societies. The word integration itself (Böhning in Werner, 1993) is not unambiguous;
it may point to the process or to a state. Concepts of (cultural) integration range
from multiculturalism to full cultural assimilation and even segregation (see also
Castles, 1993). Basically integration concepts only deal with the public domain of
society and its (immigrant) actors. There is, however, no universal definition of the
public and the private domain. Concepts of integration may differ greatly with regard
to their position on the inclusion-exclusion continuum of legal-political rights (eg
possibilities of naturalization, active and passive voting rights), and in offering access
in differing degrees to language training, housing, education, the labour market and
the social security system. Concepts of integration may differ in the role attributed to
the government and/or to the immigrants in the process of integration. It might be the
case that the government only provides the legal framework for access in society
and/or the legal framework in the fight against discrimination. The roles and duties of
immigrants in view of their participation in society may also differ.

Thoughts about integration may be based on the assumption that well developed
concepts on the integration of immigrants exist in all or most European receiving
countries. This is not the case. Concepts may be discussed at a scientific level.
Integration concepts and integration policies, however, are explicitly formulated on
governmental and policy levels in all of the receiving states. Concepts and specific
policies may be completely absent or only in an incipient phase of development like
in those countries in Southern Europe that face immigration only recently and even
more in Central and Eastern Europe. In other countries, like the Netherlands and

32
some Nordic countries, existing concepts of integration are widely discussed and
partly replaced by new ones.

Three elements at interplay in the different concepts of integration can be

distinguished: The first element is the relation between the cultural aspects of the
public and private domain. Concepts of integration differ in scope and desired degree
for cultural adaptation of immigrants (between cultural assimilation and cultural
segregation) but they all deal with the public domain. However, the public domain
may differ in the countries compared.

The second element is the degree of inclusion/exclusion of immigrants in non-

cultural aspects of the public domain: (1) legal-political (residence rights, access to
citizenship, voting rights), (2) socio-economic (language training, (professional)
education, housing, labour market, health services and social security).

The third element in integration concepts is the immigrant's role in the integration
process: what are his or her duties in order to integrate in the public domain?

Concepts of integration (policies) relate to all of the above mentioned elements. In

order to get some more insight into these concepts, these three elements will be
treated separately.

1. The different cultural approaches may logically lead to the following concepts of
immigrants' incorporation (definitions a-c from Castles, 1993):

a. assimilation, a one side process of adaption, in which migrants

have to give up their distinctive linguistic, cultural or social
characteristics and become indistinguishable from the major
population;

b. integration, a process of mutual accommodation between

immigrants and the majority population. The concept implies that
immigrant groups will cease to be distinctive in culture and
behaviour over time, but sees the adaptation as a two-way
process in which minority and majority groups learn from each

33
 other and take aspects of each other's culture;

c. multi-culturism refers to the development of immigrant

populations into ethnic communities which remain distinguishable
from the majority population with regard to language, culture,
social behaviour and autonomous associations over a long period
(usually several generations);

d. cultural segregation refers to a complete separation between

the immigrants' and host society's cultures.

2. Concepts of inclusion/exclusion of immigrants concerning non-cultural

aspects of the public domain, may be related to the following policies:

a. Policies that allow for and promote the acquisition of citizenship

by naturalisation and /or double nationalities;

b. Policies promoting active and passive voting rights for alien

immigrants;

c. Policies directed at the legal non-discriminatory treatment of

immigrants: Anti-discrimination laws, and laws on equal
treatment, and the abolishment in national laws of elements
discriminatory for immigrants;

d. Policies based on the principles of accessibility and

proportionality: The first element points at the full and equal
access to public services by immigrants and the majority
population to public services. Proportionality has to do with the
proportional distribution of one category over all relevant
positions (like eg (professional) education, housing, labour
market) compared with that of the majority population. Policies
based on positive action, or "positive discrimination" are
sometimes applied with the aim to reach proportionality.

34
3. Concepts of integration may in practice contain different ideas about the
migrants' role and duties in the process of integration:

Policies might encompass positive or negative sanctions for immigrants in

measures (voluntary or obligatory) meant to improve the immigrants'
abilities to participate in the receiving society (eg language tuition,
(professional) education, measures directed at the improvement of the
immigrants' position in the labour market etc.).

The three policy elements treated above (cultural policies: public, domain versus
private domain, inclusion/exclusion policies in non-cultural public domain and policies
with regard to the immigrants' role and duties) may lead to integration policies that
might be also be distinguished, according to their

- directness (direct or indirect immigrant policies, see also

Hammar, 1985);
- level of implementation (national, regional, local);
- target population (all immigrants, a selection of immigrant
groupings, disadvantaged (immigrant) target groups, a specific
immigrant generation etc.)

3. Problems and possibilities in comparing integration (policies)

We have to find ways to compare what seems incompariable by stressing those

elements of integration that go beyond purely normative aspects. We have been
warned that not only the composition of the immigrant population group in each of
the European receiving states is different, but that the receiving states also differ in
socio-economic, cultural and political aspects.

Part of the solution may consist in comparing the same immigrant group on relevant
aspects in different countries. But what can we conclude if the per centage of
unemployed semi-skilled Turkish workers in a specific age-group is higher in one
country than in an other, if we do not take into account general and specific
differences in the labour markets of the countries compared?

35
Variables like naturalisation rates are measured relatively easily but what do they tell
us about integration. If in one country naturalization is a simple process, while in the
other country it might be practically impossible? Is the rate of mixed marriages by
nationality a possibility to measure integration? What to do with those mixed
marriages, which are in fact marriages between two former compatriots of whom one
has changed his or her nationality by naturalization? How do we compare figures on
mixed marriages between countries with easy and difficult procedures of
naturalization?

If properly used, labour market participation rates, language abilities, the rate of
mixed marriages by nationality, naturalization rates, as indicators inform us about the
degree of integration in one country, in relation to the integration concept adopted by
the country concerned. If we want to evaluate the integration level in different
countries our comparison will need refinement and elaboration.

One aspect has not been treated yet. Since our focus is on integration, we might
easily forget to consider transnational migrant communities. Today's immigrants may
have easy access to elements of their own "high" culture:

Parabolic reflector receivers shorten the distance between the immigrant and his
home-country. The costs of travelling are lower than ever. In double nationality is
possible, immigrants do not lose their hereditary titles or ownership of immovable
properties in their home country any more. The relationship between property
ownership and hereditary rights on the one hand and actual integration on the other
hand is scarcely researched. How to interpret the sudden increase of Turkish
immigrants in the Netherlands asking for Dutch nationality since dual nationality
became a real option? As a sign of integration or as a practical act safeguarding the
immigrants rights in the country of settlement and the country of origin, or as both?
The parabolic reflector receivers, are they to be interpreted as an expression of the
freedom of the press, speech, etc. belonging to the private domain of the
immigrants? We cannot answer these questions here. They underline, however, the
difficulties related to the evaluation of the integration process and concepts of
integration and they are a sign for the immigrants loosing interest in the host
society's culture.

36
 * * * * *

Sources:

- Bauböck, Rainer (1994); The Integration of Immigrants. Documents

prepared for the Council of Europe. Strasbourg, CDMG (94) 25 E
- Brochmann, Grete (1995); European Integration and Immigration from
Third Countries. Oslo, Institute for Social Research.
- Castles, Stephen (1993); The Process of Integration of Migrant
Communities. Paper, Expert Group Meeting on Population Distribution and
Migration, Santa Cruz, Bolivia. UN/UNFPA.
- Hammar, Tomas (ed.) (1985); European Immigration Policy; A comparative
study. Cambridge University Press, Cambridge.
- Muus, Philip (1984); Immigratie, van Invloed op Integratiebeleid
(Immigration of Influence on Integration Policy). Pre-advice in charge of the
Ministry of Interior of the Netherlands; DCM, Ministerie van Binnenlandse
Zaken, The Hague.
- Muus, Philip (1995); International Migration within and Towards the
European Economic Area (EEA) - Past Trends and Expectations for Future
International Migration; in Social Security in Europe, Equality between
Nationals and Non-Nationals, European Conference, Oporto; Commission
of the European Communities/Departamento de Relaçoes Internacionais e
Convençoes de Segurança Social;
- Werner, Heinz (1993); Integration ausländischer Arbeitnehmer in den
Arbeitsmarkt- Deutschland, Frankreich, Niederlande, Schweden. ILO
WEP/Migration and Population, Working Paper, Geneva.

37
 A european research project
on migrants integration
Raimondo Cagiano de Azevedo
and Barbara Sannino

1. Introduction

During the 80's, immigration and the questions related to it have been among the
principal issues of political discussions; they still are today.

The political actors were divided on certain subjects like the regulation of migration
flows, social and economic costs of immigration, the identification of integration
models and the relationship between host society and immigrants.

In general terms, the phenomenon of immigration implies questions of coexistence,

which must be answered by endeavours to integrate ethnically and culturally different
groups. Immigration is perceived, from differing point of views, either as integration
or as assimilation. Integration means equal access to the labour market, housing, the
education system as well as to health and other public services. From a socio-
economic perspective, immigrants can be in very different situations ranging from
integration to total segregation. In a segregated position, immigrants have low skill-
jobs, live in low standard housing conditions and lack access to public services.
Integrated immigrants do enjoy the same opportunities as the host population.

The "assimilation" process concerns life styles, manners of expression, cultural and
religious value systems, human relations etc. The immigrants' position in the
assimilation process can also differ considerably. Assimilation is a long-term process
compared to integration. In the present text "assimilation" refers to the immigrants'
adaptation to the life style and value systems of the host society.

Both processes are interrelated. The links between them may sometimes also create
difficulties, in particular for the second generation of immigrants having less cultural

38
distance from the host society than their parents. They have undergone an
assimilation process, without having reached, from a socio-economic point of view, a
satisfactory degree of integration. This situation may lead to tensions. Seemingly a
higher degree of integration than of assimilation does not cause the same amount of
stress as the opposite. It results nevertheless in the immigrant groups' isolation from
the host society. If then they try to impose their cultural models, even greater
conflicts may arise.

2. Migrants' integration in europe

This very broad introduction suggests that the most important statistical data in
relation to the immigrants' integration process to be collected concern the admission
on the territory of the country of residence, residence periods, working and living
conditions and naturalisation. Integration must be analysed in its social and legal
aspects. Immigration also affects the demographic structure in the country of
residence. It also has some influence on the labour market situation, which is
currently a rather difficult one.

Recent population movements in Europe, with their consequences for social

protection systems and the labour market, have to be seen in a new dimension.
Within the coming years, political solutions have to be found urgently. The
phenomenon is still new in Southern European countries, where the administrative
and social structures are used rather to emigration flows of native workers than to
immigration of foreign citizens. The phenomenon is not new in Northern European
countries. In fact, administrations in all European countries face questions of
immigration and integration of immigrants. In their official statistics, however, only
few traces of this debate can be found. On a European level and in a scientific and
statistical approach, relevant data are not compiled or analysed systematically.

For these reasons, in a Eurostat Research Project on "Measuring Migrants'

Integration", alternative measures of integration were examined and current
statistical measurements tested on a European level on the basis of existing records
and statistical surveys. Several steps were performed and a Research Report in
three volumes, published by the University of Rome (1993-1995) concluded this
programme.

39
The sources for these infirmation on integration and integration measures were
periodical official data and official estimates and, to a wide extent, those informal
surveys and estimates which it was possible to carry out.

The data on employment, housing and education show the great diversity of
immigrants' positions in the integration process. Participation in social life and access
to health care are other relevant factors, even though it is difficult to determine a
priority. Differences in the housing situations may be particularly significant for the
immigrants' integration performance. These differences are and will be of special
informative value on the integration process.

Demographic data on fertility and intermarriages are also useful measures of

integration and assimilation. In most European countries fertility has declined. The
fertility rate of immigrant populations of European origin, but also of West-Indian and
non-Muslim Asian origin, has declined to a period level or even below that of the host
society. Muslim population groups from Turkey, North Africa and South Asia have
shown the least decline.

Intermarriage is increasing faster than expected in less well integrated immigrant

populations. Up to 40% of West Indians born in the United Kingdom appear to have
white partners. The same proportion is reached by young Maghrebians in France. If
fertility decline and intermarriages are - as a sign for a "vouloir vivre ensemble" -
good indicators for social coherence on a national level, they reveal a rapid
absorption of immigrants of European origin and sharp differences between
immigrant groups from Third World countries, between immigrants of Caribbean,
non-Muslim Asian or Muslim origin, between first or second and more generation of
immigrants.

The present analysis leads to certain conclusions. First of all we cannot talk about
only one type of integration as there is a plurality of different integration features.
Some of them are negative (eg criminality), others positive (labour demand, trade,
consumption, etc.). Integration in society and reality is much more complex than
perceived by legislative and policy institutions.

Furthermore, integration is to be considered as evolutionary process and compared

40
to the evolution of society as a whole. Integration must be seen as interactive system
of relationships and not only as static combination of individual situations. Integration
is also related to questions of choice and needs. Integration as a choice means the
freedom to live in a given place, to look for a job and to enjoy civil rights. Integration
as a need implies a more or less explicit persuasion of people concerned.

The concept of integration must thus be analysed from different perspectives: the
individual's, the family's, the social groups' and the ethnical group's concerned one.
We may also distinguish official and informal, legal and non-legal integration.
Therefore statistical sources must rely on official and informal support.

Speaking about immigration and integration, we always think about migrant workers'
integration. Today, however, we should speak, more generally, about the integration
of citizens. In our countries, immigrants should not live in worse or different
conditions than those prevailing in their countries of origin as far as basic human
rights are concerned.

The integration level in European countries can be identified by using new statistical
ratios, in particular different dependency ratios.

Information required to establish these ratios concerns less the traditional

demographic structure of our population than movements and new structures of
minorities, ethnical groups and immigrant populations. These statistical ratios would
allow different ways of measuring quality of life and social benefits, both for the
majority and the immigrant population.

Through this analysis we would get clearer pictures of the integration efforts
undertaken by European countries following different methods.

According to the German model, immigrant workers were considered as temporary

guests. This model excluded full socio-economic integration, while granting good
wages and regular employment. In Germany integration is considered as a process
of adaptation to German conditions requiring some active participation by the
immigrants themselves and their acceptance of the fundamental values and basic
principles enshrined in the German constitution (separation of state and church,

41
equality between women and men, freedom of religion, etc.). Furthermore some
knowledge of the German language, abstention from excessive nationalistic and
religious behaviour and insertion in the schooling system (compliance with the
obligation to attend school) and in the labour market are considered as prerequisites
to be fulfilled. On the other side, foreigners living in Germany may legitimately expect
the host society to be tolerant (Federal Ministry of the Interior; 1991, p.5- 6).

The French model aims at turning immigrants into French citizens. Immigrants are
granted the same rights as French citizens and equal opportunities. Therefore, in the
French model the ethnical or cultural differences seem to play a minor role.
Immigrants have to be firmly included in societies with strong national identities.
They must have the same opportunities as French citizens in the social and
economic field on condition that they adapt their behaviour to the basic values of the
host society. Immigrants are supposed to show their different religious or cultural
adherence only in the private sphere.

Integration in the United Kingdom does not follow a model of assimilation, but of
equal opportunities for everyone, irrespective of colour and ethnic origin and of
cultural diversity in an atmosphere of mutual tolerance. This model is occasionally
referred to as "multicultural" or "pluralistic integration".

The Belgian approach to integration is the concept of "insertion" showing four

structural levels:

1) The immigrants' submission to the public order, i.e. national legislation

(exceptions being accepted in the framework of international private law).

2) Promotion of the immigrants' social insertion taking into account the

"fundamental social principles of Belgian society". These principles refer to the ideas
of modernity, emancipation and pluralism, in a manner Western societies conceive
them.

3) In the context of item 1 and 2 above, unambiguous respect is required by

native Belgians and immigrants for cultural diversity as a possibility of mutual

42
enrichment, eg all the aspects related to cultural identity and practices.

4) Promotion of the structural involvement of minorities in activities and aims

of public authorities (J. Leman, M. Van Haegenborgh, 1994).

The "insertion"- concept tries to synthesize the Anglosaxon and Latin integration
strategies. The Flemish Community (Flanders and Flemish-Brussels) tends to adopt
the ideological background to the Anglosaxian model, in which "integration" should
allow immigrants' groups to preserve their own culture and also to differ from each
other, this being considered as enriching for the society as a whole. The francophone
Community (Wallonia and French-Brussels) tends to adopt the ideological
background of the Latin model, in which acculturation of the individuals and
acquisition of citizenship constitute the most important features. In such a concept
people are considered as equal individuals. The main strategy consists in preventing
marginalisation and exclusion.

In Norway, Norwegians and foreigners have the same rights granted by law with
some exceptions: Only nationals can vote in general elections, become members of
Parliament or do military service. Since 1983, immigrants have the right to vote in
local elections after a residence period of three years.

The Dutch model is similar to the English one. It respects the different cultural
identity of immigrants. It differs from the English one by including policy measures of
"positive discrimination" in view of achieving socio-economic integration.

Italy adopted extensive legislation on migration, concrete policies of integration,

however, are incomplete. There is no coherent integration system, although the
general orientation of such a system seems to favour the regularisation of
undocumented immigrants and the preservation of the culture of the country of
origin. This attitude is underlined by the granting of equal rights to non-EU nationals
and Italians in terms of access to social security, health services, the education
system and housing. The cultural identity of immigrants is preserved.

Other countries, such as Sweden, have adopted multicultural policies allowing

immigrants to keep up their way of life.

43
44
3. Statistical indicators

Statistical indicators are needed to analyse and study the socio-economic

integration. We distinguish indicators in five main areas: labour market, housing,
education, a more general one comprising demographic indices, and a
comprehensive one casting a light on the different integration levels.

Looking at the original tables in the Eurostat Research Report, it is important to

underline, that:

- indicators are already available; some of them are used in the national
reports;
- in some cases, indicators have been proposed, but are not yet in use;
- the indicators should be analysed with different variables, like citizenship,
ethnic group, sex and age group, and compared to the rate of foreigners in
each of these groups;
- indicators for the immigrants' integration should be compared with
indicators on the host population;
- the quality and validity of data depends on the kind of sources available
(official/informal).

In every country we are able to obtain information on the number of immigrants, the
per centage of the active foreign population in relation to the total foreign population,
the unemployment rate, the number of students by school type. New indicators may
be developed according to needs. In the United Kingdom, for example, 50% of the
population belonging to ethnic minorities are not foreigners any more, but nationals.
This fact leads necessarily to a new perspective and underlines that integration is a
question that concerns the national population as a whole. Belgium also shows a
high ratio of naturalized immigrants. In other countries, too, demography experts try
to measure the integration of naturalized immigrants.

Administrative sources are used by experts to collect other general information on

immigrants: e.g. reasons for staying (Italy: family reunion 17,3%, study 13,7%;
Norway: family reunion 9,4%, seasonal workers 18,9%, renewals 52,9%). Experts
also use demographic sources: foreign population by marital status, country of origin,

45
age group, sex (Germany, United Kingdom, Italy). In Germany this rate shows that
out of the total of 5,64 million foreign residents, 22,8% are younger than 15 years
(males 22%, females 23,7%). This per centage is 27,2% among five selected
nationalities. The percentages of the elderly (55 years and more) represent only
8,6% of the foreigners in Germany. The rate "foreign head of household by marital
status and length of stay" has been proposed by Germany.

In many countries (Belgium, France, United Kingdom, Italy, Norway) information

about the territorial distribution of immigrants is collected. Therefore it is possible to
show a concentration of immigrants in cities and industrial areas. This is important,
as it implies that in the same areas job applications by immigrants will concentrate.
Other demographic indicators are the per centages of marriage, of birth (France,
Italy, the Netherlands), of divorce and of death by causes (France).

Several factors influence the high - in comparison to the host society -

unemployment rate among minorities. Generally, minorities are poorly qualified.
Simultaneously the restructuring of the economy has led to a sharp reduction of
employment in the unskilled and semi-skilled sectors. Furthermore the high
unemployment among minorities is disaggregated by age (many youths) and gender
(more men than women).

The analysis of relevant data concerning France, Germany, United Kingdom, Italy
and the Netherlands show that immigrants, after arrival in the country, carry out
modest jobs. In general, immigrants don't leave their first jobs, because of lacking
alternatives. All these indicators must be analysed by sex and age. Groups with a
more recent migration history comprise more males than females. The foreign
population is much younger than the indigenous one. Employment among the
second generation is lower than among those who were born abroad. The economic
activity rate for women shows the cultural differences. It is lower for immigrants than
for the host population.

Employment rises among males and females with rising education level
(unemployment by educational level and sector of economic activity). Other
indicators of employment are the per centage of independent workers, contractual
typologies, working hours a week (part-time, full-time), income by sectors of

46
economic activities (Belgium, France, Germany, Great Britain, Italy, The
Netherlands). France has underlined the importance of information on irregular work.

Information on the housing situation of immigrants is not easily accessible yet. In the
EU member states, it seems to be difficult for third country nationals to rent flats or
houses, because of the high rents and the, sometimes, discriminatory attitudes of
owners. Housing indicators vary from country to country. In Italy, for example, the
criteria of ownership is not used. In Belgium, 62% of immigrant owners have one-
family houses in comparison with 79% of Belgian owners, 38% of immigrant owners
have apartments compared to 21% of Belgian owners. Upon arrival and in the first
period of residence, housing ownership is not widespread. In Italy, a new indicator
has been introduced, the number and per centage of immigrants staying in refugee
centers.

The "housing typology" and "housing typology by year of arrival" allows conclusions
on possible improvements in the housing situation of immigrant. Another important
indicators are "the number of persons per surface" and the "tenure and services per
accommodation" referring to WC and bath facilities, sanitary installations in general,
central heating, etc.

In the field of education the most common indicator is "the number of students by
school type". The attendance rate should be estimated, which is difficult to do. In
order to evaluate the integration of students, the performance (success or failure)
should be rated. According to the Belgian report, immigrants' children are less
successful in school than children belonging to the majority population; 29% of the
foreign pupils failed their first year of primary education compared to 8,5% of Flemish
pupils. The failure rate becomes even more important in the following years of
education. The reasons for this development are among others: the high
concentration of foreign pupils in certain schools, missing language skills, lacking
participation in pre-school education, overworked and demotivated teachers, etc.
There is a link between the language skills, the rate of school-leaving and the school
results of the immigrants' children.

Countries also use other general indicators adapted to their experience: Italy and the
United Kingdom suggested to study the criminality rate among immigrants as to
identify social unrest. It would also be of great use, though very difficult, to analyse

47
irregular and clandestine migrants. France and Italy have considered to introduce
"the number of family reunions" and "acceptance rate of family reunion applications"
as new indicators.

4. The statistical sources

The sources of principal statistical indicators on integration are:

1. Percentage of the active foreign population in relation to the total foreign

population

In Italy it has been analysed by sex, age, nationality, marital status, educational
level, sector of economic activity, language skills. Sources: residence permits,
registered at Labour Registry Office, INPS.

In Germany it has been analysed by sex, age, nationality, marital status, sector of
economic activity. Sources: Micro census, Federal Labour Office.

In the Netherlands it has been analysed by sex, age, nationality, educational level,
profession. Sources: Census, Labour Office.

In Belgium it has been analysed by sex, age, nationality, educational level, kind of
profession, marital status. Sources: periodical official data, unofficial estimates.

In Great Britain it has been analysed by sex, age, knowledge of language. Sources:
Census, labour force, local and national suveys, statistical analysis, Commission for
Racial Equality investigations, research.

In France it has been analysed by sex, age, nationality, sector of economic activity.
Sources: Ministère de l'Education Nationale, OCDE, unofficial estimates.

2. Unemployment rate

48
In Italy it has been analysed by sex, age, nationality, marital status, educational
level, sector of economic activity, knowledge of language. Sources: residence
permits, registered at L.R.O. (periodical official data, occasional unofficial estimates).

In Germany it has been analysed by sex, age, nationality, marital status, educational
level, sector of economic activity, knowledge of language. Sources: Federal Labour
Office,(occasional official survey, periodical official data, unofficial survey).

In the Netherlands it has been analysed by sex, age, nationality, marital status,
educational level. Sources: Labour Office (periodical official data).

In Belgium it has been analysed by sex, age, nationality, marital status,

educational level. Sources: National Institute for Statistics (periodical official data).

In Great Britain it has been analysed by sex, age, knowledge of language. Sources:
Census, Labour force, local and national surveys, statistical analysis, immigration
Office.

In France it has been analysed by sex, age, nationality, sector of economic activity.
Sources: Ministère de l'Education Nationale, OCDE, unofficial estimates.

3. Overcrowding rate4

In Italy it has been analysed by sex, age, nationality, marital status, educational
level, sector of economic activity, knowledge of language. Sources: Census,
occasional survey.

In Germany it has been analysed by sex, age, nationality, marital status, educational
level, sector of economic activity, knowledge of language. Sources: Census,
unofficial survey.

4 Number of people living together in relation to the number of available

rooms

49
In the Netherlands it has been analysed by nationality. Sources: Census.

In Belgium it has been analysed by sex, age, nationality, marital status,

educational level, sector of economic activity. Sources: Census, unofficial survey.

In Great Britain it has been analysed by knowledge of language. Sources: unofficial

survey.

In France it has been analysed by sex, age, nationality. Sources: Census, periodical
official data.

4. Rate of school leaving of foreign children, compared to autochthons children,

language skills, school results, rate of the increase of foreign students: Analysed by
sex, age, nationality, in every country. Sources: periodical official data (Italy,
Germany, Belgium, France, The Netherlands), occasional official survey (Germany,
Great Britain), unofficial survey (Italy, Germany, Great Britain, Belgium).

5. Conclusions

With the on-going integration process, immigrants' are to be increasingly considered

as integral part of the population of the host country. Statistical indicators should
ideally show the same figures for immigrants as for the majority population.
Therefore statistics on demographic consistence and structure (by sex, age,
education and employment), natural movement (births, deaths, marriages) and social
mobility should be established in a continuous and reliable manner.

Official statistics should include statistical indicators allowing for a comparative

interpretation of national experiences on a European level.

In Southern European countries immigration is a recent phenomenon. Historically

these countries experienced emigration of native workers due to poor living
conditions. In countries which traditionally experienced immigration, various
integration concepts were and still are discussed, leading to a wide range of
measures through which socio-economic needs should be met.

50
How to measure the success or failure of integration policies has been a widely
debated subject based on either political or scientific interest. This debate is not
much reflected in the official statistics of European countries and relevant data are
not collected in a systematic way.

Recently, a certain progress could be noted. In statistical terms, however, the crucial
moment will reached when the same variables and indicators will be applied to
immigrants, ethnic groups and the host population.

In the present European context, two points seem particularly relevant:

a) On a micro level to study the motives and circumstances in the

environment in the country of origin motivating the decision to migrate:
family and village structures and institutions, socio-economic situation in
the municipality or region of origin.

b) On a macro level, it is fundamental that the authorities of sending countries

participate in international cooperation mechanisms and structures.

We are obviously shifting towards an approach in which statistical aspects are more
closely linked to the political aspects of the integration of immigrants.

The first statistical indicators to be developed should focus on stock data rather than
flow data. They should refer to sex, age, education, profession, maritage rate, birth
rate, mortality rate and on geographic and social mobility. They should also deal with
housing, health care and education of immigrants and on their participation in
political life.

The scientific and technical debate on these issues must be continued on a national
level as well as on a regional or municipal level. Further financial and organisational
resources could be devoted to research programmes carried out in a coordinated and
systematic manner in various European countries, as called for in one of the
Recommendations of the 1993 European Population Conference in Geneva.

51
Summary

In the first half of the nineties, an important and complex research programme and
studies on migration movements affecting European countries were implemented in
the framework of the European Union. One part of this research was devoted to the
question of measuring the integration of immigrants in Europe.

This activity has been concluded with the publication of three volumes corresponding
to the three main parts of the investigation. Firstly, an extensive comparative study
on existing statistical indicators relating to the integration of immigrants was
undertaken in the then twelve member States. In the second stage of the project
cooperation structures and mechanisms between receiving countries and countries of
origin were studied on the basis of some proposals and indications made in the first
stage. In the third part of the study case studies on the basis of five variables were
carried out in seven countries: Belgium, France, Germany, Italy, Netherlands,
Norway, United Kingdom. During the implementation of the research project a
network of experts contributed to the discussion through studies and in international
seminars. The findings of the discussions are also reflected in the present document.

6. Bibliographical references

- AA. VV. (1989) "Il mondo arabo, l'Italia e l'Europa", Turin,

Fondazione G. Angeli, in: Abitare il pianeta, vol.1.

- M. Anwar (1991) "Race Relations Policies in Britain: Agenda for

the 1990s", University of Warwick, Coventry, Centre for research
in Ethnic Relations.

- J. Arango (1991) "Perspectivas migratorias en el Sur de Europa",

"La Sapienza", University of Rome, Rome, 2nd European
Forum."

- C. Bonifazi, G. Gesano (1993) "Immigrazione straniera tra

regolazione dei flussi e politiche di accoglimento", Rome, IRP,

52
 Tendenze demografiche e politiche della popolazione.

- R. Cagiano de Azevedo (1992) "Migration from Mediterranean

developing countries to Europe. Demographic background and
political implication", Milan, Franco Angeli, Affari Sociali
Internazionali.

- D. A. Coleman (1994) "Trends in fertility and intermarriage

among immigrant populations in Western Europe as measures of
integration", University of Oxford, Departement of Applied Social
Studies.

- A. Golini, L. Ascoli (1991) "Movimenti della popolazione in

Europa e minoranze etniche","La Sapienza" University of Rome,
2nd European Forum.

- T. Hammar (1992) "European immigration policy. A comparative

Study", Cambridge.

- J. Leman (1991), "Intégrité, intégration. Innovation pédagogique

et pluralité culturelle", Bruxelles, De Boeck Université.

- C.V. Marie (1992), "Travail clandestin, trafic de main-d'oevre et

formes illégales d'emploi", Paris, Ministère des Affaires Sociales,
Rapport établi pour le Conseil National des Populations
Immigrées (CNPI).

- D.G. Massey (1988), "Economic Development and International

Migration in Comparative Perspective", in: Population and
Development Review, vol. 14, n.3.

- R. Penninx (1984), "Les populations immigrées et l'évolution

démographique dans les Etats membres du Conseil de l'Europe",
Strasbourg, Conseil de l'Europe, dans la: Série d'Etudes
Démographiques, no13.

53
 - M. Poulain (1993), "Migrations en Belgique en 1991", Université
Catholique de Louvain, Institut de Démographie, Rapport
SOPEMI, p.101.

- J. Rex (1988), "Policy Problems relating to Immigrant Minorities

in North-West Europe", Florence, in: Symposium on Population
Change and European Society.

- University "La Sapienza", Rome:

(1992) European Seminar:"Measuring Migrants Integration".

G.C.Blangiardo, C. Bonifazi, M. Bretz, R. Cagiano de Azevedo
and R. Di Prospero, O. Casacchia, A. Furcht, J.S. Macdonald
and A. Reid, E. Moretti, R. Penninx, M. Reginato, M. Ricci, M.
Roig, G. Siampos, M. Termote, E. Todisco, M. Tribalat, P.
Willems L. Deschamps and K. Pauwels.

(1993) European Seminar: "International Migrations as a Factor

of Security and Cooperation in Europe and in the Mediterranean
Region".
M. Behroun, G.C. Blangiardo, R. Cagiano de Azevedo, C.
Calvaruso, G. Caselli, A. Cavaterra, J.D. Nicolas, G. Gesano, M.
Van Haegenborgh, C. Hodgens, J.S. MacDonald, J.Malacic,
U.Mammey, P.M. Masotti, M. Miller, C. Nigoul, M. Poulain, N.
Safir, J.S. Martinez, A. Rizzi, P. Schatzer, G. Siampos, L.F.
Tomasi.
(1994) Eurostat and DG/V Research Project: Final Report
"Immigrantintegration policies in seven European Countries". M.
Anwar, G. Brochmann, R. Cagiano de Azevedo R. Di Prospero,
J. Leman, U. Mammey, C.V. Marie, B. Sannino, M. Van
Haegenborgh, S. Van Wersch.

- D.J. Van de Kaa (1991), "European Migration at the end of

54
 History", Wassenaar, NIAS, PDOD, paper n. 8.

- M. Weiner (1985), "International Migration and International

Relations", New York, in: Population and Development Review,
n.3,

55
 Measurements and indicators of integration
Marian Fitzgerald

Introduction

In this presentation, three main areas are covered:

- questions of how we actually measure integration;

- the availability of British data on those measures and at what they tell us;
- and some questions about the wider problems and dilemmas that they may
pose.

The perspective taken in the present document is that of a researcher who has been
working on these issues for many years in Britain, during which time she has closely
observed the development and implementation of the sorts of measures under
discussion here. And, more recently - not least through my involvement with the work
of the Council of Europe's Commission against Racism and Intolerance - I have
become increasingly conscious of the problems of international comparisons in this
field.

The measurement of integration

In international discussions of questions of integration, the statistics which are most

likely to be traded tend primarily to focus on the economic situation of different
groups. This may largely be because economic data are most readily available
anyway and, therefore, the most likely to be broken down, if at all, by national, ethnic
or racial categories.

If it is possible at all to 'measure' integration one needs statistics which meet three
main criteria:

- We need statistical data on a battery of measures - including economic

56
 measures (obviously) but with at least equal emphasis on social and
political indicators.

- The data should be available at a meaningful level of desegregation both

between and within the relevant groups.

- And finally the data need to have been collected consistently over a
sufficient length of time for trends to be identifiable.

The measures I suggest should include as a minimum the following illustrative list,
although there are many other very important measures, e.g. measures of housing,
of geographical dispersal or concentration and of levels of health).

The second aspect is the question of the availability of British data on these
measures and the broad patterns shown by what exists. Inevitably the different areas
can be touched only very superficially and, it is not practical here to talk about
figures.

Economic Social and Political

Job levels Naturalisation

Unemployment Intermarriage
Income Racial harassment and attacks
Language skills Discrimination
Educational attainment Political participation
Crime and relations with the police

Availability of british data

It's worth starting here with some background. Broadly speaking, the largest
minorities in Britain who have been the focus of `race relations' policies, have their
origins in countries formerly governed by the United Kingdom and which, as they
gained independence in the postwar period, maintained their links with the United
Kingdom as members of the Commonwealth. Among these, the groups of particular

57
concern have been non-whites. Speaking very generally, the three main reasons for
this initially were language problems, cultural differences, and racial prejudice on the
grounds of skin colour.

In the 1960s/1970s the main information available on groups of immigrant origin

were census data based on place of birth. However, starting in the late 1960s and
particularly in the mid 1970s, surveys increasingly began to quantify the
disadvantage suffered by these groups and to show that these disadvantages were
continuing into the non-immigrant generation. From the late 1970s, many local
authorities began to keep systematic records on these groups - but on the basis of
their ethnic origins rather than their place of birth. This added momentum to moves
towards get an ethnic origin question as such into the national Census in 1981.
These moves failed and a surrogate measure was invented whereby the 1981
Census data were used to estimate the ethnic minority population on the basis of the
numbers of people living in households where the head had been born in one of the
countries of the New Commonwealth or Pakistan. Clearly, though, this method
missed minority households where the head had already been born in this country
and it was, for this reason, likely to become increasingly less useful over time.

But from 1981 two major national data series adopted an identical ethnic origin
question -the General Household Survey and the Labour Force Survey. (The Labour
Force Survey, it may be worth mentioning, is conducted under the terms of a
provision of the Treaty of Rome and, therefore, has counterparts in all other EU
countries - although not in terms of all the specific items of information collected, and
certainly not in terms of ethnic origin data). And, during the 1980s, we saw a massive
increase in the extent to which local authorities kept ethnic records both on
employment within the authority and with respect to the delivery of its services.
Notionally at least, these data were analysed in order to `monitor' the authorities'
performance in one or more of three related areas: the impact of their policies on
different ethnic groups; identifying whether groups had different needs (and if so,
whether these were being met); and checking for possible discrimination - whether
direct or indirect.

This practice of ethnic monitoring used the census data as a benchmark for
comparison and it began from the mid 1980s also to extend to other agencies - for

58
example, the collection of ethnic data on the prison population. Further one-off
surveys were conducted specifically concerned with aspects of racial disadvantage.
Other surveys also began to include an ethnic breakdown routinely in the
demographic information which they collected on respondents. And finally, in 1991,
we had a completely new question on ethnic origin (as well as place of birth) in the
national Census which used the following categories. The published information
further subdivided the `Other' group into an `Asian Other' and an `Other' category.

Ethnic Question used in the 1991 Census:

White
Black-Caribbean
Black-African
Black-Other (Please describe).....................

Indian
Pakistani
Bangladeshi
Chinese
Any other ethnic group (Please describe)............

So the current situation regarding the data available nationally under the headings
listed is very broadly as follows:

- For job levels and unemployment information comes from the Census and
the Labour Force Survey.

- On income information comes from the General Household Survey,

although the overall sample size is relatively small and information is
inevitably limited since minorities form only about 6 per cent of the total
population. A question on income has been included in the Labour Force
Survey also since the end of 1992.

59
- In terms of language skills there is no information from regular national
survey series although a one-off survey with regards to the health of ethnic
minorities recently produced some useful information and more definitive
indicators may become available in the near-future from some important
one-off surveys.

- On educational attainment information about qualifications by ethnic group

comes from a range of sources - most importantly and most regularly the
Census and the Labour Force Survey. Other national sources include the
Youth Cohort Study sponsored by the Employment Department and ethnic
monitoring of applications, acceptances and performance in higher
education.

- Information on naturalisation is routinely available - although in the British

context this has tended to be a less important indicator of integration than
it might be elsewhere since a large proportion of the main minority groups
already had British nationality when they arrived as immigrants; and nearly
half of these groups were now born in this country and almost all will have
British nationality automatically as a consequence.

- In terms of inter-marriage information comes from the Labour Force Survey

and from the Census.

- On racial harassment and attacks data have been compiled by the police
using a common definition since the mid 1980s. (These cover incidents
reported to them which they have recorded.) And the British Crime Survey
since 1988 has included an ethnic minority booster sample which has
provided useful information on ethnic differences in victimisation and on
the proportion of victimisation which is perceived to be racially motivated.

- On discrimination the main data come from major one-off surveys of racial
disadvantage conducted, with government support, by an independent
research institute in the mid-70s and early 80s. Another is due to report
next year. There is also evidence of the numbers of cases brought under
the Race Relations Act before the courts and Industrial Tribunals.

60
- In terms of political participation there have been relatively few studies with
any quantitative component, although some election surveys have asked
about voting intentions and levels of registration among different ethnic
groups. There has been further information on electoral registration from
the post-1991 Census survey carried out by OPCS. The number of ethnic
minority MPs is now six (which doesn't pose a major data collection task);
and various bodies have made occasional trawls to establish the number of
ethnic minority councillors nationally.

- And finally in terms of crime and relations with the police we have
information from a variety of sources. Certainly in terms of attitudes
towards and perceptions of the police the survey data (including data from
the British Crime Survey) are by now quite extensive. But in the two most
important areas - ethnic minorities as victims of crime and ethnic minorities
as suspects and offenders within the Criminal Justice System - the sources
are diverse and rather patchy. On victimisation, the British Crime Survey is
now beginning to provide valuable information. In terms of ethnic minorities
as suspects and offenders there have been a number of ad hoc studies
over many years, but the main source of national data over time has been
the prisons' data mentioned earlier, which has been available since the
mid-1980s. This, however, only represents the end point of the Criminal
Justice System and it has been quite difficult to tell what happened at
earlier stages in that system. Only since 1993 there has been a
requirement of ethnic monitoring on police forces at a national level. To
date, this requirement has only covered occasions when the police formally
stop and search members of the public. It will extend to later stages of the
criminal justice process from April 1996. As yet, however, there is no
systematic information on decisions by the police, the Crown Prosecution
Service or the courts between these stops (only a minority of which actually
result in arrests) and the point of imprisonment.

A very superficial look at the picture which emerges on these various indicators from
the data available suggests the following:

- In terms of job levels and unemployment, the position of the Indian group

61
 comes closest to that of whites, with Black-Caribbeans overall not faring
very well and Pakistanis and Bangladeshis still worse.

- In terms of income there are similar patterns.

- Language skills, obviously, are more of an issue for some groups than for
others, in particular for the more recently arrived groups, including the
Bangladeshis.

- In terms of educational attainment the evidence is mixed when one takes

local variation into account. The national picture mirrors that employment
and income -with a very poor showing by males in the Black Caribbean
group masked by the relatively better performance of female co-ethnics
and with the Bangladeshis are areas of particular concern.

- On the question of inter-marriage very valuable data were collected in the

1980s' from the Labour Force Survey (LFS). Today the census categories
have been adopted but throughout the 1980's a different categorisation
system was used which specifically included a `mixed' group. The LFS
suggested that already by the mid-1980's in the under-15 age group there
were more children of mixed ethnic origins (who were predominantly black
and white) than there were black children as such. Fairly high rates of
Black-Caribbean/white unions have now been confirmed by the 1991
Census and seem to be rising steadily among younger people; although
the same pattern is not true as yet of the Asian groups.

- In terms of racial harassment and attacks, there is some disparity between

the various sources of information available but it would appear overall that
it's the Asian groups who are most susceptible to racial harassment.

- The evidence of discrimination as reported by survey respondents, on the

other hand, is that Black Caribbeans are more likely than Asians to see
themselves as victims of discrimination.

62
- In terms of political participation, it is important to remember that the main
groups have more or less since their arrival had the right both to elect and
to be elected. We have seen over time an increase in the number of local
councillors of all ethnic minority origins and more recently the number of
MPs, although this still remains small. On the other hand surveys have
suggested a falling-off in electoral registration particularly among young
men from ethnic minorities and low rates of voting particularly among Black
Caribbeans.

- Finally, the evidence on crime and relations with the police suggests that
all minorities hold a lower opinion of the police than whites. It appears that
the greatest disenchantment is among younger Black Caribbeans but older
Asians. In terms of victimisation, the British Crime Survey suggests that -
relative to whites -all the main minority groups are disproportionately likely
to be victims of crime, but it also shows that the reasons for these apparent
ethnic differences are quite complex. In terms of ethnic minorities as
suspects and offenders, the concern has always been about the
involvement of the Black groups with the criminal justice system and the
prisons data have consistently shown very high levels of over-
representation of these groups.

Questions arising

In looking at the questions and dilemmas concerning the measurement of integration

we have to determine criteria to be used and investigate in how far the British data fit
these criteria.

- On the first criterion we have by no means comprehensive data across

even the limited range of economic, social and political measures
identified.

- On the second criterion (the need for sufficient desegregation), there are
two main sets of problems. One is that because, relatively speaking, even
the largest groups are quite small, surveys will often fail to pick up
sufficient numbers for refined comparisons to be made. In particular the
Asian groups often tend to be lumped together in surveys simply to

63
 generate sufficient numbers. Yet it is evident that there are major
differences between these groups, with Indians on the one hand, and
Pakistanis and Bangladeshis on the other, occupying opposite ends of the
spectrum of disadvantage and the Caribbeans somewhere in between. In
the second place, to understand ethnic differences shown up by surveys
and other data, it is often imperative to be able to explore the data in terms
of other variables such as socio-economic group, area, family size, gender,
educational attainment, etc. Very often information on these variables is
not available; nor, realistically speaking, would it be practical to obtain it
without rendering the data collection exercise inordinately complicated and
expensive.

In as much as data are available, they tend to be limited almost exclusively

to the Black and Asian groups.

- Turning then to the third criterion (the need to be able to identify trends),
we have an obvious problem of inconsistency and discontinuity because of
the changing bases of ethnic categorisation which have been used over
time. Moreover, where broad trends are nonetheless discernible for the
main groups, it is questionable whether the available data are managing to
keep up with changes in the composition of the ethnic minority population
in general and the populations of most concern in particular. This is very
specifically illustrated by the fact that in the 1991 Census over 20% of all
ethnic minorities were classified in categories `other' than the ones which
were explicitly specified by the Census question. Indeed it is worth noting
that while our main preoccupation remains the groups of Caribbean and
Indian sub-continent origin, the proportion which these groups constitute of
the overall ethnic minority population has dropped really quite markedly. In
1966 they formed 86% of the minority `coloured' population (as it was then
called); yet in 1991 they accounted for only 66% of all non-whites. We
know very little about the remaining 34% of non-whites (and almost nothing
about minorities in the `white' group).

Nonetheless, the United Kingdom data are far more comprehensive and more nearly
meet the criteria set out than any other European statistics. On the other hand, it is

64
apparent that the United Kingdom data are far less comprehensive than United
States data which are in a league of their own in terms of the availability of
breakdowns by ethnic origin on virtually all socio-economic indicators and have been
available over a long period. Even so, the Americans have had their own problems
with changes in categorisation and a significant ongoing difficulty with incomplete
and unreliable census data on certain groups. Most striking, though, is the way in
which statistical evidence of ethnic differences has come in the United States almost
to stand as an explanation for social and economic variations in the population rather
than the other way round.

The problem seems to be that, if a particular configuration of questions becomes

identified with certain groups over a period, an ethnic pecking order begins to
establish itself. This tends to occur especially where the problems appear not to be
amenable to policy intervention. The result - implicitly or otherwise - is that certain
ethnic groups become designated as those who will do well (perhaps with a little bit
of help in starting up) while other groups remain 'a problem' on an almost permanent
basis. This of course readily feeds in to the current and growing debate about an
'underclass' which, certainly in the United States, has readily tended to be conducted
in ethnic terms.

What does all this add up to in terms of the implication for our ability to "measure"
integration? To go back to the outline of the UK data, one can see a picture building
up in which it becomes clear that not all minority groups are equally disadvantaged
and that the most disadvantaged aren't necessarily disadvantaged in the same ways.
At one level there is no problem with this. In policy terms a generic approach based
on the assumption that all minorities are disadvantaged and are disadvantaged in the
same way and for the same reasons, would not only be enormously wasteful in
resource terms, it would be doomed to failure. On the other hand one can readily
see how the necessary distinctions between ethnic groups can lend themselves over
time to the positive stereotyping of some groups and the negative stereotyping of
others. In the criminological literature, for example, one finds even self-consciously
'liberal' criminologists speculating about why (as they see it) Caribbeans are
disproportionately involved in crime when Asians aren't. Some of the terms of this
debate are so fanciful and so ill-informed as to be laughable, but it has a far more
sinister side. Currently we are seeing moves afoot amongst certain American and
European criminologists to explore at an international level why - as they see it -
certain minorities simply are more 'criminal' than others despite facing apparently

65
similar levels of disadvantage.

There is then a tension and a difficult balance to be struck.

On the one hand we can agree that, in principle, we need more and more refined
forms of measurement across a very wide range of indicators of integration and we
need these for four main reasons.

- First we need statistics to establish the extent and the specific ways in
which different groups are not yet integrated within the mainstream of
society. (And by `integrated' I mean able - without surrendering their ethnic
identity and culture - fully to participate in all aspects of national life and to
benefit from that participation at both an individual and a group level.)
- The second purpose is to try to establish the reasons for the extent and
form of this exclusion.

- Thirdly we need the information under the previous two headings to inform
the development of policies intended to promote integration.

- And fourthly, we need the information over time as a scientific check on

whether such policies are proving effective.

On the other hand we are faced with a number of questions to which there are no
easy answers but which are simply set out for debate. They are of three very
different types:

- Firstly, is it realistic to expect to generate data which would meet the

criteria set out earlier? And, if not, what is the best we can hope for? Which
indicators should be prioritised and for which groups? How can the need
for continuity be reconciled with the need to keep track of change? Above
all, given (a) that reliable and comprehensive data are probably too much
for hope for and (b) that data in any case do not explain themselves, what
other tools and research approaches are required - and in what
combination?

66
- Secondly, there may be serious dangers in aspiring to make international
comparisons given a) the pitfalls at national level b) that progress towards
overcoming them will be very uneven indeed and c) that the type of data
needed and the basis on which it is collected may, in any case, legitimately
differ quite markedly from one country to another.

- Thirdly, is it possible to avoid scoring an own goal? That is, is it possible to

generate, publicise, use and compare such data over time without the risk
that they may end up militating against integration rather than working in its
favour?

67
 International migration statistics
Thana Chrissanthaki and Lars Ostby
(EUROSTAT)

Introduction

Comparability ...

Eurostat launched its programme on international migration statistics in 1988. An

initial study was carried out by M. Poulain, University of Louvain, on behalf of
Eurostat on the collection systems existing in the EU member states, for international
migration flows and on the possibility of harmonising these statistics. The study
showed the broad differences that existed - and still exist - among the various
countries with respect to the collection methods, definitions and characteristics. The
conclusions of the study included a set of tables that could be used as a starting
point for collecting data.

Mostly, these data cover the flow of migrants and the stock of immigrant population.
On the level of the European Union, not much information is available on the
situation of migrants and none at all relating directly to their integration. The present
paper will focus on Eurostat flows and stock data.

The current data collection, based on a revised set of tables agreed by all member
states, started in 1991. The at that time 12 EU countries and the 7 EFTA countries
were asked to transmit to Eurostat detailed data from 1985 onwards for detailed
information and from 1960 onwards for more aggregated data. The processing of
that huge amount of information lowered the initial enthusiasm and optimism. The
problems were more complicated than expected. The work programme was thus
modified according to these findings. It was decided to work on two parallel axes:

- methodology, and in particular the improvement of data quality and

comparability;
- collection of existing information.

68
Flows

Migration movements can be measured in two different perspectives: at the place of

departure and at the place of arrival. In theory, figures reported by the countries of
immigration should be identical to the figures reported by the countries of emigration.
In practice, however, substantial differences may exist when both figures are
compared. Why? Several factors might help to explain this phenomenon.

- Firstly, especially for international migration, procedures for obtaining a

residence permit differ from country to country. In some countries, in order to obtain
a residence permit one has to register. For example, an immigrant from Africa
(non-EU citizen) obtains a residence permit by registering in Belgium, Denmark,
Germany, Spain, Italy, Luxembourg and the Netherlands. In the Netherlands (s)he
has to reside for a certain period (6 months). In Germany (s)he should be residing as
owner-occupier, tenant or sub-tenant in order to be allowed to register.

The treatment of different citizenship groups may vary. For example, a Dutch citizen
returning to the Netherlands has to reside there for at least 1 month, whereas for a
person without Dutch citizenship the minimum period is 6 months.

- Secondly, another source of disparity is the different collection systems

used. Some countries use registers while others use surveys. Each of the systems
has its advantages and disadvantages: a survey might have a sample problem, a
register allows only those who comply with a certain set of registration rules to enter,
and it records entries more efficiently than departures. Comparing figures derived
from these sources is not straightforward and of course difficulties arise when doing
so.

- Thirdly, not all groups of migrants are included in the statistics. For
example, asylum-seekers are sometimes excluded from the immigration and
emigration statistics, sometimes not. Only Portugal seems to include all
asylum-seekers in its immigration statistics (its emigration statistics only include
nationals). For the other countries it is not known what portion of the asylum-seekers
is included. On the other hand, invited or resettled refugees are counted as
immigrants in almost all EU countries, but cannot, with some exceptions, be

69
identified as such.

- Fourthly, one should mention the availability of data. For example, France
does not provide any international emigration figures at all and its international
immigration figures refer only to non-nationals. On the other hand, Spanish data
capture only a small part of the emigration flows (nationals). Of course, recent
developments such as the establishment of the free movement of EEA citizens
within the EEA countries has also meant that certain sources of data, such as
residence or work permits for those people, have been eclipsed. The difference
between international and internal migration is fading in the case of EEA citizens
moving within the EEA region.

Stocks

The definition of foreign population in a country can change according to the

purpose. For example, for integration purposes information on the foreign-born
population is sometimes more useful than that on foreign citizens. Attitudes or acts
of discrimination will hardly stop after the acquisition of citizenship. On the other
hand, for the right to participate in elections, and in other legal aspects, the
citizenship element will be crucial. For some purposes neither citizenship nor country
of birth is appropriate. Countries with a good population register have developed
classification systems taking into account even the parents' country of birth.

Again, as with flows, there are difficulties in the data collection of stock statistics.

- The concept of the migrant stock differs from country to country: some use
the foreign-born persons, others the non-national population or ethnic groups,
etc.

- Not only are there differences among countries in the collection systems
used but there might also exist different sources within one and the same country; for
instance, censuses and administrative records or surveys. The census is an
important source in many countries. The problem is that it cannot provide annual
data, and inter-census estimates are difficult to calculate, especially for small

70
population groups like non-nationals of a certain age from a specific country.

... and harmonisation

1. United Nations recommendations on international migration statistics

At Eurostat's annual meetings with the member states (Working Party on Migration
Statistics), where Eurostat's future programme on international migration statistics is
presented and discussed, many EU-, EFTA- and also several central and eastern
European countries have expressed their dissatisfaction with the existing United
Nations recommendations on international migration statistics. The United Nations
drew up these recommendations in 1976 in a period when the international political
scheme was completely different. Countries were strongly encouraged to use these
recommendations. Unfortunately few, if any, do so completely. These
recommendations, being the result of a consensus between all UN-countries, were
too complicated, and national collection systems could not comply with them.
Needless to say, the 1976 recommendations are out of date now. Eurostat was
urged to do something.

At the beginning of 1994 Eurostat started to work in very close co-operation with the
Statistical Division of the United Nations, New York, on the revision of these
recommendations. At their first meeting UN and Eurostat representatives drew up a
plan for the next few years. The first step was the launching of Eurostat's special
study on international migration statistics. This study included a detailed
questionnaire on the implementation of the existing UN recommendations on
international migration statistics. Eurostat has processed the answers, and
presented results in different meetings. A report was finalised in 1995.

All member countries of the Economic Commission of Europe (UN-ECE) have been
invited to participate in the special study on international migration statistics. In a
second stage selected countries from Asia, Africa and America were involved, as
well as the other regional commissions of the United Nations.

Based on these preparations, a new round of discussions started in 1995, including

meetings with international experts and discussions with representatives of countries

71
(or groups of countries). On the basis of the results of an expert meeting at global
level in New York on 10-14 July 1995, the work on a suggested new set of
recommendations will be finalised.

A publication will be issued later, based on the decisions taken by the Statistical
Commission.

2. Statistics on asylum-seekers and refugees

Special attention is given to the issue of asylum-seekers and refugees. The

statistical definitions differ from country to country and are not comparable. There
are no clear links between asylum-seeker statistics and international migration
statistics. The same applies to statistics on refugees. These issues are not treated
at all in the current UN recommendations. At the time when these recommendations
were discussed, the problem of asylum-seekers was not at all as pressing as it is
today.

Eurostat started to work on the recommendations on asylum-seeker and refugee

statistics. It is not clear yet whether such recommendations will be linked to those
on international migration statistics or not, and it remains to be seen whether they
will be at United Nations or European level. A draft paper was prepared, and
discussed with some countries at a joint UNHCR/Eurostat Workshop in Geneva in
May 1995. A revised paper was discussed at the Expert Group Meeting on
International Migration Statistics in New York.

The Netherlands Interdisciplinary Demographic Institute (NIDI) prepared, on behalf of

Eurostat, a study on the statistics of asylum-seekers and refugees in the 15 EU and
4 EFTA countries. The study focuses on the relationship between asylum-seekers
and refugee statistics on the one hand and on the definition of an asylum-seeker and
a refugee on the other. One of the main conclusions of this study is that there is
such a big difference in the way these groups are counted in the different countries
that it is not advisable to produce comparative tables.

Data collection: better and faster ...

72
The second part of Eurostat's work is the collection of existing data. There is scope
for improvement given the existing data source constraints. A series of projects
aimed at improving data quality and timeliness will be financed in the EU and EFTA
countries. These projects aim at improving existing data sources, using new
sources, examining differences between figures reported by the countries of
immigration and emigration or examining special issues (such as the link between
stocks and flows, family reunification and family formation, etc.). These projects are
carried out mainly by the National Statistical Institutes in the various countries.

... and more...

Other areas of interest of Eurostat are the integration of migrants and illegal
migration. After extensive discussions on integration, the next step would be to
construct a series of indicators to be tested in some countries in order to see whether
the results are reliable. As for the issue of integration, it is unnecessary to explain
why more work has to be done in the field of illegal migration. It is an issue of
interest to many countries, especially countries in southern Europe, and it is a very
difficult one. Again caution is needed, as it is a phenomenon very hard to quantify.
"Statistics Denmark" will carry out a project in co-operation with Eurostat and Inter
Governmental Consultations to collect information on the estimated number of
illegals in the member countries, and to evaluate as far as possible the estimation
methods.

... but first of all ...

The data collection is the backbone of Eurostat's work. Data on international

migration are collected from the 15 EU member states and the 4 EFTA countries.
Eurostat sends an annual questionnaire to the National Statistical Institutes. This
questionnaire was drawn up together with the ECE in 1993. From 1994 onwards, it
was used jointly by Eurostat, ECE and UNSTAT. Further efforts will be made to
reduce the burden on member states by trying to merge questionnaires containing
the same questions.

As those who are familiar with Eurostat publications on demography and migration
will know, the coverage of our statistics is not wide enough for its purposes: from a

73
regional point of view and concerning the time factor. Within a reasonable period
after the end of a reference year, hardly fifty per cent of the figures needed to
complete a migration matrix were sent it. When it comes to break down by sex, age
region etc., the situation is even worse. In order to enhance comparative research
possibilities, it might be better to improve the data situation in countries with bad
data or none at all, than to improve the information from countries with well
developed systems. If migration questions will maintain, or even increase, their
political significance, it will be important to have better data to support the political
decision-making process.

The data received by Eurostat are official statistics. All information is stored in
Eurostat's Social Data Bank. This data bank is not yet public. Researchers of
University College of London, assist in making the documentation of the migration
data as clear and comprehensive as needed for external users. When this work is
completed (early 1996), the migration data will be opened for external users.

... what we got...

The information gathered, refers to both migration stocks and flows:

- immigration and emigration by different characteristics;

- asylum-seekers and refugees;
- acquisition of citizenship;
- population by citizenship and other characteristics.

The following three items cover the situation of migrants, but there is only a very
limited number of variables included. Neither the timeliness of the data, nor the
country coverage are satisfactory:

- economically active population by citizenship;

- employees by citizenship;
- students in higher education by citizenship and other characteristics.

74
A list of the tables can be found in Annex I.

The period covered is, in principle, from 1985 onwards. Historical aggregate series
exist from 1960 onwards and data on foreign employees from 1983 onwards.

If you want to know more...

Eurostat has been engaged since the beginning of 1993 in a very active publication
programme in the field of migration statistics. The first publications were in the form
of what is called "Statistics in Focus", (previously "Rapid Reports"), small information
leaflets giving quick information on a specific subject. Four such "Rapid Reports"
have been published: on non-national population, migrant women, flows and
asylum-seekers, and a fifth, on naturalisation, is under way.

The results of the study conducted by the NIDI on behalf of Eurostat concerning
asylum-seekers and refugees have been published in two volumes. The first volume
dealing with the EU countries was published in early 1994, the second volume
dealing with the EFTA countries was published in late 1994.

Finally, a Yearbook on migration statistics was published in late 1994. A new edition,
somewhat different from the previous one will appear before the end of 1995.

75
 ANNEX I
Eurostat publications

Statistics in Focus (Rapid Reports) "Population and Social Conditions":

1993-6 Population by citizenship in the EC - 1.1.1991

1993-8 Female population by citizenship in the EC - 1.1.1991
1993-12 International Migration Flows in selected EC countries - 1991
1994-1 Asylum-seekers in the EU: Better data needed
1994-7 Non-nationals form over four per cent of total population in the European
Union - 1.1.1992
1995-3 International Migration in the EU Member States - 1992
1995-? Acquisition of citizenship 1980-1993

Other publications:

Migration statistics - 1994

- Asylum-seekers and refugees, a statistical report, Volume 1: EC Member

States, 1994
- Asylum-seekers and refugees, a statistical report, Volume 2: EFTA
countries, 1994
- Demographic Statistics 1995
- Labour Force Survey - Results 1993
- Education across the European Union - Statistics and Indicators
- Causes of international migration - Proceedings of a workshop,
Luxembourg, 14-16
December 1994
- Proposals for the harmonisation of European Community statistics on
international migration, 1990
- Proposals for the harmonisation of EFTA statistics on international
migration, 1991

76
 Equivalency and equal access as standards
for ethnic minority policy
in The Netherlands
J. Dagevos

Social integration of minorities in the netherlands: equivalency and equal

access

1. Introduction

Although, in a welfare state, the answer to the question: Who gets what and why?
should be based on well defined principles of social justice, decisions regarding the
distribution of social goods (eg wealth, income, education, labour, housing) are
based on political realities. As a consequence, it remains unclear what standards are
valid if we want to make a judgement on the extent to which social inequality is
acceptable and whether a certain distribution of social goods is fair. In the
Netherlands there is one main exception; in order to formulate the main objective of,
and also to evaluate the policy aimed at ethnic minorities, the Dutch government has
developed a well defined standard of distributive justice.

The foundation for the policy aimed at ethnic minorities dates from 1983 when the
policy document Minderhedennota was published. In this document the distribution
of social goods is seen as fair if members of minority groups have an equal share in
the distribution of these goods. As a consequence, the main objective of this policy is
to realize that minorities will have their equal share. This is the standard known as
equivalency. To stimulate this, minorities should have equal access to important
government provisions, for example the employment office and the educational
system.

The crucial point here is the operationalization of equal share or equivalency. This is
the translation of the idea of distributive justice in this social area. The government
has chosen clear indicators in order to measure equivalency (eg educational

77
achievements, unemployment rates). Furthermore, the idea of equal share has to be
measured against indigenous Dutch with comparable characteristics. The Dutch
government wants to monitor the development of the socio-economic position of
ethnic minorities. Since 1986 the Institute for Sociological and Economic Research
(ISEO) of the Erasmus University Rotterdam has been responsible for this project,
known as the T&E project.

2. Ethnic minorities in the Netherlands

More than 2 million inhabitants of the Netherlands are not indigenous Dutch people
(15%). This number includes people from industrialized countries such as Germany,
Japan, the United Kingdom and the United States. We do not consider them to be
ethnic minorities. In defining an ethnic minority two elements are considered to be
important:

- the, on average, low (or disadvantaged) socio-economic position

in Dutch society, and
- the fact that the group does not originate from the Netherlands.

An implication of this operational definition is that people from the former colony "The
Netherlands East Indies" (which is now the Republic of Indonesia) - who were
repatriated at the end of the 1940s and the beginning of the 1950s - are not to be
considered as ethnic minorities because they absorbed quite well into Dutch society.
For the Italian and Spanish group this could be in prospect: because of the fact that
the social position of these groups has improved greatly in the last decade, voices
are heard demanding that Italians and Spaniards should no longer be considered as
ethnic minorities.
Taken together, about 950,000 people belong to a minority group. This is about 6%
of the Dutch population. There are four main minority groups in the Netherlands;
Turks (240,000), Moroccans (195,000), Surinamese (260,000) and Antilleans
(90,000). Turks and Moroccans were originally labour migrants who came to the
Netherlands in the late sixties and early seventies. This was followed by family
reunification (wives and children) and family formation (marriages). The main inflow
of Surinamese into the Netherlands occured in the mid-seventies when Surinam
became an independent republic. Nowadays more than one third of the Surinamese

78
population lives in the Netherlands. The migration of Antilleans is an ongoing
process in the past decades. Their motives are very different: education, work, social
security, marriage, et cetera. They are Dutch citizens and because of this Antilleans
do not encounter restrictions when they want to migrate to the Netherlands.

3. Research project T&E

The purpose of the research project is threefold. First of all, it has to describe the
(developments in the) socio-economic position of minority groups in the Netherlands.
In the descriptive studies, the main purpose is to determine to what extent minorities
have an equal share in the distribution of social goods. These social goods are
represented by a set of indicators. The second purpose of the T&E project is more
analytical. In the analytical studies we try to determine to what factors the unequal
share in the distribution of social goods can be attributed. The third aspect of T&E is
to gather reliable data concerning the socio-economic position of ethnic minorities. In
the following we shall give a brief outline of these three purposes of T&E.

3.1 Description

When the project started in 1986, the indicators used for the description of the socio-
economic position of minority groups were more diverse than they are now.
Indicators were not only chosen for education, labour and income, and housing but
also for health care and welfare. During the research project there was more and
more emphasis on: (a) education, (b) labour and income, and (c) housing. These
aspects of the socio-economic position have got most emphasis in Dutch policies
aimed at ethnic minorities. For these areas the relevant indicators are:

Education
1. the percentage of a category/group attending school
2. the percentage of a category/group per educational level
3. the percentage of the category attending school per educational
level
4. exit level (with diploma)

Labour and income

79
 1. labour market participation
2. unemployment (rate and duration)
3. labour position
a. labour income
b. job level
c. type of labour contract (permanent vs. temporary jobs)
d. number of hours worked per week
e. occupation
f. quality of work
h. the percentage of wage earners vs. entrepreneurs
4. source of income (labour, social security)
5. size of income
Housing
1. tenure
2. dwelling characteristics
a. type of dwelling
b. age of dwelling
c. housing quality (equipment)
d. number of persons per room/dwelling
e. per centage of the income used for housing (rents/mor-
tgage)

The main objective of the policies aimed at minorities is the realization of

equivalency. The question is: when can one speak of equivalency? Our starting point
is that the position of ethnic minorities has to be measured in comparison with native
Dutch who have the same characteristics. It is useless to compare minorities and
native Dutch without taking these differences in characteristics into consideration.
We illustrate this with the following example: We compare the unemployment rates
of indigenous Dutch and minority groups. In the Netherlands the unemployment rates
of ethnic minorities are high: 20 to 30% of the workforce is unemployed. The
unemployment rate for indigenous Dutch is much lower: 6%. Following the logic of
the monitoring system it is not yet possible to conclude whether there is equivalency
or not. It is necessary to make a comparison between two groups with the same

80
socio-economic characteristics. As we all know, the unemployment rate is heavily
influenced by educational achievement. Therefore, a more appropriate way is to
compare the unemployment rates of indigenous Dutch and members of minority
groups with the same educational achievements. The results of this analysis are in
Table 1, see page ......, Appendix 1).

Even if we compare those categories with the same educational levels, the
unemployment rates of ethnic minorities and indigenous Dutch still differ. We
therefore have a strong indication that there is no equivalency as far as unem-
ployment is concerned.

3.2 Analysis

In the analytical studies we try to focus on the explanation of the disadvantageous

socio-economic position of various minority groups. A major feature is to elaborate
on the comparison between native Dutch and minorities; in analytical studies the
comparison is more detailed since we control for more social characteristics and also
include cultural attributes of minority groups in the analyses. Let us return to our
example of different unemployment rates. Unemployment rates differ not only on
educational achievements. Other factors like gender, age, occupational experience,
residential area and language proficiency are also important. Sophisticated statistical
methods are necessary to adequately compare native Dutch and ethnic minorities on
more than one or two variables. The results of these analyses show that members of
ethnic minority groups have higher unemployment rates than native Dutch with
similar characteristics. We have done the same type of analysis for other areas of
social life such as education, income and housing.

3.3 Data provision and the SPVA survey

The third objective of T&E is the improvement of data provision on minorities. When
we started the T&E project in 1986, it was almost impossible to gather information
about the unemployment rate of minority groups, let alone figures about the quality of
housing or labour contracts. At that time minorities were either not registered as such
in the public data systems or nationality was the only criterion of identification. The
Antilleans and most of the Surinamese are Dutch citizens and therefore cannot be
identified as minorities with reference to their nationality. To make the monitoring of

81
the socio-economic position of minorities possible, the government asked ISEO to
hold a survey among the four largest minority groups. Because of the fact that
equivalency is a central objective of the government's minority policy, the survey
should make it possible to compare minorities with native Dutch. In 1988 the first
survey was held. A representative number of Turks, Moroccans, Surinamese and
Antilleans were interviewed on their situation on the labour market, their educational
achievements, their income and their housing situation. The next survey was held in
1991 and recently we finished one that was held in 1994. Those three surveys
enabled us to monitor carefully the developments of the socio-economic position of
ethnic minorities in the Netherlands.

Another reason why we now know much more about minorities is that methods of
registering minorities in official statistics have improved and, more important, the
methods of registration in those statistics are becoming increasingly comparable. In
the early nineties ISEO examined the pros and cons of the different possibilities
concerning the identification of minorities. This examination eventually resulted in the
development of a registration system. This system is based on the person's place of
birth and the place of birth of (one of) the parents. If he or one of the parents is born
abroad, the person is not considered to be an indigenous Dutch. This system, though
controversial for some time, is now accepted as the standard and is in use in many
official statistics. (The results of the study concerning the identification of minorities
are summarized in Appendix 1.)

4. Results

After almost ten years of monitoring ethnic minorities in general two conclusions can
be drawn. Firstly, their socio-economic position has improved if we compare minority
groups now and ten years ago. The conclusion has to be more pessimistic if we
make a comparison between minority groups and indigenous Dutch. The gap that
existed in different areas of social life (education, labour, income) has not decreased
in the last ten years, on the contrary in some areas the gap has even widened. The
main reason for this is that the improvement of the socio-economic position of the
minorities did not keep pace with the improvement of the socio-economic position of
the Dutch population. There is one exception to this. The gap between minorities and
the Dutch narrowed as far as housing is concerned. This is noteworthy, since in the

82
Netherlands this area of social life is heavily influenced by the government. This is at
least an indication that government policy mitigates market forces and leads to a
distribution which is fairer as far as the socially disadvantaged are concerned.

5. T&E: Criticism and comments

Although the T&E report system has its merits, it also has its limitations. A first
objection against T&E is an ideological one concerning the notion of equivalency.
The assumption which hides behind the notion of equivalency is that the behaviour of
the `average' Dutch population is taken as the standard for the behaviour of the
members of minority groups. The opinion of minorities does not seem to matter. A
subjective test on equivalency, in which the needs, wishes and necessities of the
minorities are considered does not exist.
To a certain extent we agree with this kind of criticism. It is especially valid as far as
the descriptive studies are concerned. But here we would also like to stress that in
our descriptive studies we only use the notion of equivalency as a first test - without
any normative implications and being very aware of the limitations of the notion of
equivalency. The objection of ethnocentrism is not valid, or to a much lesser extent,
for the analytical studies. In these studies we explicitly include cultural factors in our
analysis and try to determine to what extent they contribute to the explanation
whether equivalency is realised or not. We therefore strongly disagree with the critics
who state that we do not ke into consideration any differences in culturally based
wishes and aspirations between minorities and native Dutch.

A second main objection concerns the comparison of ethnic minorities and a

reference group of native Dutch. A point that has been made is that one is never
sure if one selects the right characteristics when defining the reference group. First of
all we would like to emphasize the importance of the comparison of a reference
group and the ethnic minority group. It is our expercience that the results are much
more convincing in discussions with policy makers and public opinion than without
the aforementioned comparison. Secondly, we do not randomly choose the
characteristics of the reference group. For example, if we want to study the labour
market position we do not include the brand of someone's trousers into the analysis.
The selection of characteristics is based on the study of relevant findings in research
publications. On the basis of these findings, we consider for example age,
occupational experience, gender, language proficiency, duration of stay in Holland,

83
educational achievements to be relevant characteristics if we want to find out
whether or not labour market equivalency is realized.

APPENDIX II

In theory there are different methods to identify minority groups. Nationality is a

criterion, but in the Dutch context it is rather useless because of the fact that one
cannot identify Antilleans and most of the Surinamese with this criterion. In addition,
because of the increasing numbers of mediterrenean people who opt for Dutch
nationality this criterion will be no longer adequate to identify a major category of
these groups. The criterion country of birth of a person does not have this
disadvantage. It is especially suitable for identifying minorities who belong to the first
generation. The main drawback of this criterion is that it is not possible to identify
members of the second and third generation, i.e. persons who are born in the
Netherlands. If one wants to identify the second and third generation, one needs
information on the country of birth of the parents (second generation) and even the
country of birth of the grandparents if it concerns the third generation. The gathering
of information concerning the country of birth of grandparents is not without
problems. Firstly, one has to add four more questions on the origins of the
grandparents. Secondly, and most important, people who have to answer this
question often do not understand the ratio behind this question and/or do not know
the country of birth of their grandparents. In a sample of 240 ethnic minorities 21% of
the respondents due to the reasons mentioned above did not want or were not able
to answer the question on the country of birth of their grandparents. Mainly because
of these practical arguments the criterion "country of birth of the grandparents" has
not been added in the registration system of minorities in Holland.

The criterion of self-identification - the group to which he/she considers to belong -

meets some of the objections mentioned above. With this criterion it is possible to
identify third generation minorities without bothering them with questions about their
grandparents. Notwithstanding these important advantages the government decided
that self- identification is not a proper way of identifying minorities. According to the
policy makers self-identification may lead to opportunistic behaviour by minorities. As
a consequence, for registering minorities it is no longer allowed to use the criterion of

84
self-identification.
It is important to note that the choice for a certain combination of criteria of
identification has important consequences for the extent to which an ethnic group
can be identified. In table 2 we show some results for major minority groups in
Holland.

85
 The Dutch Law of equal participation on immigrants in the labour
market (WBEAA)
Karin Alfenaar

Introduction

This paper describes the background of the law on equal participation, some legal
details and its implementation. Furthermore, reasons identified by employers and
others for the present unsatisfactory state of implementation of the law will be
analyzed. Finally, means that could be used to increase efforts to achieve equal
participation of immigrants in the labour market are suggested.

Since 1 July, 1994, the Dutch Law on equal participation of immigrants in the labour
market (WBEAA) compels Dutch employers to report annually to the Chamber of
Commerce on the number of ethnic minorities employed and on efforts made to that
end. The law's objective is to increase the number of immigrants employed by
companies and public services, in proportion to the number of immigrants available
on the labour market in that region. It is important to note that no legal quota has
been fixed.

The law has been applicable for more than one year. However, the number of
employers complying with it is disappointing. A total of only 3,000 out of 30,000
(10%) of employers in the Netherlands, meet the legal obligations.

Although all companies should have reported to the Chamber of Commerce by 1

June 1995, only 737 employers have done so. It was revealed that out of the 13
ministries only 4 had reported and the Lower Chamber stressed that ministries had to
give the right example to the private sector. Finally, all ministries complied with the
law.

An evaluation of the law will take place in mid-1996. It is likely that prior to this
evaluation, no sanctions will be imposed on companies which do not comply with it.

86
1. Background

In the Netherlands a broad political consensus exists on the idea that concrete
measures should be taken in order to come to a more equal allocation of work
among the participants in the labour market. This is due to the fact that, as in many
other European countries, unemployment among ethnic minorities is much higher
than the average unemployment rate. This subject takes therefore an important place
on the political agenda.

In 1986 the Dutch (governmental) Advisory Committee for Research on Ethnic

Minorities (ACO) advised the government to provide for equal opportunities by law.
Three years later the Scientific Council for Government Policy (WRR) advised to
introduce a law on equal participation in the labour market.

The WRR based its advice on experiences in Canada where an 'Employment Equity
Act' has been in force since 1986. This law compels employers to draw up a plan
every year with a strategy to realise equal participation. Canadian employers also
have to write an annual report on their progress. The law also contains the
instrument of contract compliance. "Contract compliance" implies that only
companies that are able to prove that serious attempts have been made to recruit
ethnic minorities, can obtain orders by public authorities.

The WRR advised to adopt a law similar to the Canadian Employment Equity Act in
the Netherlands. It also recommended to introduce "contract compliance". The Dutch
government indicated that the adaption of a law on equal participation would be a
envisaged. However, it rejected the "contract compliance" - clause.

In 1990, there was quite a good political basis for a law on equal participation.
Nevertheless, it took four more years before such a law was actually introduced.

The reasons can be attributed to the unions and employers' organisations that
managed to postpone the introduction of a law by concluding a joint agreement on
the achievement of equal participation of immigrants by the end of 1994. This meant
that these parties promised to create 60,000 extra jobs for ethnic minorities before
the end of 1994. Especially the employers were strongly opposed to a legal

87
arrangement. The unions also feared that the consultative structure among
government, unions and employers would be damaged, if legal measures were
introduced. This consultative structure is typical for the social-economic relations in
the Netherlands.

In the first few years, perceptible efforts were made in order to implement the
agreement of the social partners. Evaluations made clear that the majority of the
employers did not even know about the existence of the so-called 60,000-jobs
agreement. The number of immigrants employed as a result of the agreement
amounted not to 60,000 but only to a few hundred. These poor results brought the
subject of a law on equal participation back again on the political agenda.

After a difficult discussion and in spite of a strong opposing lobby by employers'

organisations, Dutch parliament finally adopted the Law on equal participation of
immigrants in the labour market in 1994.

The Dutch Centre for Immigrants (NCB) was very happy with this development
having pleaded in favour of legal measures for equal participation for a long time.
The unsatisfactory implementation of the earlier agreement on the 60,000 jobs
emphasised the importance of this kind of measures. NCB thus supported the bill in
public discussion and in its contacts with government and parliament.

2. The law in brief

As already underlined in the introduction, the law does not include a quota system.
The law compels employers to report on the equal participation of immigrants in their
company, and to develop an internal policy by means of a plan of action in order to
correct any inequalities.

In the public report the employer is to indicate the total amount of employees working
for his company, and the per centage belonging to an ethnic minority. An explanation
must be added. A description should be given of the means by which the employer
hopes to establish an equal participation in the following year. The report must be
made available to the Chamber of Commerce and have an auditors' certificate. The

88
failure to submit such a report on the previous year by 1 June of the following year
consitutes an economic offense. It is possible to prosecute offenders and sentence
them to a considerable fine or even imprisonment as the publication of the data
alone may not stimulate employers sufficiently to take the necessary measures. The
WBEAA also contains an obligation for the employer to adopt a plan of action
indicating:

1. the employers assessment of the qualitative and quantitative participation

of immigrants in his company;

2. which employer's concrete objective on how to achieve equal participation

of immigrants in his company;

3. which concrete measures he considers essential, and is willing to take, in

order to realise this objective.

The works council of the company has the right to be consulted on the plan of action
and to take initiatives in this field.

The legislator considers the drawing-up of a plan of action an internal matter of the
company. Therefore the WBEAA does not refer to a date limit by which the plan of
action must be finalised, neither does it contain the obligation to publish the plan.
The most important details are already included in the annual report. Furthermore,
publication could stop employers from stating clear and ambitious objectives in their
plan of action out of fear of external attacks.

3. Assistance to the implementation of the WBEAA

After the adoption of the WBEAA, extensive information material was developed.
The Ministry of Social Affairs, responsible for the implementation of the law,
published several information brochures. In cooperation with NCB, a manual for
employers was developed. It includes models for the drawing - up of the annual
report and plan of action.

Several other institutions have also developed models and guidelines for employers.

89
Moreover, the Department of Employment offers special services to employers
willing to recruit more immigrants. Their management consultants on immigrants give
advice, free of charge, on drafting-methods for the plan of action and on available
subsidies.

4. Objections to the WBEAA

The four main objections to the WBEAA concern registration and protection of the
private sphere, bureaucracy and costs, the effectiveness of the law, the fear of public
criticism.

1. The most important argument concerns the registration of immigrants. Registration

of ethnicity is a very sensitive subject in the Netherlands since the Second World
War. Nevertheless, this argument is generally not heard by immigrants or their
organisations. It is more common for employers to use the privacy argument when
expressing their views against the law.

The law has been drafted very carefully and offers sufficient remedy against the
misuse of information on the national background of employees. There are a few
laws in the Netherlands which put privacy more at risk than the WBEAA.
Nevertheless these laws meet much less opposition. The Dutch Registration Office,
responsible for privacy protection in the Netherlands, concluded in a preliminary
examination in 1995, that the safeguards and standards on privacy protection as laid
down in an earlier stage, are complied with in practice.

According to NCB resistance against registration can be softened by a careful

introduction in companies. Consultation should take place at all levels. Sufficient
information on the reasons for and implications of the law is needed. In contacts with
companies NCB receives positive feed-back about this procedure. It can be
concluded that a careful approach leads to the desired results. Resistance may
disappear step by step, since registration will become a matter of routine, generally
taking place at recruitement. NCB also hopes that the better results of companies
with an ethnically mixed labour force will become evident.

90
2. Another complaint by employers concerns bureaucracy and costs. Our counter-
argument is that the employer has to draw-up only one report per year and one plan
of action. Furthermore, the aforementioned 60,000-jobs agreement of the employers
and unions is still in force. In order to implement this agreement an employer also
has to register the immigrants working in his company, which entails the very same
costs.

Many special services have been made available to the employers, such as the
above-mentioned manuals, or the management consultants on minorities. Good use
of these facilities will greatly limit the costs and bureaucratic fuss.

3. Sometimes it is voiced that the WBEAA will not induce an increase in employment
of immigrants. NCB doubts that. The voluntary 60,000-jobs plan was not successful
in creating immigrant job opportunities.

The advantage of this law is that it confers responsibility to the individual employer. It
should also be noted that some employers were already conducting a successful
positive-action policy before the WBEAA entered into force. There certainly are
employers who are willing to deal with the problem of socio-economic deprivation.

4. The main fear of employers concerns the possibility of public criticism for the lack
of compliance with the law's objective. NCB prefers cooperation to opposition and
rather to draw up a list of top ten best perfroming companies, than to sanction a
company that does not make an effort to implement equal participation.

5. The Canadian Employment Equity Act compared to the WBEAA

In spite of all efforts, the implementation of the WBEAA cannot be called a success.
Why does the Canadian Employment Equity Act work and the WBEAA not? In
Canada, registration takes place on a larger scale. First of all, it seems that
Canadian employers are more aware of the advantages of equal participation of
immigrants. In Canada, the instrument of "contract compliance" is used in
combination with compulsory registration. There is a very solid institutional structure
in Canada to deal with equal participation, positive action, and, in particular, the

91
Employment Equity Act. The Canadian Ministry of Employment and Immigration and
the Canadian Human Rights Commission play a vital and active role in this field.
Such structures does not yet exist in the Netherlands.

NCB therefore proposes to establish centres per branch of business that:

- monitor the implementation of the WBEAA;

- provide information on the law;
- collect the annual reports and convert them to central annual
reports;
- initiate closer examination when the annual reports give rise to
that;
- initiate discussions with companies where necessary;
- give advice on the approval of the law;
- and, bring successes to the public attention.

6. Conclusion

The government, parliament, unions, and employers agree on the purpose of the
WBEAA: the equal participation of immigrants.

Although registration, information gathering and long-term planning belong to the

day-to-day work of many companies, there is reluctance to establishing a rather
uncomplicated human resources administrative system and reporting certain simple
facts to the Chamber of Commerce.

NCB thinks it is the duty of the governmental institutions responsible to guarantee

the orderly implementation of WBEAA. Legal measures should be taken against
reluctant employers and action is needed to improve the understanding and the
implementation of the law.

The Ministry of Social Affairs, though, is not willing to take any action before the
evaluation of the law in 1996.

92
Currently no legal action is taken against employers who do not register. The
ministry holds that this should not be a governmental task but rather a minority
organisations' initiative.

The current political climate in the Netherlands is not very immigrant-friendly, and it
is difficult for NCB to draw public attention to possible improvements in the
implementation of the law and to lift it to a higher rank on the political agenda. Equal
participation is a vital element in view of the integration of immigrants in our society.
NCB therefore keeps working for an effective law of equal participation of immigrants
in the labour market.

*****

Sources:

- A. Kuijer and M.N. Stulemeijer, Dossier, Evenredige Arbeidsdeelname van

allochtonen, 1995;
- Introduction by I. Akel, director Dutch Centre for Immigrants, seminar
WBEAA, Rotterdam, December 2, 1994;
- Marc van der Loo, De Wet bevordering evenredige arbeidsdeelname
allochtonen, thesis, June 26, 1995;
- Registratiekamer, Registratie op etniciteit en privacywaarborgen, Verslag
van een oriënterend onderzoek na een jaar WBEAA, July 18, 1995;
- Several newspapers and magazines: Staatscourant, Algemeen Dagblad,
Binnenlands Bestuur, NRC-Handelsblad, Communicatie, Babylon
Achtergrond, Volkskrant, Parool, Utrechts Nieuwsblad.

Appropriate tools for measuring assimilation/integration: the

French experience
Michèle Tribalat

93
During the 1980s France became increasingly exasperated with questions
concerning immigration, its discomfiture exacerbated by the inability of its statistical
machinery to correctly measure the phenomenon of migration. Persistent
misunderstanding of the term "foreign population" made it clear that the notions in
use required an urgent overhaul. It was becoming necessary to answer the question
of "How many" with something other than ready-made replies. The INED5 did not try
to stick as closely as possible to the general meaning of "foreign population", which
would have meant little and been quite impossible, but developed a methodological
line of thought indicating two directions in which more accurate measurement might
be pursued. In view of the inadequate data on the living conditions of immigrant
populations and their descendants, it has also worked on developing a specific data
gathering tool. We are placing the emphasis on methods and sources here, as
results are strongly dependent on the relevance of the latter.

Problems of categories

Categories better suited to methodological requirements

Perceptions of the communities which form in the wake of immigration are

conditioned by highly subjective factors, in which physical appearance plays a major
role and tends to magnify the size of the most visible communities; phenomena of
concentration have a similar effect. The actual nationality of individuals is rarely a
consideration.

By centering on the scope of research and therefore the validity of certain categories,
the methodological approach renders nationality irrelevant. Obviously, when it is a
question of studying the process of assimilation of immigrant communities, it is not
sufficient to survey foreigners while omitting those who have acquired French
nationality.

Nor can the use of ethnic categories employed in Great Britain or America help to
divide up the field of observation because it suffers a priori from the same defects as
breakdown by nationality: claims of ethnic identity vary in time and between

5 "Institut National d'Etudes Démographiques"

94
generations. By observing only those who identify with an ethnic group, we risk
introducing a huge bias. Moreover, in the context of France, a statistical variable
based on self-identification with an ethnic group is not likely to work. So this kind of
category can be regarded only as a secondary variable, in the same way as
nationality.

A fresh look at existing statistics

Nothing in the information gathered on census forms justifies the priority generally
attached to nationality. The techniques usually applied to census data are not
therefore engendered by the tool itself. This most valuable source of statistics can
be used in other ways. At the INED, we have used the last three censuses (1975,
1982 and 1990) to highlight not the category of foreigners but that of immigrants,
obtained by adding the number of foreigners born abroad and naturalised foreigners
born abroad (thus combining country of birth and present or former nationality). We
call this category the "direct migratory gain". Furthermore, the census provides other
elements with which to evaluate the population gained through immigration and
correct the picture obtained by analysing the foreign population alone: such elements
are the number of people living in households headed by an immigrant, which in
1990 amounted to just over 6 million individuals (compared with 3.4 million
foreigners or 4 million immigrants). While this notion provides an incomplete view of
communities of foreign origin as a whole, it is more useful when considering children,
and specifically those who still subject to potential authority. It seems particularly
well suited to providing information on "ethnic colour" in given localities; when
applied at département level, it reveals proportions of young people living in
households headed by immigrants which are considerably higher than the recorded
proportion of foreign children: 38.2% for example in Seine Saint-Denis compared
with 19%. We could achieve an appreciable gain in accuracy by extending this
notion to the persons living in a household in which the spouse of the head of
household is an immigrant.

Alongside these categories which may be used more or less immediately, we have
developed notions of "indirect demographic gain" and "population of foreign origin",
the evaluation of which requires a clear view of the migratory process (through which
mechanisms does the population gained via foreign immigration become integrated
by nationality into the French population?) and lengthy work to reconstitute and

95
estimate time sequences. The sum total of a hundred years of immigration, taken at
1 January 1986, gives us the following figures: the indirect gain is slightly over 6
million, a figure which is pushed up to 10 million by adding the direct gain; the
number of people of foreign origin born in France (with at least one parent or
grandparent who immigrated in the previous hundred years) is also put at 10 million.

Creating the right tools

Although we can provide more accurate answers to the question "How many?" using
census data, we cannot do much more on that basis. At best, this information
enables us to produce a snapshot of the situation of immigrants in France; it does
not offer a dynamic view of their behaviour. While the next census will definitely be
improved by the probable inclusion of the year of entry into France, it will not show
how the situation developed over time. It will nevertheless highlight the structural
effects linked to length of stay and above all to the age at entry into France and
enable us to generate more accurate immigrant samples for a post-census survey.
In more general terms, the national surveys carried out by the INSEE do not
incorporate the variables which are indispensable to migration analysis. It was
therefore necessary to create a specific observation tool free of any links with the
criterion of nationality.

The INED has just carried out, with the help of the Institut national de la Statistique
et des études economiques (INSEE), a major national survey on geographic mobility
and social integration of immigrants and their children, providing an insight into how
their living conditions have developed and, more generally, the process of
assimilation. This was an entirely new experiment which, for its implementation in
1992-93, required outside funding (Office des migrations internationales - OMI,
Fonds d'Actions Sociale FAS, Direction de la population et des migration, DPM).

It differs from conventional sources in several respects. It focuses on immigrants, ie

persons having foreign nationality upon their arrival in France (some 9,000
respondents) and the next generation, consisting of the persons of foreign origin born
in France (some 2,000 respondents). For the purposes of comparison, a
representative sample of the French population was also included (some 2,000). For
financial reasons, we had to select a number of groups representing about 60% of

96
the immigrant population aged 20-59, which was the range of ages covered (from
Algeria, Spain, Portugal, Morocco, Turkey, Black Africa and South-East Asia). For
young people born in France, the same line of thought and sampling method
constraints led us to limit the age range to 20-29 and parents countries of birth to
Algeria, Spain and Portugal. For this, we had to use separate samples of immigrants
and young people born in France, which made data gathering more difficult and
costly. To make the tool comprehensive, we added a third "control" sample,
representative of the French population aged 20-59. It consisted of individuals
surveyed at random within a sample of households, which was a satisfying and less
costly solution.

The survey also differed in its polling method (random and not quota-based).

Finally, it did not seek to give a picture of the present situation which would soon be
out of date, but a dynamic analysis covering the necessary span of time.

A retrospective survey

The survey is retrospective and we were very careful to obtain the year of entry into
France.

In opting for a retrospective survey, we were obliged to place the emphasis on

factual information. The questionnaire centres on historical data, concerning
migration, family, work and residence. This information was supplemented by data
on nationality and ways in which it was changed, fertility and contraception, mother
tongue and language used, social behaviour etc.

The accuracy of information as regards date of entry is vital, as it marks a starting-

point or a break with the past. It can be used to deduce age upon entry, which is
equally significant. In this way, for one given phenomenon, three "generation
references" can be defined: adult migrants (entry after 15 years of age), child
migrants (entry before 16 years of age) and persons born in France to immigrant
parents.

On the basis of tests, we opted for a line of repeat questioning which indicated the

97
first time when the migrant arrived in France for a stay of more than one year.

Problems of conventional analyses of migration

The methodological considerations underlying the definition of the field of

observation also apply to the analysis of results and must not be discarded at this
stage. We must call conventional measurement techniques into question and go
back to the basics of demographic analysis. If the moment at which subjects entered
our field of observation is important, it is because that moment indicates the starting
point of the timescale to be taken into account. But the dynamic approach is both
the most instructive and the least misleading. For the study of assimilation of
immigrant populations, it is indispensable.

The necessity of a dynamic approach

A cautious attitude should be adopted towards definitive conclusions drawn from

snapshot information and any resulting league tables of countries of origin showing
differing degrees of assimilation or even differing aptitudes or desires between
migratory waves.

For this reason, a dynamic approach must be preferred. Assimilation is a process

occurring over a variable timespan which depends not only on the country of origin
but also on the characteristics, which themselves evolve, of migrants within the same
migratory wave. The influence of cultures and traditions will have a variable
influence according to the country of origin and differ from the customs of the host
country to varying degrees. Social origin and level of education are influential in
assimilation in France. There is every good reason, therefore, to make time a key
factor in measuring assimilation.

Two notions of time may be used to study the assimilation process: the length of
stay (possibly in combination with another time element) and the succession of
generations.

However, length of stay as a time reference for the assimilation process soon reveals

98
its limits. Most of the events studied occur only once and it is very difficult to
demonstrate the effect of the length of stay on the manner in which French
nationality was acquired for example, as it is a once-in-a-lifetime event for an
immigrant. Any cross-measurement, for the same country of origin, will reflect the
effects of a given period and more generally of structure more than the effect of time
itself. It can even lead to complete misinterpretations if we are not careful. As far as
events occurring more than once are concerned, it is often necessary to take account
of their order, which takes us back to the previous situation. Furthermore, if we take
the example of fertility, reference on to a dual timescale (year of birth and year of
entry) is very often sufficient to exhaust a sample.

In practice the succession of generations is of more use in analysing the process of

assimilation.

Initial hypothesis of "a" strongly heterogeneous immigrant population

Heterogeneity is a feature of any population and is not in itself an obstacle to

analysis. It is the modification over time of the factors making up that heterogeneity
which introduces biases. The study of behaviour in relation to the length of stay of
the entire immigrant population is meaningful as long as its heterogeneous elements
carry a weight equal to for each length of stay. This is rarely the case in cross-
analyses, since waves of migration, which we must initially assume to be
heterogeneous, have followed one another over time: the greatest length of stay will
be found among migrants from Spain and Algeria for example, while those from
Black Africa will have stayed the shortest time. Longitudinal analysis reduces these
structural effects since taking account of a cohort of fixed entry date, at least within a
retrospective study, fixes the composition of the population under observation.
Structural effects have an influence only in the comparison of different cohorts with
equal lengths of stay: taken at ten years' length of stay, the migrants who arrived in
the 1950s are not the same as those who arrived in the 1970s; again, we are back to
previous structural effects. Moreover, the point at which the period of observation
ceases introduces an additional factor of heterogeneity.

Some factors of heterogeneity between migratory waves are of a purely mechanical

nature, the age upon arrived at entry into France being one example. Depending on
the period of activity of a migratory wave, the resulting population structure by age

99
upon entry will vary: in 1992, in the 20-59 age-group, the proportion of migrants
under 16 years of age stood at over 50% for persons born in Spain, whereas it was
approximately 20% for those born in Morocco. But there is a difference between
entry at 5 years, for example, and at 35. Our survey analysis indicated major
differences between child and adult migrants. Schooling is one area where the
contrasts are most apparent: a high proportion of adult migrants from Algeria (41% of
men) have had no schooling for example, while nearly all migrant children (3%) have
had school education, at least part of it in France.

Other factors of heterogeneity are influential within the same migratory wave, often
period-related and reflecting not only changes within the respective country of origin
but also and above all changes in migration control.

Problems of analyses considering the migratory phenomenon to be of secondary

importance

The tendency of French statisticians to consider the characteristics "immigrant", or

"of foreign origin" but mostly just "foreigner", as secondary variables obviously stems
from the system of statistics used in France and the limitations of targeted
quantitative surveys of these groups. This has generally led people to believe that
models explaining the behaviour of France's population as a whole were universal
and perfectly applicable to immigrants or populations of foreign-origin. In this
context, virtually all research has focused on the secondary question of whether,
once the effect of variables thought to determine the phenomenon studied has been
cancelled out, there are differences between foreigners and the native French
population and what those differences are. This assumes that the model illustrating
the overall situation in France can be applied either to all immigrants (or foreigners)
or to each of the constituent nationalities. It is true that the number of foreigners
covered by national surveys is generally insufficient to permit independent analyses
by nationality. Our geographic mobility and social integration survey provides an
opportunity to test such models with regard to precise origin which, instead of serving
as a variable within the model serves to define its scope. It enables us therefore to
compare models illustrating certain phenomena according to national origins
(whether just one is involved or not) and, in some cases, to compare them also with
those applied to the native French population. This approach takes us to the logical

100
limits of "all things being equal in all other respects", with one category embracing a
variable sociological situation: this is the case with analysis by social group, applied
systematically to even out all other differences. It also enables us to gauge new
situations which have no parallel among the native French population and cannot be
dealt with by conventional means.

International comparisons

The need to undertake comparative analyses in Europe is a further reason for taking
the utmost care in preparing adequate data. Publications describing the situation in
Europe, either the regular bulletins of Eurostat or those of the OECD, deal too
exclusively with nationality. We cannot content ourselves with simply adding up
foreign populations. Although some analysts have become aware of the problem,
they still skirt round the issue too often, with introductory precautions which have no
effect on the manner of dealing with the phenomenon of migration. Discrepancies in
legislation on nationality engender substantial errors of evaluation.

Nationality codes differ greatly in Europe in terms of both their content and the
rigidity with which they are applied. The tendency in many cases is to introduce one
form or another of jus soli. Ten or so countries have already done so, with:

- strict jus solis in Ireland;

- jus soli with a condition of residence in Great Britain;

- systems whereby children born in Belgium, Italy, France, Luxembourg, the

Netherlands or Portugal may opt for, declare or manifest a desire to
acquire the nationality of the country concerned;

- dual strict jus solis in Austria, Spain and France, with right of residence for
Belgium and the Netherlands.

Other countries have a policy of offering strong incentives for naturalisation, which
encourages a great many immigrants to take the nationality of the host country; this
is generally the case with the Scandinavian countries.

101
Germany displays both great concern and considerable backwardness in this area.

By taking the case of France and Germany at the beginning of the 1990s, we can
see how biases are introduced by using the foreign population to reflect the
proportion of foreign immigration: although there are twice as many Italians in
Germany as in France, an equivalent number may be obtained for France by
reincluding the Italian migrants who have become French: 253,000 Italians and
523,000 Italian immigrants (indeed, close to 700,000 if we include all the people
where the head of household comes from Italy). Similarly, the number of Spanish
immigrants in France (413,000 as compared with just 216,000 Spaniards) is further
removed than it seems from the number resident in Germany (135,000).

Moreover, in Sweden, the number of persons born in other countries is double that of
foreigners: 430,000 foreigners compared with 869,000 immigrants, or even 1.2
million if we add the second generation, accounting for 13% of the population as
compared with the 4% represented by foreigners. In the Netherlands, over half of
those born abroad have Dutch nationality and, in the country's big cities (Amsterdam,
Rotterdam, the Hague and Utrecht), over half the young people aged under 16
belong to what is known as the second generation.

"City contracts"
Marie Poinsot

1. Pilot sites for integration as part of the integration section of "city

contracts": quantitative and qualitative evaluation of integration schemes
for communities of foreign origin.

2. How can integration schemes in France be evaluated?

3. Evaluation of public and private initiatives: producing knowledge and

producing change

102
Since 1977, the Agency for the Development of Intercultural Relations ("Agence pour
le Développement des Relations Interculturelles" - ADRI) has existed to "foster, by all
appropriate means, the integration of immigrant communities in France and the
harmonious development of relations between French and foreign communities".

It does this by studying and evaluating integration policies, and state and local
integration initiatives. It disseminates its findings in publications and also via
seminars and training programmes for all those concerned with integration:
government and local government officials, elected representatives, social workers,
heads of private associations, etc. In 1995, it made a nationwide study of the way in
which foreigners are dealt with by public services, and of "women go-betweens". It
essentially operates at the point where field initiatives and academic research
connect.

The "Agence pour le Développement des Relations Interculturelles" (ADRI) has

become skilled in collecting data on the ground and publishing details of various
schemes and their results. Its work on the realities of integration in France is backed
by regular news bulletins and thematic bibliographies, which help to illuminate ways
of transferring the lessons learned from local experiments from one place or sector to
another. It wants to set up a co-ordinated network of European partners to evaluate
integration schemes.

The "Agence pour le Développement des Relations Interculturelles" (ADRI) also

organises forums for the discussion of institutional and social practices in the
integration field, city policies, the fight against exclusion, etc. In January 1995, its
local integration workshops gave people working in this field in the Nord-Pas de
Calais region and in other parts of France and Europe an opportunity to meet and
talk about their experiences, practices, prospects and problems. In May 1996, the
"Agence pour le Développement des Relations Interculturelles" (ADRI) will be
holding more of these workshops in Strasbourg, with a greater emphasis on the
European dimension.

1. Pilot sites for integration as part of the integration section of "city

contracts": quantitative and qualitative evaluation of integration
schemes for communities of foreign origin

103
Since the beginning of the 1990s, integration policies have made considerable
progress in France, in spite of the heated debate on immigration, the electoral
successes of the National Front and the upsurge of racism and xenophobia. Local
initiatives for the integration of foreign or immigrant communities form part of a
national integration policy, covering action programmes and contributing to the
setting-up of national, regional and local schemes.

This progress is due to the expertise which the Social Action Fund ("Fonds d'Action
Sociale" (FAS)) has built up since the 1970s (Zecca, 1993), the flexibility of its
activities and the thorough knowledge of local networks which its 10 delegations
have acquired since it was decentralised in 1984. Integration has become a
government priority, and this is reflected in the setting-up of an integration ministry
within the Ministry of Social Affairs. The National Council for Integration, set up in
1990, is another sign of the state's determimation to take positive action in this field.

Since 1990, an increased emphasis on integration policy in some neighbourhoods

has made it possible to implement a full range of projects which foster integration
and also to try new approaches. The pilot committee, which is chaired by Hubert
Prévot, Executive Secretary for Integration, comprises the "Fonds d'Action Sociale"
(FAS), the "Délégation interministérielle à la Ville" (DIV), local and regional
authorities, and associations. Together, these bodies worked out a programme of 60
pilot integration sites, which was evaluated in 1994.

"The success of city policy depends on integrating immigrant communities. Any

action which contributes to this integration should be seen as an important aspect of
negotiating "city contracts" (DPM No. 93/30, November 1993). In 1994, 165 of the
214 city contracts concluded under the government's policy on cities included an
"integration" element. Local and regional authorities (for a long time slow to do
anything about integration) now play a decisive role in this area, depending on their
powers and the policies they pursue. The decentralised state services, which work
for all sections of community without discrimination, also contribute here. The fact
that most cities are working with the state and the "Fonds d'Action Sociale" on
projects to help immigrant communities shows that integration has indeed become a
state priority. In addition to spending more than 110 million FF (Francs Français)
under these contracts, the "Fonds d'Action Sociale" contributes more than 180

104
million FF (Francs Français) to non-contract projects. At present, it supports more
than 4,700 associations which are working to integrate foreign communities. This is
why it is so hard to keep track of all these associations' activities, which are
obviously highly varied.

How can projects aimed at integrating immigrants be singled out within an overall
programme covering the whole population? To identify them in 1995, the city
contracts defined four priority areas of action for 1995:

- Action to de-isolate hostels for migrant workers and integrate residents

within the community

- Basic language training and social integration programmes for immigrant

women
- Treatment of foreigners by the public services;
- Socio-vocational integration for young people of foreign origin.

105
The integration projects implemented under city contracts are now being evaluated.
A number of national studies and surveys have been carried out and quantitatively
evaluated. They cover:

- the 15 "département" plans for the reception of families joining immigrants,

(DPM, 28 February 1994);
- initiatives aimed at very young children as part of city policies, with a
special emphasis on reception points for families;
- school solidarity networks, launched in 1992 on 77 sites and continued in
1993 on 80 sites, benefiting around 2,000 adolescents (FAS Circular No.
94/233, July 1994);
- special summer schemes to stop the social links between the generations
from deteriorating and reduce the risk of young people's being
marginalised;
- the 15 economic sponsorship networks set up in five French regions in
1993 to help 3,000 young people of foreign origin aged 15-25;
- the treatment of foreigners by the public services;
- "women go-betweens";
- results secured with projects carried out by the FAS-delegations.

2. How can integration schemes in france be evaluated?

"The evaluation of integration schemes requires a clearly defined approach to the

communities studied, based on target groups and specific objectives" (Lebon, 1994,
p. 21). The problem here is that the way in which statistics are kept in France makes
it hard to pinpoint indicators relating to the immigrant origins, or ethnic or religious
affiliations of specific groups - a point discussed by Michèle Tribalat in her report
which, as European readers hardly need reminding, has been hotly discussed in
France in recent months. As a result, evaluation of public integration schemes is
usually limited to partial, single-subject or sectoral analyses of the socio-economic
situation of foreign communities, carried out by agencies which have nothing to do
with planning projects or running them.

This situation is now changing, however, thanks to the priority mapping of integration

106
patterns which the Directorate of Population and Migration (DPM) is masterminding.
Through its regional delegations, the "Fonds d'Action Sociale" (FAS) is trying to get a
really clear, site-by-site picture of foreign communities and communities of immigrant
origin.

Integration observatories, like those in Montpellier and Strasbourg, and resource

centres, like the one in Roubaix, were set up in the late 1980s to help with this task.
They have the job of producing regional - and sometimes very local - data for the
people involved in integration projects, to help them in their work. They also have the
job of collecting information on integration schemes for immigrant communities and
assessing them. Last but not least, they act as meeting places and training centres
for integration professionals, who need to explore new methods and keep abreast of
new trends.

Another thing which makes it hard to obtain a coherent set of indicators for
evaluation of integration schemes is the multi-dimensional nature of the integration
process, in which foreign communities adjust to French society in successive stages.
The fact that migration and family patterns vary, and that the individuals and groups
who make up these foreign communities have different social and cultural capacities,
makes this process a complex one, which changes with the passing of time.

As a result, any attempt at evaluation faces the problem of coming up with a usable
definition of integration - based, for example, on the concept suggested by the
National Council for Integration. The first step, in other words, is to go some way
towards overcoming the disagreement on what integration actually means. Two
conceptions of integration are actually current in France, and public policy oscillates
between them (Lebon, 1994, p. 25).

The first conception is close to assimilation. It wants the communities to mix and
mingle, and sets out to smooth away distinctively foreign cultural characteristics by
easy stages. The idea is that foreign communities, by living in proximity to French
ones, will gradually take on their cultural values and social behaviour patterns. The
second is closer to the notion of "insertion": it tries to ward off the social problems
caused by the presence of foreign communities, while preserving those communities'
cultural characteristics. Public action here aims more at maintaining social cohesion
than at assimilation (Lebon, 1194, p. 25).

107
Depending on the state of French society, and particularly the extent to which public
discussion of immigration has been radicalised, integration policy focuses on one of
these two conceptions, and sets itself aims which are later taken into account when
the local or national impact of integration schemes on foreign communities and their
environment is being measured. These variations in the yardsticks applied make it
harder to gauge the effectiveness of public projects - and integration needs,
moreover, to be assessed in comparative terms, with reference to the performance of
other social groups.

Sayad (Hommes et Migrations, 1994) argues that integration cannot be measured,

and must be related to the structural changes taking place in French society to yield
a picture of that society's capacity to absorb and integrate exogenous communities.
At a time of economic crisis, when pockets of social exclusion are developing,
integration - long a question which arose solely in connection with communities of
foreign origin - has become a matter of concern to the whole of French society,
whose social cohesion and national identity are threatened by the growing number of
marginalised social groups.

In these conditions, the groups originally targeted by integration policy lose their
separate identity and become just one of those covered by the "integration" aspect of
city policies and the fight against exclusion. This has the healthy effect of removing
the stigma which results from immigrant communities' being the only target, but there
is a long-term danger that it may make people forget that the integration of immigrant
communities raises very specific social, economic and cultural issues which cannot
be boiled down to those raised by exclusion in general. To take only one example,
public action must continue to pay special attention to the problems of discrimination
which foreign communities face in the employment and housing fields - otherwise,
social relations will be ethnicised, and France will get full-scale ethnic ghettos of the
kind which exist in the United States.

Integration consequently raises questions concerning the extent to which action by

the state, local and regional authorities, social workers and private associations can
really support and facilitate the process of adjustment to French society for
communities of foreign origin. This is why evaluation of integration schemes in

108
France - which are of immense political importance for those who run them - can
easily be used to legitimise the things done, the lessons learned and the results
obtained in terms of their declared aims at starting.

Given France's highly centralised tradition, some evaluations suggest that more
emphasis should be placed on regional and local operatives, since it is very hard to
put the state in touch with the grass-roots and give it a true picture of local realities.
The recipients of subsidies manipulate the evaluation process in an effort to hang on
to their budgets, or, when their contracted aims have been accomplished, to renew
their activity. This results in a tendency to keep coming back to the handicaps which
can be used to justify the action taken for these communities, instead of planning a
new strategy.

The state itself is not proof against this temptation to focus on pilot projects which
are quickly counted and taken as a sign that it is doing something, instead of thinking
about ways of making the most of integration schemes and reproducing or
transferring them.

3. Evaluation of public and private initiatives: producing knowledge and

producing change

There are two approaches to the evaluation of integration schemes. They are
complementary, but rarely combined.

- An approach which looks at the integration problems of immigrant

communities in terms of sectors - housing, health, training, etc. - or groups
- infants, women, old people. This makes it possible to analyse integration
schemes in detail and quantitatively, and to follow the various stages in
their development.

- A transverse approach, focusing on local, regional and national intervention

systems, and the ways in which responsibilities are divided between
government and state, local and regional authorities, specialised services
and associations, in accordance with guidelines set out in the partnership
agreements signed between them. The 60 FAS pilot integration sites were

109
 evaluated using this approach, which is more stimulating when it includes
close analysis of the successive stages: fact-finding visits prior to selection
of the sites, the ways in which objectives are worked out for each, the
setting-up and functioning of teams, strategies, the development and
typology of activities (experimental, structural, "flagships", knock-on effects
(FAS p.14) and any impact on private initiatives

But this approach, emphasising areas of action rather than projects, encourages any
tendency which operatives may have to see the things they do and the procedures
they follow in very general terms. Combined with sectoral evaluation, it can provide
detailed information on the practicalities and specifics of local initiatives. Desirable
as it is, however, this combination of the two approaches is not easily secured in
practice.

So far, we have been considering evaluation in terms of its ability to generate

knowledge and tell us something about the quality of integration projects. But
effective evaluation procedures must also help those in charge of these projects to
think usefully about their future development, and formulate way-stage and progress
indicators for the process they are piloting. To be genuinely useful, evaluation must
therefore tell them more about the way in which this process should develop over
time. This makes it possible to readjust the criteria for public intervention, and makes
it easier to see what it can do, in terms of future action, matching future funding to
the needs identified, local impact, etc.

In this way, evaluation generates change in public action, becoming a source of

innovation and a means of experiment. Fostering a "constant check" approach to
public action, it gives the various protagonists of integration an instrument of self-
training and self-qualification which allows them to see how their work should
connect, both with that of other partners and with the community they are working
for.

How can we get beyond sweeping, simplistic conceptions of foreign communities?

The whole purpose of integration projects is to eliminate the stigmatising effects of
immigration. But their specific natures makes them, too, a source of stigma. This is
why politicians are slow to come out clearly in favour of integration, which can cost

110
them dear at election time. Social workers also feel that their status is diminished by
working with foreign communities.

In general, projects aimed at foreign communities have had positive effects on all
sites. This emerges from the evaluation of the integration aspect of "city contracts"
carried out in 1995. They stimulate local partnerships between institutions which are
not in the habit of working together, strengthen private associations, and help to
improve the area all round. "It was feared that developing specific programmes for
immigrants might stigmatise them and isolate them from other local residents, but
this has not happened" (FAS, 1995, p. 29).

How can politicians be brought into integration projects? The aim of public action
must be to foster awareness, explain the issues, and train local teams to cope with
the problems and challenges of integration. In the long run, this work of clarification
and explanation will undoubtedly help to change public attitudes on immigration.

*****

Sources:

- Agence pour le Développement des Relations Interculturelles (ADRI),

Action culturelle et intégration, Bilan des connaissances, May 1994
- Agence pour le Développement des Relations Interculturelles (ADRI),
Chahla Beski, Les femmes relais, March 1994
- Agence pour le Développement des Relations Interculturelles (ADRI),
Martine Fournier, L'accueil des étrangers dans les services publics, March
1994
- Danielle Boyzon Fradet, ENS Fontenay, CREDIF, Un dispositif d'accueil
original pour les enfants rejoignants de la ZEP de Montpellier, rapport
d'évaluation et de perspectives, Jan. 1994
- Direction de la population et des migrations (DPM), Les pratiques et
méthodes des réseaux solidarité Ecole à travers quelques actions
remarquables, 1995
- Espinasse M.-T., Contrats de ville et intégration des immigrés, bilan 1994

111
 de l'application de l'accord cadre Ministère des Affaires Sociales de la
Santé et de la Ville, Dec. 1994
- Fonds d'Action Sociale (FAS), Petite enfance et politique de la ville,
Mulhouse 14-15 May 1993, Ed. Syros, 1993
- Fonds d'Action Sociale (FAS), Bilan et perspectives du programme 60 sites
pilotes pour l'intégration, 1990-1993, July 1994
- Fonds d'Action Sociale (FAS), Sylvie Brun, Dispositif d'évaluation des
actions santé, bilan quantitatif 1992-1993, Oct. 1994
- Fonds d'Action Sociale (FAS), Contrats de ville: bilan d'étape, Dec. 1994
- Lebon, A., DPM, Situation de l'immigration et présence étrangère en
France, 1993-1994. Doc. Française, 1994
- Lebon, A., Les indicateurs d'intégration: où en est la France? Feb. - March
1994, University of Montreal, pp. 21-43
- Zecca Marine, Politiques publiques et capitalisation des savoir-faire de
l'intégration, Secrétariat Général à l'intégration, March 1993.

112
 The development of an ethnic group question for the British
Census
Andy Teague
Office of Population Censuses and Surveys (OPCS)

Abstract

The 1991 Census was the first in Britain to include a question on ethnic group. This
followed a period of some sixteen years in which a number of possible questions
were tested. It was important in this process to both gain acceptability of asking a
question on ethnic group from various ethnic minority organisations and the public as
a whole, and to find a question that would provide sufficiently accurate results. This
paper provides a summary of the processes through which a question on ethnic
group was developed for inclusion in the British Census of Population. Fuller details
are provided in other papers to which reference is made. The fullest account is given
by White and Sillitoe6, on which much of this paper is based.

Introduction and background to the need for measurement of ethnic group

The Census in Britain is taken with the authority of the 1920 Census Act. Although
the Act specifies that 'Nationality, birthplace, race, [and] language' may be part of the
Census, it was not until the 1991 Census that a question on ethnic group was
included.

Throughout its history Britain has received immigrants from elsewhere. In all
Censuses since 1841, people have been asked to provide information on country of
birth and sometimes nationality as well. This information proved to be sufficient in
the compilation of statistics on the immigrant population until the large scale influx of
people from the West Indies, East Africa and South Asia in the 1950s. Unlike their
predecessors, these people were predominantly distinguishable from the indigenous

6 Sillitoe K. and White P., Ethnic group and the British Census:
the search for a question. Journal of the Royal Statistical Society (Series
A), Vol 155, No. 2 (1992).

113
population (and former immigrants) by their skin colour. Sample surveys began to
show that people of Black and Asian origin were possible suffering from
discriminatory practices; less well paid jobs, higher unemployment, poorer housing
conditions, and so on. For these reasons, and to help monitor the effectiveness of
equal opportunities policies, the demand for information increased.

As more and more children of the original Black and Asian immigrants were born in
Britain, information on country of birth became less useful. The 1971 Census
included a question on parents' country of birth. However, this proved to be
unsatisfactory as well. There were a sizeable number of white people whose parents
were born in the Indian Sub-Continent. Also most people born, or had parents who
were born in East Africa were in fact of Asian descent. Further, during the 1970s the
numbers of Black and Asian children with parents born in Britain would increase. All
of this led to the need to search for a question on ethnic group or race as opposed to
country of birth or nationality.

Development of an ethnic group classification in britain

In many countries that have populations of differing origins and diverse cultures, it
has been the practice over recent decades, in particular, for questions on race or
ethnic group to be asked in the Census. The types of questions asked depend on a
mix of historical, social and political factors. In some countries such as Canada,
ancestry plays an important part. In others (such as India) religion and language, or
caste/tribe are more influential. Elsewhere (for example, the USA and the
Caribbean), a variety of criteria have been used such as skin colour, national origin,
language and culture.

The message therefore is that there is no universally acceptable criteria on which to

base an ethnic group classification. This was recognised by the United Nations in
19777 and 19878. Pursuant to this, the development of a classification in Britain has

7 United Nations Economic and Social Council. Principles and

Recommendations for the 1970 Population Censuses, as amended by
Recommendations for 1980 Censuses of Population in the ECE Region.
United Nations (1977).

114
been based around one which would meet national needs.

Sillitoe and White describe some of the problems in determining definitions of

ethnicity and race. They conclude that it is difficult to distinguish the two in trying to
meet the objectives of a question that will be acceptable to the majority and provide
information in a useable form. In general, ethnic group is now preferred as a term
rather than race. Ethnic group is defined by Sillitoe and White as "a socially distinct
community of people who share a common history and culture and often language
and religion as well". In identifying socially distinct groups a contributing factor is
skin colour which introduces a "racial" element to the classification. As will be found,
this race element persisted throughout the development phase into the final question
for the 1991 Census.

Trials of questions on ethnic group in the period 1975-1979

In 1975, OPCS9 began a series of trials to see if a census question could be devised
which would be both acceptable to the public and the answers to which would
provide more reliable information on race and ethnicity. In general, the trials
simulated the census itself: an interviewer delivered a trial census form to each
household in a sample and then called back a few days later to collect and check the
form. Participation in the trials was voluntary and, apart from 1979 and 1989 when
they were carried out as part of the final census tests, were not subject to the
attendant publicity which accompanies a full census.

Four alternative designs were carried out in three separate trials between 1975 and
1977. Details of these are given by Sillitoe10. Sillitoe recommended the following

8 United Nations Statistical Commission and Economic

Commission for Europe, Conference of European Statisticians, Economic
Commission for Europe Committee on Housing, Building and Planning,
Statistical Standards and Studies - No. 40. Recommendations for the
1990 Censuses of Population and Housing in the ECE Region. United
Nations (1987).
9 Office of Population Censuses and Surveys
10 Sillitoe K., Ethnic Origins 1, 2 and 3. OPCS Occasional Papers 8,

115
"tick box" categories following the 1975-77 trials:
White
West Indian
African
Indian
Pakistani
Bangladeshi
Arab
Turkish
Chinese
Any other race or ethnic group (with a request to write in the
description)

The main difficulty with the recommended design was in relation to the classification
of persons of West Indian descent. Many West Indians, unlike the Asian groups,
were uncomfortable with describing themselves as such particularly if they were
British born. It was suggested that introducing a Black category(ies), would get
round this problem particularly if "Black-British" was given as an option as well.
However the term "Black" (and for that matter "white") was deemed unacceptable by
the Government at the time, and further trials were requested avoiding such "racial"
descriptions. None of the three designs in the next trial in 1978 was deemed
useable (see Sillitoe11), and so a further alternative was tried in the 1979 Census
Test (see below) alongside a question in parents' birthplace. Forms with the country
of birth of parents' question were delivered to alternate households; the other 50 per
cent of households received a form with the ethnic question.

Categories in the ethnic group question in the 1979 Census Test

English, Welsh, Scottish or Irish

Other European

9 and 10 (1978).
11 Sillitoe K., Ethnic Origins 4. OPCS Occasional Paper 24 (1981).

116
West Indian or Guyanese
African
Indian
Pakistani
Bangladeshi
Arab
Chinese
Any other racial or ethnic group

117
Partly due to a campaign organised by some local organisations in the test area of
parts of the London borough of Haringey, co-operation from the public was seriously
affected. The campaign advised people to not answer any questions about ethnicity,
birthplace or related questions on the grounds that the collection of such information
was linked with proposed changes to the nationality laws. The response rate was
only 54 per cent compared with around 70 per cent in previous Census Tests. As
many as 32 per cent of West Indians and Asians objected in principle to the inclusion
of a question on ethnic group.

In considering these findings and also through consultation with ethnic minority
organisations, it was decided that including a question on ethnic group or parents'
country of birth in the full 1981 Census would be too risky.

Trials of questions on ethnic group in the 1980s

Following the 1981 Census which, while successful in itself, was deficient in not
collecting information on ethnic group, the House of Commons Sub-Committee on
Race Relations and Immigration reviewed the need for ethnic group information. In
1983, they reported12 with a recommendation that OPCS be asked to carry out
further trials. The Government in its response13 accepted these recommendations in
principle and invited OPCS and the General Register Office for Scotland (GRO(S)) to
proceed with the aim of funding a suitable ethnic group question for inclusion in the
1991 Census.

"Black" and "white" were now deemed terms that were inevitable and therefore the
starting point was in effect the stage reached in 1977 (see earlier). A design very
similar to this had been used successfully in the Labour Force Survey since 1979.

12House of Commons. Ethnic and Racial Questions in the Census.

Second Report from the Home Affairs Committee Session 1982-83,
Volume 1, HC33-1, London (HMSO) (1983).
13 Her majesty's Government. The Government Reply to the Second
Report from the Home Affairs Committee Session 1982-83 HC 33-1.
Command 9238, London (HMSO) (1984).

118
Trials took place in 1985 and 1986 and are described in detail by Sillitoe14. In the
first trial in 1985, the category "British Asian" was not acceptable. The introduction
of a religion sub-section for South Asian groups worked well but did not provide any
extra information to qualify ethnicity.

The trials carried out in 1986 tested a number of different designs none of which
were wholly satisfactory. The best of these proved to be the following (tested in
October 1986):

White
Black, West Indian or African
Indian
Pakistani
Bangladeshi
East African Asian
Chinese
Arab
Any other ethnic or social group (with space to provide a written
answer)

The most noticeable aspect of this question was the exclusion of the term "British".
This had proved unworkable in all the trials in which it had been tested.

During 1985-86 and in the subsequent two years ethnic minority organisations and,
in particular, the Commission for Racial Equality were consulted. This proved useful
in gaining support for the question and in ascertaining likely acceptability of certain
categories to the public.

April 1989 census test

Following the extensive consultation period during 1987 and early 1988, the

14 Sillitoe K., Developing Questions on Ethnicity and Related Topics

for the Census. OPCS Occasional Paper 36 (1987).

119
Government published its proposals for a question on ethnic group in a White
Paper15. The question proposed had the following categories:

Recommended categories for 1989 Census Test

White
Black
Indian
Pakistani
Bangladeshi
Chinese
Any other ethnic group (with space to provide a written answer)
The White Paper stated that the Government proposed to include a question on
ethnic group in the forthcoming (April 1989) Census Test and to study the results
before deciding whether such a question could be included in the full census in 1991.
Further comments were also invited from the public and organisations representing
ethnic groups. As a result of these consultations, the black group was split into
three. The following categories were those used in the 1989 Test (and ultimately in
the 1991 Census).

White
Black-Caribbean
Black-African
Black-other (with space to provide a written answer)
Indian
Pakistani
Bangladeshi
Chinese
Any other ethnic group (with space to provide a written answer)

Her Majesty's Government.

15 1991 Census of Population.
Command 430, London (HMSO) (1988).

120
Response rates to the 1989 Census Test were in line with those expected based on
previous voluntary Census Tests. Although these varied from area to area, and for
different ethnic groups, the ethnic group question was rarely (less than one per cent
overall) given as a reason for not taking part. When prompted by interviewers, a
higher proportion (up to 20 per cent of Black informants) objected to the inclusion of
an ethnic group question. However it was demonstrated that these objections would
not prevent them from taking part in the Census Test; the proportion objecting was
about the same for those who took part in the Test as it was for those who didn't.

Further, the quality of answers to the question was sufficiently high. Based on a
comparison of answers given in the test and in a follow-up interview for a sample of
households, 90% of answers for white group were the same, 86% of Black
respondents, 89% of Asian respondents, and 82% of 'others'. (See White16 for further
details of the evaluation of the 1989 Census Test).

The 1991 census and the future

Based on these findings, the Government took the decision to include a question
identical to that used in the 1989 Census Test in the full Census in 1991.

In general, the question worked well. There was no organised objection to the
question and enumerators did not report any real problem with the question with the
public. There were some minor objections to the form of the question -
predominantly from those in the 'white' group that this category was not segregated.
This takes us back to the main aim of the question: to measure where disadvantage
or discriminatory practices are taking place, further that most evidence (although not
exclusively) was that discrimination takes place on the basis of skin colour, towards
groups other than white.

It is recognised that the question is not perfect; no ethnic group question ever will be.
It is impossible to meet all user needs, be acceptable to the public and provide

16 White P., A question on ethnic group for the Census: findings

from the 1989 Census Test Post-Enumeration Survey. Population Trends
59, London (HMSO) (1990).

121
reliable results. Variations to the question are being considered as part of the
planning towards 2001. In doing so, it will be important to consider comparability
with the existing question (which is also being used now in other ethnic monitoring
schemes and surveys) alongside new and changing needs and categories which are
both politically acceptable and understood by the public.

The results of the 1991 Census, which showed that 5.5% of Britain's population
belonged to ethnic minority groups (ethnic groups other than 'white'), are being used
extensively in central and local government, health authorities, academia, and the
private sector. A large number of analyses have already been produced. A series of
four volumes will be published by OPCS later this year or early in 1996. These have
largely been written and edited by outside (OPCS) contributors.

Conclusion and lessons learned

The decision to include an ethnic group question in the 1991 Census was seen as
one of the main successes of the 1991 Census. It took a long time to find a question
which was suitable but the wait is considered to have been worthwhile. A wealth of
statistics on ethnicity is now available, the like of which was not previously available
particularly for the smaller geographical areas (the Labour Force Survey had
previously provided basic estimates at national and broad regional levels).

The key lessons learned in developing the question are as follows:

• Establish the case for information.

• Establish precisely what is required.

• Test different versions of questions with the public. These tests

should involve in-depth interviews (as part of the research into
new and modified questions for the 2001 Census, OPCS has
been using the technique of cognitive research which attempts to
gain a picture of how respondents read and understand
questionnaires).

122
• Consult widely with organisations representing the different
ethnic groups. Establish both formal and informal links with such
organisations and customers of the resulting data.

• Allow plenty of time for testing and consultation.

• Accept that ethnicity is a largely subjective rather than objective

concept. Some members of the public find this difficult to do.

123
 Four national surveys of ethnic minorities
Richard Berthoud

This is the story of 30 years of research carried out by one organisation. That's not to
pretend that many other organisations have not also studied the social and economic
position of ethnic minorities. But the focus on one series of surveys provides a
central theme, and can be used to illustrate some of the changing pre-occupations of
race relations in Britain. The second half of this paper, concentrating on the current
survey - the fourth in the series - illustrates the kinds of information which seem to be
relevant to a multicultural society.

Discrimination in the 1960s

Concern about ethnic minorities in the 1960s focused on the issue of discrimination -
`the colour bar' as it was often called. The first Race Relations Act in 1965 mirrored
American civil rights legislation by forbidding segregated access to public services
such as restaurants. But discrimination in ultimately more important fields such as
employment and housing remained not only legal, but overt: notices specifying `no
coloureds' were openly displayed.

The first survey was directly concerned with the issue of discrimination. A sample of
1,000 members of ethnic minorities in six towns was interviewed about their
perceived experience of discrimination. Interviews were also carried out with
employers, landlords, banks and other organisations in a position to discriminate.
And, crucially, objective tests of the extent of discrimination were devised, so that the
measurements did not depend simply on subjective opinions.

For example, 40 white actors, 40 `coloured' actors and 40 Hungarians were sent to
apply for the same 40 low-level jobs. In ten firms, all three applicants were rejected,
and the vacancy may have been filled. But in the remaining 30, all the white
applicants received some kind of encouraging answer; 17 of the Hungarians, and
only 3 of the coloured immigrants.

124
This research was one of the influences which led to the second Race Relations Act
of 1967, which outlawed direct discrimination in employment, housing and other
fields.

One of the findings of the study was that the victims of discrimination often did not
know for certain, or perhaps even suspect, that it had occurred. Even the
perpetrators of discrimination were not necessarily conscious of it. The objective
tests of discrimination developed for the first survey were extended and repeated a
number of times over the following 20 years; and tended to show that discrimination
persisted, even though it was illegal.

Disadvantage in the 1970s

There has been much less concern in recent years about direct measurements of
discrimination. We know that discrimination exists, though it is difficult to prove in
individual cases. The emphasis has moved on to identifying the areas of social and
economic life where ethnic minorities face disadvantage.

The second study was a much wider-ranging enquiry. It included further

discrimination testing, and studies of the recruitment and allocation policies of
employers and municipal housing authorities. But the centrepiece was a survey of a
random sample of 3,000 members of minority groups, with a comparison sample of
white people. Thus disadvantage could be identified in terms of differences between
groups - in employment, in housing, or in other fields - to the extent that minorities
experienced worse conditions than white people. For example, one fifth of white
households lived in accommodation which did not have the exclusive use of a bath,
running hot water or a WC. The proportions were one third for minority groups,
except among Pakistanis and Bangladeshis, where it was more than half.

According to this way of looking at things, disadvantage is defined as an outcome.

Differences might be explained in terms of the characteristics and behaviours of the
minority group; or the attitudes and behaviour of the majority group; or in the
operation of the market and other processes by which economic resources are
generated or distributed. Detailed investigation of each of the potential processes (of
which direct discrimination is only one) is required if inequality is to be reduced.

125
The research was, again, one of the influences behind the Third Race Relations Act,
which took a much broader approach to the definition and identification of direct and
indirect forms of discrimination.

Dualism in the 1980s

It was nevertheless important to look for an over-arching explanation for ethnic

minorities' experiences. This was found in the concept of `exclusion' - whatever the
details of the situation, non-white people were denied access to social and economic
privileges in ways which paralleled the exclusion of women, the working class and
disabled people. Ethnic minorities were thus defined in relation to the white majority,
and this was consistent with a political solidarity under the banner of the common
name `black'.

The third survey was primarily concerned to compare the experiences of the majority
and minority groups, including a larger sample, this time, of 5,000 members of the
minorities. The study design had much in common with the fourth survey, to be
described in more detail in the next section. It showed that in some respects,
conditions were improving for all groups, black as well as white. For example, the
proportion of households lacking basic amenities had fallen to only 5 per cent of
whites, 5 per cent among West Indians and 7 per cent among Asians (compared with
the much higher figures quoted above for 1974). But in other respects disadvantage
was just as great as it had been ten and twenty years before. Unemployment, for
example, ranged between 13 per cent for white men to 25 per cent for West Indians
and 29 per cent for Pakistanis and Bangladeshis. The gap was actually wider than it
had been in 1974.

Since many members of the minority groups had by now lived in Britain for decades,
or for the whole of their lives, it was clear that the problem was not simply a matter of
adjustment following a period of migration.

Diversity in the 1990s

Much of the discussion over this period had been in terms of `integration' - the word

126
stressed in the title of this seminar. This seemed to imply that everyone should be
treated exactly alike, and in the end there would be no discernible difference
between one ethnic group and another. But a pair of questions in the third survey
helped to cast doubt on that objective. The overwhelming majority of both West
Indian and Asian respondents rejected the idea that they should `keep themselves
apart from white people'. On the other hand, the overwhelming majority said that
they should `try to preserve as much of their own way of life and culture as possible'.
In consequence, the emphasis now is not so much on `integration' as on
`multiculturalism'. And, because groups originating in different countries each have
their own ways of life, as reflected in language, religion, style of dress and so on,
much of the focus is on the diverse needs of different minorities, rather than on the
common experience defined by non-whiteness.

The fourth national survey of ethnic minorities is now in progress. We have

interviewed 5,250 people of Caribbean, South Asian and Chinese origin. This is a
difficult task: a sample of members of the minorities (only six per cent of the
population of Britain), had first to be identified by a technique known as `focused
enumeration'; second, they had to be interviewed by a member of their own ethnic
group, and often in their language of origin. The survey has collected information
about a far wider range of issues than any of the previous studies, including:

Family structures
Education
Employment
Income and standard of living
Physical and mental health
Housing
Area of residence
Relations between groups, including racial harassment
Identity, religion and language

The data are now being analysed, and the results will not be published until next
year.

At the time of the first survey in the 1960s, it was widely thought that these issues

127
were far too sensitive to be studied by social science techniques. Nowadays, as the
previous paper has explained, large quantities of basic data about ethnic minorities
can be obtained from the Census, the Labour Force Survey, and many other
sources. One of the special roles of our own survey is to cover topics which have not
been studied elsewhere - such as health, the experience of racial harassment, or
variations in the perception of ethnic identity. But another is to link together the
information about a variety of issues, to show how far there is, or is not, a common
pattern.

If there is a pattern, it is that in many respects the minorities are at least as different
from each other, as they are from the white majority. Here is a series of examples.

♦ An obvious difference between South Asians and people of Caribbean

origin is that the former have their own languages and religions, distinct
from those of the white community. Most black people in Britain were
brought up to speak English and to practice Christianity, though they have
their own cultural identities.

♦ Some minorities, such as Chinese and Indians, appear to do better than

expected in the competition for places at universities, while others, such as
Caribbeans and Pakistanis, do worse.

♦ Among Caribbeans, marriage rates are low, and a very high proportion of
children are born to single women. South Asians, in contrast, are more
likely to get married, and remain married, than white British people.
Pakistani and Bangladeshi families (but not Indians or African Asians) tend
to have more children than is common nowadays.

♦ Caribbean women are more likely to take paid work than their white
counterparts; but Pakistani and Bangladeshi women have very low rates of
economic activity.

♦ While most minority groups experience more unemployment than white

people, unemployment rates are especially high among young black men,

128
 and among middle-aged Pakistanis and Bangladeshis.

♦ Such data as are so far available indicate that Indian and African Asian
families have incomes at least level with those of white households -
perhaps higher. But Caribbean families are worse off than average. And
Pakistani and Bangladeshi families are at high risk of poverty.

The theme is `diversity'. It needs to be said, perhaps, that diversity can be both a
good and a bad thing. A good thing when we think of the variations in culture and
perspective which the minorities bring to our increasingly cosmopolitan society. A
bad thing when we think of the unacceptable variations in outcome for particular
groups in terms of unemployment, poverty, poor housing and so on. Equality means
equal access to rights and resources. It does not mean everyone being the same.

So there is a real challenge to explain the differences in educational and economic

success between minority groups. These explanations have to take account of
minority cultures, racial exclusion and broader structural changes. Appropriate
policies can be formulated only when the interactions between these three elements
have been understood.

129
 An action-research approach to the understanding of integration
Berit Berg

Introduction

Outlining briefly the theoretical foundation of my understanding of the concept of

integration, I refer to the Swedish sociologist Jose Alberto Diaz who has offered the
following definition of integration:

We define and describe integration in relation to the complex participation of the

immigrant in more-or-less organized and also institutionalized fields of social,
economic, occupational, residential and communicative systems of social action in
the host society... Full integration into the host society, for example, implies a state
of complete similarity between immigrants and native people in their participation in
the socially regulated distribution of valuable resources. (Diaz 1995, p. 202)

This definition is influenced by Bernard (1973) and Esser (1980), and also by
Scandinavian researchers such as Ålund and Schierup (1987, 1991, 1993), who find
that integration is the:

...immigrants' adaptation to the institutions, norms and culture of the 'majority

society' to the extent necessary for the group's members to function in the society,
while at the same time keeping intact its own ethnic identity. (Ålund and Schierup
1991, p. 14)

A central issue in integration studies has been the attempt to identify a set of factors
explaining progressive, regressive, or stable patterns in the process of integration of
immigrants in the host society. In this context it is important to ask: Are the most
important factors individual motivation, the social or economic background or
language abilities? Or are the various conditions existing in the new environment the
most important factors for explaining successful integration?

Diaz answers "yes" to both these questions. We need a multi-dimensional approach

130
if we are to understand the complexity of this issue. He defines several key
dimensions of immigrant integration: structural and political integration at the system
level, social, residential and communicative integration at the interactional level of
the society, and personal integration at the actor level. All these dimensions must be
taken into consideration. His main focus is on the relationship between individual and
contextual factors.

"Integration", both for refugees and labour immigrants, has long been a main
objective of Norwegian immigration policy. The definition of what was meant by
integration in the 1970s and early 1980s was founded on the concepts of
"segregation" and "assimilation". But this has changed over the last decade so that
today the concept of integration in Norwegian immigration policy mainly builds upon
the objectives of equality and participation.

This change reflects the problems that arose when attempting to make the
integration concept operative. When such problems of definition arise, and when
difficulties are encountered in making the concept operative, it follows that there will
also be implementation problems. As they are so general, it is possible to satisfy the
objectives with very different content. What does it mean to work on the basis of an
integration objective? What indicators do we have with which to ascertain that the
objectives have been reached?

Empirical studies based on a multi-methodological design

This has been part of the background of two research projects which have been
carried out at the research institution "SINTEF IFIM" during the period 1991-95
(commissioned by the Norwegian Research Council, and the Programme for
International Migration and Ethnic Relations). It is useful to look at the two projects
together as they complement each other both thematically and methodologically. The
first project has focused on the reception of refugees in the local communities
(Project I). The project, which has been carried out in 28 Norwegian municipalities,
has been based upon an action-research design. The objective of the project has
been to find measures which can contribute to a more expeditious integration of
refugees in the municipalities. The target group consisted of personnel from local
administrations who work with refugees during their first years in Norway. The
second project has focused on the individual refugee's participation and

131
integration in society (Project II). The objective of this project was to study the
relationship between the objectives of immigration policy, the measures used and the
practical results. How has the implementation of new measures functioned? What
effect have they had, and to what degree have they contributed to greater
independence and participation in society and the labour market? It is important to
have such an ambitious approach, but it presents many methodological challenges
for the researcher. What in fact can be measured, and what are the contexts in which
we find the reasons? In our two projects we have been interested in evaluating the
integration process, and the degree to which equality and participation of refugees
has been achieved after some years in Norway. The methodological questions are,
however, the same whether they are centred on refugees or the more established
immigrant population.

We have chosen to combine both quantitative and qualitative methods in the two
projects. The qualitative methods have been especially important in obtaining an
overview of indicators such as work, education, political engagement and
organizational association. These are some of the indicators which recur in the
research on integration and equality. However, it has been necessary for us to study
these issues more deeply. In part it has been necessary for us to question how
suitable these objective indicators are in studying social processes, while it has also
been necessary to study the conditions which contribute to achieving, or not,
equality and integration.

In order to answer these last questions we have used different forms of qualitative
approaches. In the local-community project (Project I) we have studied the
implementation of refugee policy on the local level and through this we have gained
insight into how the various actors (the local authorities, politicians, professionals,
local population and refugees) interpret and implement refugee and immigration
policy. In the integration project (Project II) we followed this up with more extensive
qualitative interviews of 70 to 80 refugees.

We chose this methodology because we know too little about how refugees
themselves experience the situation. If we are to attain the objectives of our
immigration policy it is important that the principal persons are heard. Immigrants
and former refugees in administrative roles have also assisted in making and

132
deciding the interview guide and methods. In these interviews we have used a
relatively open interview guide. The informants, who come from four different
municipalities, have been selected on the basis of a larger survey in which we used a
number of objective criteria to measure integration. The qualitative study therefore
has the character of being a follow-up in-depth study. It has also been used to
correct and elaborate on findings from the survey on several essential points. One of
these is the relationship with the local community and the neighbourhood, where the
survey gave the impression that there was a considerable degree of isolation and a
lack of contact between immigrants and Norwegians. The more in-depth interviews
revealed this to be a more complex issue by describing the qualitative aspects of
contact, different contact patterns etc.

Success or failure?

It is difficult to measure the results stemming from a development project like this.
The positive results will eventually emerge as long-term effects. The project lasted
for a period of three years. This is not long enough to record a measurable change in
terms of integration level. Nevertheless, it is possible to assess the success or failure
of the project by changing the success criteria (the indicators of success). This can
be done with this project. Our main focus has been on the local authorities' efforts to
implement the integration policy. In the action-research section of the project we
analysed refugee assistance work in the 28 cases and network-municipalities. These
municipalities setup their own local projects with the objective to implement new
models and methods for refugee assistance work.

Most of the municipalities succeeded in their efforts. Within the project period they
were able to prove that new methods could improve their integration work - in terms
of lower rates of unemployment, higher degree of language abilities among the
refugees, more social contact between refugees and Norwegians and so on. This, of
course, is a possibility of measuring the results of the project. But - as we have
already emphasized - our main interest in this project has been to study the
conditions for the results. This means that we should not only study (and measure)
the results per se (unemployment rates, language abilities, social contact etc.), but
also strive to answer questions such as "how" and "why".

During the project period we have gained experience with many different models,

133
methods and organizational practices. The municipalities were very inventive both in
utilizing local possibilities and adopting new ways of thinking. We have stressed that
the solutions must take at least two issues into consideration: Firstly, the overall
goals or objectives; secondly, the local situation - with it's limitations and
possibilities. We quite often see that individuals and organisations forget the second
issue. They try to arrive at universal solutions, but (at least in this area) there are
very few universal ways of doing things. Of course it is possible to establish some
sort of guidelines, but of necessity they must be quite general. All the details and all
the practical solutions have to be closely related to the local situation. In short: We
have to avoid any kind of blueprint!

Nevertheless, this does not mean that general advice is not of any interest. The
findings from the 28 municipalities reveal some important general factors:

- The integration work has to be anchored in the local political and ad-
ministrative system. There must be a common attitude on the modalities of
how to implement integration policy.

- The integration work has to be based on cooperation between professional

groups and organisational boundaries.

- The integration work has to be based on the participation of the immigrant

groups themselves.

- The integration work has to be organized in a way which makes it possible

to choose flexibility and differentiation based on the needs of each
individual immigrant.

These are some of the general factors. If the practice is to be changed, these
general factors have to be "translated", made more concrete and adapted to the local
situation in the municipality. This transformation is often difficult. To look into the
efforts of 28 municipalities to improve their integration work has been a learning
process - both for the participants and the researchers. The participants were able to
choose better methods and models for their work while the researchers received

134
valuable knowledge about integration work in general, and about the implementation
barriers in particular. It is difficult to obtain relevant information when using more
quantitatively oriented approaches. As mentioned above, we have also
supplemented this project with quantitative methods (surveys) and more traditional
qualitative methods (in-depth studies), but knowledge about the implementation
process has mainly been derived from the action-research project.

In this project the researcher had the role of an observer, a participant, an adviser
and an assessor. This mixture of roles is unusual in the perspective of traditional re-
search, but it is a quite common occurrence in action research. Criticism of the
action-research method has centred on problems of objectivity. The great advantage
of this method is, however, that it enables the researcher to get close to the subjects,
who in this particular case were both refugees and representatives from the local
authorities. Moreover, it is more appropriate for the collection of process-data. As
these two aspects have been of great importance to us; this method was the most
suitable for us.

Has the project been a success or a failure? In our opinion it has been a success. It
is always possible to improve the results, but altogether we are satisfied. We were
also asked to mention if the project was monitored during its implementation. In one
sense - yes. Representatives from the government and from independent interest
groups have followed the project closely, and we have had a special monitoring
group. Furthermore, every year we have submitted special reports to the project's
client, the Norwegian Research Council. An "independent evaluation", however, did
not take place.

*****

135
Sources

Bernard, William S (1973): "Indices of Integration in the American Community" in

International Migration Review, Vol 2, pp. 23-33

Diaz, Jose Alberto (1995): "Integration: A basic Definition and Methodological

Strategy" in Hjarnø, Jan (ed): Multiculturalism in the Nordic Societies, Copenhagen,
Nordic Council of Ministers, TemaNord 1995:516

Esser, Hartmut (1980): Aspekte des Wanderungssoziologie. Assimilation und

Integration von Wandreren, Ethnischen Gruppen und Mindernheiten. Darmstadt:
Luchterhand

Ålund, Aleksandra and Carl-Ulrik Schierup (1991): Paradoxes of Multiculturalism.

Avebury, England

136
 The national immigration statistics project
Eva Hagensen

Introduction

The background of the project was the recognition of the need to generate more and
better statistical and factual information about migration flows and living conditions of
the immigrant population in Norway. Most of the statistical data available in 1992
was situational. Longitudinal data were almost totally lacking.

The initial goals of the project were to build up the necessary competence and lay
the foundation to produce situational and longitudinal statistics on migration flows
(emigration and immigration) and the same criteria of living conditions as available
for the Norwegian population in general.

Phase 1

Norway - like the other Northern countries - has based much of its statistics on the
utilization of administrative registers (register statistics). Statistics Norway is
responsible for almost all official demographic statistics, including vital statistics. The
Central Population Register (CPR) was established in 1964. The main pillar of the
CPR ist the National Identification Number, a unique and permanent 11-digit number
that identifies each person who ever lived in Norway since 1 October 1964. This
individual number system is used as an identification in many registers, and as a
linking key when registers are matched. The Statistics Act of 1989 makes it easier for
Statistics Norway to utilize data from public systems established for administrative
purposes. An Inspectorate for Data Security was established to secure individual
privacy and to prevent misuse of I.D. numbers.

In addition to the I.D. number, the register contains i.a. name, address, citizenship,
marital status and family relations (children, parents, spouses).

The general idea of the National Immigration Statistics Project is to incorporate,

within the limits set by the Inspectorate for Data Security, an immigrant dimension

137
into the relevant existing administrative registers. By creating a link to the Central
Population Register it can be determined who belongs to the category "immigrant
population".

The first central question to be settled within the project was how to classify
"immigrants" for statistical purposes. A new standard for the classification of
immigrants was decided upon in 1993. According to this the "immigrant population"
is classified as first generation immigrants without Norwegian background and their
children (the second generation). In more popular terms, "immigrant population"
means all persons with both parents foreign born.

The classification of "immigrant population" comes in addition to the formerly

established classifications "foreign citizen" and "foreign born". All three classifications
will be used. The classification of "immigrant population" was well received by
various users of statistics.

138
Phase 2

In order to modify the classification of "immigrants", it was decided to build in

information on the type of immigration status. This will be completed during 1995.
This is done by taking information from the Immigration Control Register
(FREMKON). All foreign nationals who apply for residence in Norway, including
asylum seekers and refugees, are registered there. When a residence permit in
Norway is granted for a period of more than 6 months, the Central Population
Register is automatically notified. The reason for linking this information to the
"immigrant population" is to be able to give situational and longitudinal data as to the
situation of eg refugee groups compared to labour immigrant groups and/or the
situation of the total population.

Phase 3

When discussing what to measure when it comes to "immigrants` integration",

account is taken of the aim of Norwegian immigration policy, that is to achieve a
genuine equal status for legal immigrants. As far as possible, legal immigrants are to
have the same opportunities, rights and obligations as the rest of the population.
What should be measured is situational and longitudinal genuine equal status.
Another partial policy goal is that all categories of immigrants should become
economically self-supporting as early as possible. One way to measure this is to
document people's living resources.

A clear cut and precise definition of genuine equal status for legal immigrants and
Norwegians has not been established. It was decided - very much for data reasons -
to limit measurement to begin with to the same components of living conditions that
are available for the Norwegian population, that is: health, employment, family and
social relations, housing and local environment, recreational and cultural activities,
security for life and property and political participation and rights. Living conditions
are measured by drawing a holistic picture of these components and demographic
and economical statistical data.

The Nordic way of describing the populations' living conditions is defined as follows:
"Living conditions are determined by the resources of the individuals when it comes
to income, property, health and education". This "resource approach" means that

139
living conditions are looked upon not only as something that "happens" to people. It
is also something that can be influenced and changed through active steps taken
either by the individual himself or by policy measures.

The production of statistical data to describe the complete picture of living conditions
will still take years.

It is further recognised that some aspects of the living situation and conditions of the
immigrant population or immigrant groups are different from those of the majority
population; eg daily life harassment and discrimination on racial, religious, ethnical
grounds and other forms of negative treatment. The attitudes of the majority
population to immigration and immigrants rights are also considered to play an
important role on the immigrant groups' strategies to adaption and integration. Work
has started to improve data collection and measurements of different aspects of
ethnical discrimination and rascism. The attitudes of the majority population are
measured annually. Statistics Norway summed up in a report (only in Norwegian)
dated June 1995 the available data on migration flows, stock of immigrants and
some aspects of their living conditions.

140
 Integration of foreign workers into the labour market - France,
Germany, The Netherlands and Sweden
Heinz Werner

1. On the necessity of integration efforts

For a number of European countries the presence of a more or less large number of
foreign employees and their dependants is a fact to which the countries must adjust.
With the increasing duration of residence a return to the country of origin becomes
increasingly less probable. Experience demonstrates that a large proportion of the
foreign workers and their families will remain in the country permanently. Seen
against this background, the most comprehensive integration possible of the foreign
workers and their dependants becomes necessary on humanitarian, social and
economic grounds.

On humanitarian grounds, because the people have lived and worked in the country
for some years and have made a contribution to the development of the receiving
country. If they decide to remain in the country permanently it must be made
possible for them to take part in the life of the society of the receiving country as
equals, like nationals of the receiving country. On social grounds integration is
necessary because discrimination against certain groups can lead to their
marginalisation. This in turn can lead to socially undesirable behaviour which is
expressed, for example, in the form of criminal activity or social unrest. From an
economic point of view successful integration is necessary because discrimination in
the labour market has the effect that the human resources (in this case the foreign
workers) are not employed in accordance with their qualifications, which means their
inefficient use. As a result, productivity potentials are not used, and avoidable costs
are incurred.

Also in future the immigration of foreign workers into the industrialised European
countries can be expected to continue. A large proportion of the immigrants will stay.
Integration efforts will be necessary for them, too. Immigration will occur for two
reasons: firstly because of the enormous migration pressure from the South (eg from
the Maghreb countries) and from central and eastern Europe, secondly because of

141
the ageing of the working population of the receiving countries.

2. The concept of integration

Integration can be understood as a process and at the same time as a state. The
integration process proceeds in the direction of a desired state, i.e. the achievement
of successful integration. The integration process has proceeded successfully if, with
respect to participation in social areas of the receiving country, differences between
comparable groups of nationals and foreigners/ethnic minorities no longer exist.
Important areas for integration are housing, education, and work. The pre-requisite
for the successful integration of a person is legal security/legal equality for the
migrants.

But opportunities must not only be the same in law for nationals, non-nationals and
members of ethnic minorities, but conditions must also be the same in actual
practice. In competitive societies, equality before the law rarely suffices by itself to
ensure that people with different cultural backgrounds can take advantage of
society’s opportunities on equal terms. Therefore, disadvantaged persons must be
assisted or enabled to compete on comparable terms with those nationals with
similar characteristics.

Gainful employment is an important component of society and in the life of each

individual as it is necessary for one’s livelihood and determines status in society.
With regard to the labour market a successful integration would have to show the
same employment structures between comparable groups of nationals and
foreigners (eg in unemployment, earnings, employment rate). The pre-requisite for
this is the same access to occupation/work, including the existence of certain
preconditions which enable the foreign workers to practise an occupation on the
same terms as nationals, such as access to the education systems and to training
schemes, housing or social security. These areas are not independent of each other
and can influence each other.

3. Measurement of integration into the labour market

142
Integration into the labour market means a process in the direction of the same
gainful employment structures for foreign economically active persons/ethnic
minorities as for national employees. Similar gainful employment structures signalise
successful integration. Applied specifically to labour market conditions this means
that at the end of the integration process no more differences should be found
between comparable groups of nationals and migrants in employment participation
(activity rate), in unemployment or income, etc.

As integration means a process in the direction of a desired state (same employment

structures), integration could be measured by the deviation (or similarity) of
employment structures between the foreign and national workforces. In the most
simple case of measurement the labour market indicators (eg activity,
unemployment, incomes) for national and foreign gainfully employed persons can be
juxtaposed and compared. Indices could also be calculated which, for example,
indicate the degree of deviation.

A great problem is the delimitation of a comparable national group: in order to be

able to asses the level of integration, comparable groups of nationals and
foreigners/ethnic minorities must be juxtaposed. The national contrast group must, in
its characteristics, as far as possible be identical with the group of foreigners/ethnic
minority. The most important characteristics concerning the labour market are age,
sex, qualifications, occupation, economic sector and region.

Integration also has a time-related aspect. The personal situation and the situation of
the successive generations change in the course of time. If one wants to measure
integration by way of indicators, the situation is compared at a particular point in
time. Individual integration processes showing different developments, at the time of
the measurement can be at different stages of integration. These measurements are
then added up and standardised at a given point in time (eg by the formation of
ratios). The difference in the time sequence and between the generations are,
therefore, blurred since only one mean value is retained at a point in time. Behind
this average value, considerable differences eg of nationality/country of origin etc.,
can continue to exist. A more detailed breakdown, as far as this is statistically
possible, could make these differences visible.

As a rule, one can assume that integration has proceeded further the longer the

143
foreigner has stayed in the receiving country. This must not, however, necessarily be
so as is shown in the report of the French Haut Conseil à l’Immigration. In the current
situation it can quite possibly happen that the children of immigrants, even if they
have French nationality, are less integrated than their parents. The poor labour-
market situation with high unemployment and the consequences of an unbalanced
housing policy partly marginalise the second generations.

Because of the time dimension of integration, the degree of integration can only be
correctly determined in the time sequence, that is with the help of a longitudinal
survey. The individual development of the immigrants and that of the following
generations within the society of the receiving country would have to be followed up
and compared with the national population. This is, for example, possible with a
retrospective survey during which the stages of integration are individually surveyed
on the basis of past development up to the time of the survey.

The measurement of integration in the form of indicators is therefore not

unproblematic. Nevertheless, some indicators are to be presented below which can
be seen as pointers to the level of integration of the foreign workers into the labour
market.

There is no single indicator which depicts the stage of integration reached. Only by
simultaneously looking at several indicators can an overall picture be obtained.
Assessment of the indicators must not be carried out schematically on the basis of
one figure. Behind the one figure in each case other processes can be hidden which
are attributable to differences in the functioning of the labour markets from country to
country.

The number of indicators providing useful information must be small if they are to
have any weight for the assessment of integration and if importance is attached to
their empirical verifiability and international comparability. Below some indicators
selected according to these points of view are presented and discussed.

To assure comparability the following indicators are based on the annual labour force
sample survey of Eurostat and on a comparable sample survey for Sweden.

144
4. Some indicators for the measurement of integration into the labour
market

4.1 Unemployment rate

Unemployment has an unfavourable effect on the integration of foreign workers.

Repeated or lengthy periods of unemployment lead to de-skilling and involve the risk
of marginalisation. In table 1 the unemployment rates for nationals and foreigners for
the four countries considered here are shown for 1993. Table 1 shows that in all four
countries the unemployment rate for foreigners is considerably higher than that for
nationals. It usually amounts to double the unemployment rate for nationals. If one
splits up ages into under 25 (young people) and over 25, the differences between the
rates are even more marked (table 2). The young foreigners reach - except in the
Federal Republic of Germany -considerably higher unemployment rates as the
comparable national age group. This is to be understood as a clear signal for the
education and labour market policies that there is a need for corrective measures to
be taken!

This situation has not improved for the foreign workers since the early 1980s. In
particular, the unemployment rates for young foreign people have remained
frighteningly high. Only in Germany has the unemployment rate for foreign young
people fallen considerably. This is due to the fact that, as a result of demographic
developments and the extended duration of education, fewer German young people
enter the labour market through the so-called dual vocational training system. Thus
foreign young people have more opportunities in the market for training
places/apprenticeships.

The overall rate of unemployment of foreigners conceals marked differences

according to nationalities. For certain groups the unemployment rate is much higher
than for others.

How can these differences in unemployment between nationals and foreigners be

explained? The following reasons can be put forward.

The foreign workers are to a high degree employed in the manufacturing industry.

145
Employment in this sector is generally declining. Moreover, this sector is undergoing
a restructuring and rationalisation process which requires better qualifications on the
part of those employed there. Foreign workers often work in jobs not requiring any
qualifications. Such jobs are increasingly being lost in the course of rationalisation
and restructuring. Redundancies and unemployment of the less qualified workers,
including many foreigners, are the result. Access to further training and retraining is
more difficult for them because foreign workers usually do not have a good
educational basis and also from language deficiencies.

In the case of foreigners, massive losses of industrial jobs are only partly
compensated for by employment in the tertiary sector. In the case of more
demanding jobs or those closer to the customers, such as consulting/sales, they
encounter competition with the national employees. As regards the establishment of
businesses or self-employment, foreigners are usually short of financial capital.

But even if one compares nationals and foreigners who have the same
characteristics such as qualifications, age and sex, the unemployment rate of
foreigners is higher than that of nationals. A residue of unemployment remains which
cannot be accounted for by other features (apart from nationality). This residue, can
only be related to nationality, but does not necessarily have anything directly to do
with the nationality. It can, for example, involve discrimination due to prejudices
regarding ethnic origin.

This discrimination against foreign workers occurs, for example, in the recruitment
and selection processes of the companies. National workers tend to follow formal
application routes, whereas foreigners look to a greater extent for work through a
network of friends and relatives. Thus they do not make use of certain application
procedures.

The labour market problems of the foreign worker are greater during a recession: the
less qualified persons (including many foreigners) are affected to a disproportionate
extent by redundancies. For new recruits, the companies can select from a larger
pool of persons looking for work. This larger pool not only consists of workers who
have become redundant, but also of new additions to the labour market, such as
young people having completed training or women who want to take up employment

146
after a period of occupational inactivity. When they have a greater choice the
companies prefer, all other conditions being equal, to take national workers.

4.2 Income

For the integration of migrants their economic success is important. Integration can
only be successful if material equality between nationals and foreigners (or ethnic
groups) is achieved under otherwise equal conditions. An important prerequisite for
this is equal pay for equal work, as well as equal work in the case of equal
qualification between nationals and foreigners. The same earnings of nationals and
foreigners with otherwise identical characteristics is an indicator of integration into
the labour market. In order to achieve a comparison which provides meaningful
statements, groups of foreign and national employees with equal characteristics
must be compared. Groups with the same characteristics such as sex, age and equal
qualifications are the minimum prerequisites. Unfortunately this statistical basis is
only seldom available. Moreover, qualification is not so easy to cover statistically.
Qualifications are made up of education and abilities and skills learned during
working life. In most statistics only the educational level is given.

With regard to the differences in earnings, it can be stated that for the four countries
looked at here there are differences in payment between nationals and foreigners
with otherwise identical characteristics. However, the foreigners do not form a
homogeneous group. Substantial differences in earnings can be found, for example,
between the foreign workers from different countries of origin. The differences in
earnings between nationals and foreigners are explained by, among other things, the
fact that the qualifications of the foreign workers are usually low so that they are
concentrated in the secondary labour market in which below-average wages are
paid.

Even if one tries to eliminate such determinants by comparing groups of national and
foreign workers which are identical according to economic sectors, qualification and
other features, there are certain differences, even though they are not spectacular.
Ethnic origin/nationality often has a negative effect on earnings.

The differences in earnings between groups with similar qualifications appear at first
sight to be of minor magnitude. However, statistical studies are based on categories

147
of wage and salary scales in companies. There are signs that this in-company
classification does not adequately take into to account the training/qualification of the
foreign employees. The foreign employee is not infrequently put into a category
within the company’s wage hierarchy below the level of his training and
qualifications. Salary categories often exhibit rather slight differences between
national and foreign workers, above all if the wage structure has been fixed by
collective agreement. The bulk of the foreign workers is as a rule found in the lower
segment of the wage and salary hierarchy. Foreign workers remain in this lower part
of the wage and salary scale more often or longer than nationals since, for a number
of reasons, they participate less in company promotion processes.

4.3 Female activity rate

Gainful employment seen in itself is an integration factor and is the basis for covering
the cost of living. It also facilitates personal contact and social relationships.
Employment behaviour is in the most simple case expressed by the activity rate. This
relates the working population (= persons in employment + unemployed persons) to
the population of the same basic total group eg age group. The employment
behaviour and thus the employment rates vary between nationals and foreigners.
However, they have the tendency to converge in the course of time. Therefore, one
can regard the employment rate as an indicator of integration: the closer the
employment rates of the nationals and foreigners have converged, the more similar
is the employment behaviour, and the further integration has progressed. However, it
is not the overall rate, but the activity rate of women which is to be considered. To
explain this choice some general remarks on international migratory processes and
on female employment activity are necessary.

In the initial phases of the migratory process typical in Europe, it is mainly male
workers who move. They look for work with better earnings in comparison with their
home country. The activity rate of foreign men is therefore high. This could not be a
significant indicator of integration into the labour market since the motive for their
coming was gainful employment. For the wives of foreign workers this is different.
They follow their husbands. For these women gainful employment was originally not
the only motive. The foreign worker, as a rule, occupies a place at the bottom end of
the job and income hierarchy. It therefore becomes necessary to increase the

148
family’s income in order to avoid material marginalisation and in order to catch up
with national population groups in comparable occupations. The gainful employment
of the spouse of the foreign worker becomes a necessity.

It is true that foreign women have greater difficulties in finding a job than national
women or foreign men. Their unemployment rate is, generally, above the total rate
for foreigners and considerably higher than the unemployment rate of the national
women (table 3). While the male employee often has established contacts through
his working environment and socialisation processes thus become possible in the
receiving country, this is difficult for foreign women. Their access to the labour
market is therefore not easy. It is made even more difficult due to language
deficiencies, lack of qualifications and cultural distance from working life. These
factors represent an obstacle to integration. Once in the labour market the foreign
women - even more than the foreign men - can be found in the less attractive, badly
paid jobs which are also unstable (highly dependent on the economic cycle, with
limited contracts, dismissal at any time etc.). Periods of employment and
unemployment alternate.

The special employment problems of foreign women are reflected in different

employment rates in comparison with the national women. In the course of time the
employment rates between foreign and national women should converge, which
would mean a step in the direction of integration. The difference between the two
employment rates can therefore be regarded as an indicator of integration or non-
integration into the labour market.

In table 4 the activity rates of foreign women and national women of the four
countries examined here are shown. As the employment activity of women varies
according to age, activity rates for specific age groups are compared. Table 4 shows
that the activity rates of foreign women were, across the board, considerably below
those of national women. Compared with 1983 no major convergence has taken
place. These deviations between rates indicate a lack of integration of foreign
women. If one makes a comparison with women from EU countries their activity rate
is often close to that of national women. This may be linked to the fact that, on
average, women from EU countries stay in the particular country for a longer time
than women from third countries. Thus the integration process has proceeded
further. Furthermore, EU foreigners will, due to the ruling on free movement, tend to

149
return to their home countries sooner in the event of economic or other difficulties -
such as integration problems. For them this does not mean an irreversible decision
as it does for dependants from third countries of the EU, who, having returned to
their countries of origin, as a rule, are not able to go back to the former of country of
employment.

4. 4 Other indicators

As already mentioned during the discussion on the indicators of integration into the
labour market, there is no single indicator which reliably reflects the degree or the
rate of integration into the labour market. Only by simultaneous consideration of
several indicators can an overall picture be obtained. Each of the indicators
mentioned so far describes a particular partial situation in the labour market. Besides
the indicators of integration into the labour market so far mentioned one can imagine
others (without any empirical evidence of them being provided here):

(1) The degree of employment of nationals and foreigners in atypical forms of

employment can be compared. Examples of such forms are contracts of
employment for a limited period, as well as part-time work and manpower
hire arrangements.

For employers, the use of atypical forms of employment are attractive because in
that way they can adjust personnel to fluctuating production requirements. From the
point of view of employees, atypical forms of employment are not always
automatically the worse alternative compared with normal contracts of employment.
In part, they correspond to their interests. This applies, eg in the case of women with
children who wish to have part-time work or in the case of schoolchildren and
students who are not interested in permanent work. In most cases, as a result of the
employment situation, persons looking for work as a rule have no alternative to
atypical employment.

Typical forms of employment are part of the so-called secondary labour market, and
therefore reflect a precarious situation for employees in the labour market. If foreign
employees are found in such employment to a disproportionately high extent this
would be an indication of inadequate integration into the labour market. There are

150
indications that foreigners can be found to a disproportionately high extent in so-
called precarious employment.

(2) A further indication of the integration of foreigners/ethnic groups into the

labour market can be the per centage of foreign self-employed persons
among active foreigners. The decision to take up self-employment can, as
a rule, only be made after a certain knowledge of the local labour market
has been obtained, which presupposes a certain familiarity with local
conditions. Self-employment also tends to suggest the intention of staying
permanently in the receiving country. The possibility of a foreigner setting
up business indicates a certain freedom of access to this part of the labour
market in the receiving country. As a matter of principle, free access to the
labour market has an integration-promoting effect.

Access of foreigner can be restricted if the practice of certain forms of self-

employment is linked to the nationality of the country of residence (for example
‘tabac’ shops in France). Access can also be restrictive as a result of the
requirements for admission to certain forms of self-employment being difficult to fulfil
by foreigners. Thus for example, the running of a tradesman’s business in Germany
is tied to the "Meisterprüfung" (master tradesman’s examination). The latter
presupposes a final examination in a recognised occupation. This precondition
cannot be fulfilled by foreigners who have not passed through the German vocational
training system. About two thirds of young Germans pass through this vocational
training system as compared to one third of young foreigners. The training takes
three years as a rule. It should also be pointed out that within the scope of the
realisation of the single European market the access conditions to self-employment
in trades and crafts for EU citizens will no longer be tied to this rigid requirement of
the "Meisterprüfung".

The percentage of self-employed foreigners can, with a certain justification, be

regarded as an indicator of integration into the labour market. However, the
circumstances of the development of self-employed activity and the type of
companies must be taken into account. Were the foreigners forced into self-
employment due to the poor labour market situation? What income can be realised
with it and how viable are the companies? In order to answer these questions more
research is necessary.

151
According to the labour force survey by the Statistical Office of the European
Communities (EUROSTAT) in 1991 the per centage of self-employed persons
among the foreign economically active persons - with a rising tendency - was for
France 9.0 per cent, for Germany 6.4 per cent, for the Netherlands 8.1 per cent and,
from a comparative survey for Sweden, 7.2 per cent. From the 1990 census for
France it is interesting to note that the share of self-employed persons among the
naturalised French (français par acquisition) is one third higher than for foreigners.
The role of nationality in connection with the establishment of companies needs to
be taken into account in international comparisons.

(3) Training represents a decisive factor for integration into the labour market.
As a rule education/training promotes access to employment and makes
equality of opportunity possible with regard to incomes and choice of
occupation. Success or lack of success at school not only determines later
opportunities in the labour market, but leads to social behaviour which can
influence the course of life positively or negatively (marginalisation,
criminality). The proportions of the second and third generations of
foreigners which pass through the various levels of the education/training
system are an important indicator of integration into the education system
and the society of the particular country. The figures that lend themselves
to comparison are those according to age groups of foreign school children
at the various levels of the education system, their drop-out rates, failure
rates and success rates in comparison with national schoolchildren. These
shares give indications of subsequent problems of opportunity in the labour
market. Within the scope of this report, however, it is not possible to also
provide empirical evidence of these key figures for the four countries.

5. Concluding remarks

In conclusion it must be stated that a general convergence of the labour market

conditions between foreigners/ethnic groups and national workers cannot be
expected in the near future. The foreign employees, as a rule, accept jobs in the
lower wage/salary segment. (There is also an increasing "migration" of high-skilled
manpower. But their numbers are still relatively low). Upward mobility is generally not

152
easy, and for foreigners it is especially difficult, but even more so where there is
competition with nationals (language deficiencies and lack of qualifications,
discrimination during the recruiting process etc). Improvements in the vocational
situation are therefore primarily to be expected from one generation to another. This
takes time. However, it depends to a high degree on integration policies that must
counter or eliminate segregation within the socially important areas of housing,
education and the labour market.

In the case of integration policy a distinction can be made between direct (special)
and indirect (general) measures. The former start from the special needs of the
immigrants who require direct measures in order to improve their situation and
integration. Closely associated with the direct integration policy is the concept of
minorities’ policy. This is based on special regulations and measures for the
particular minorities. The indirect integration policy is part of the general social policy
which, to a large extent, includes the migrants on equal terms. The more
comprehensive the policies in the areas of housing, education and social security
and the more equality forms the basic principle of the policies, the more one can
assume that the system also contributes to the integration of the foreigners/ethnic
groups.

In the development of integration policies of the four countries looked at here a

convergence can be detected. Countries which have pursued a multicultural
integration policy, such as the Netherlands and partly Sweden, are converging with
the policies of mono-culturally oriented countries such as France. On the other hand,
in the Federal Republic of Germany during the last few years an intensive discussion
on the multi-cultural society has been taking place. The massive influx of asylum
seekers has meanwhile unfortunately led to ugly incidents and interrupted this
discussion.

Furthermore, it should be pointed out that despite a large range of integration efforts,
even if they are made with differing intensity, the results tend to be similar: the
unemployment of foreigners/ethnic minorities is considerably above average, partly
with a rising tendency. The comparison also revealed considerable differences in the
labour market situation of the foreigners/ethnic minorities among the countries as
well as between the immigrant nationalities within any country.

153
Successful integration can better be achieved if one manages to control immigration.
Uncontrolled, massive admission would make integration efforts more difficult,
competition between nationals and foreigners in the labour and housing markets
would rise. The overburdening of the education system and the social security
system and the resultant high costs would not be accepted by the national
population. Sweden, which of the countries looked at here, pursues the most
conceptually sophisticated and most consistent immigration and integration policies,
has always regulated immigration to a high degree. The system of admission,
accommodation, training and integration into the labour market can otherwise not
take into account the special needs of the migrants, especially during the first years.
The longer the migrants remain without work or in employment well below their
potential, the greater is the devaluation of their human capital, i.e. of their abilities
and skills.

154
Table 1: Unemployment rates by ethnic group and educational level, in %
Turks Morocca Surinam Antillean Dutch
ns ese s

-no education/primary educati- 33 39 39 53 9

on 29 16 29 28 5
-junior secondary vocational 23 23 22 33 7
education 23 32 14 21 5
-junior secondary general edu- 29 29 12 11 4
cation
-senior secondary vocational
* 28 10 11 9
education
-senior secondary general edu-
cation and grammar schools
-higher vocational education
and university education

* = data insufficient
source: ISEO/SPVA

155
 Table 2: Ethnic origin based on different criteria of identification

Criteria of identification

Nationality Nationality Country of birth country of birth country of birth ethnic origin
person (GLP) father (GLF) mother (GLF) (combination
GLP/GLF/GLM)

Index
- Dutch 112 108 105 104 100
- Italians 53 45 87 51 100

- Antilleans 0 69 61 65 100
- Surinam 6 66 87 87 100

- Moroccans 88 69 98 95 100

- Turks 93 69 99 97 100

156
 Table 1: Unemployment rates of nationals and non-nationals in France, Germany, The Netherlands and Sweden in 1993 by age
(below and above 25 years of age) an European Community nationals or Non-EC-nationals

nationals foreigners
total below 25 year Total below 25 EC-nationals Non-
nationals

France 10.8 25.2 20.6 32.3 11.5

Germany(W) 7.1 4.9 12.7 14.1 7.9

Netherlands 5.7 9.8 19.6 25.0 8.1

Sweden 7.1 18.0 20.8 27.8 12.0

Sources: Eurostat: Community labour force survey (for F, NL, D)

SCB (Central Statistical Office): Labour force sample survey, annual averages (for S) and own calculations

157
Table 2: Unemployment rates for nationals and foreigners by age in France, Germany, the Netherlands and Sweden 1983 - 1993/94
France Fed. Rep. of Germany Netherlands
Nationals Foreigners Nationals Foreigners Nationals Foreign
Total < 25 Total <25 Total < 25 Total <25 Total < 25

1983 7.4 19.1 14.5 30.0 6.0 10.1 11.3 18.2 11.3 20.5

1984 9.0 23.8 16.6 34.3 6.3 9.8 11.3 17.1

1985 9.6 24.8 18.5 39.1 6.4 9.3 12.0 17.4 9.9 16.9

1986 9.7 23.2 18.6 36.8 6.1 7.3 12.0 14.8

1987 10.2 22.8 19.0 34.0 6.3 6.9 12.5 15.4 9.4 16.4

1988 9.6 21.7 18.5 30.0 5.9 6.4 10.9 12.7 8.8 13.7

1989 9.0 19.0 17.8 27.9 5.4 5.2 9.3 9.9 8.1 12.5

1990 8.8 19.2 17.0 29.0 4.5 4.3 8.6 7.0 7.1 10.5

1991 8.7 19.4 16.7 25.8 3.7 3.2 8.0 7.1 6.6 10.4

1992 9.7 21.2 18.8 28.8 3.6 3.4 8.9 10.1 5.1 7.7

1993 10.8 25.2 20.6 32.3 4.9 4.8 12.7 14.1 5.7 9.8

1994 11.7 27.6 24.4 41.9

Sources: EUROSTAT: Community labour force survey (for France, Germany and the Netherlands); for Sweden: Labour force sample sur
and own calculations. EUROSTAT’s methodology was modified slightly after 1991, which affected the rates for the Netherlands more tha

158
Table 3: Unemployment rates for national and foreign women in France, Germany, the Netherlands and Sweden 1983 - 1993/94 in per cent.

Country France Germany Netherlands

Year Nationals Foreigners Nationals Foreigners Nationals Foreig

1983 10.0 19.1 7.0 14.2 13.3

1984 11.7 19.8 7.8 13.7

1985 12.1 23.8 8.0 14.3 12.0

1986 11.7 23.1 7.7 15.6

1987 13.0 23.6 7.4 15.2 13.5

1988 12.2 24.8 7.5 14.5 12.3

1989 12.0 24.7 7.0 13.6 11.4

1990 11.7 22.7 5.8 10.2 10.4

1991 11.2 21.1 4.5 8.6 9.5

1992 12.3 25.3 4.3 9.4 7.5

1993 13.0 23.5 5.4 13.1 7.4

1994 13.7 28.3

Source: EUROSTAT: Community labour force survey (for F, D, NL) SCB (Central Statistical Office): Labour force sample survey, annual averages (f

159
Table 4: Female activity rates of nationals and foreigners by age groups1993, 1991 (1983)

France
age group nationals foreigners
total EC-nationals
1993 1991 1983 1993 1991 1983 1993 1991 1983
15 - 24 37 35 46 31 29 34 41 39
25 - 49 79 77 70 56 51 46 67 69
50 - 64 39 39 39 37 32 33 42 39

Germany (West)
age group nationals foreigners
total EC-nationals
1993 1991 1983 1993 1991 1983 1993 1991 1983
15 - 24 56 54 48 42 37 37 48 43
25 - 49 75 69 58 59 55 57 64 63
50 - 64 42 39 33 50 44 50 49 47

Netherlands
age group nationals foreigners
total EC-nationals
1993 1991 1983 1993 1991 1983 1993 1991
15 - 24 61 58 45 38 37 30 50 57
25 - 49 67 63 46 43 41 39 56 63
50 - 64 27 25 18 17

160
 Sweden
age group nationals foreigners
Total EC-nationals
1993 1991 1983 1993 1991 1983 1993 1991
16 - 24 53 65 64 36 56 52
25 - 49 89 92 89 65 73 77
50 - 64 64 68 63 36 46 45
1993 refers to total Germany
Source: Eurostat (Community Labour Force Survey) and SCB (Central Statistical Office): Labour Force Survey, annual averages (Sweden) and

Heinz Werner

161
Table 1: Unemployment rates of nationals and non-nationals in France, Germany, The Netherlands and Sweden in 1993 by age
(below and above 25 years of age) an European Community nationals or Non-EC-nationals

nationals foreigners
total below 25 year Total below 25 EC-nationals Non
nation

France 10.8 25.2 20.6 32.3 11.5

Germany(W) 7.1 4.9 12.7 14.1 7.9

Netherlands 5.7 9.8 19.6 25.0 8.1

Sweden 7.1 18.0 20.8 27.8 12.0

Sources: Eurostat: Community labour force survey (for F, NL, D)

SCB (Central Statistical Office): Labour force sample survey, annual averages (for S) and own calculations

162
 Table 2: Unemployment rates for nationals and foreigners by age in France, Germany, the Netherlands and Sweden 1983 - 1993/94
France Fed. Rep. of Germany Netherlands
Nationals Foreigners Nationals Foreigners Nationals Foreigners
Total < 25 Total <25 Total < 25 Total <25 Total < 25 Total

1983 7.4 19.1 14.5 30.0 6.0 10.1 11.3 18.2 11.3 20.5 23.7

1984 9.0 23.8 16.6 34.3 6.3 9.8 11.3 17.1

1985 9.6 24.8 18.5 39.1 6.4 9.3 12.0 17.4 9.9 16.9 25.6

1986 9.7 23.2 18.6 36.8 6.1 7.3 12.0 14.8

1987 10.2 22.8 19.0 34.0 6.3 6.9 12.5 15.4 9.4 16.4 23.5

1988 9.6 21.7 18.5 30.0 5.9 6.4 10.9 12.7 8.8 13.7 24.9

1989 9.0 19.0 17.8 27.9 5.4 5.2 9.3 9.9 8.1 12.5 25.8

1990 8.8 19.2 17.0 29.0 4.5 4.3 8.6 7.0 7.1 10.5 23.9

1991 8.7 19.4 16.7 25.8 3.7 3.2 8.0 7.1 6.6 10.4 24.0

1992 9.7 21.2 18.8 28.8 3.6 3.4 8.9 10.1 5.1 7.7 16.4

1993 10.8 25.2 20.6 32.3 4.9 4.8 12.7 14.1 5.7 9.8 19.6

1994 11.7 27.6 24.4 41.9

Sources: EUROSTAT: Community labour force survey (for France, Germany and the Netherlands); for Sweden: Labour force sam
Office; and own calculations. EUROSTAT’s methodology was modified slightly after 1991, which affected the rates for the Netherlan

163
Table 3: Unemployment rates for national and foreign women in France, Germany, the Netherlands and Sweden 1983 - 1993/94 in per cent.

Country France Germany Netherlands

Year Nationals Foreigners Nationals Foreigners Nationals

1983 10.0 19.1 7.0 14.2 13.3

1984 11.7 19.8 7.8 13.7

1985 12.1 23.8 8.0 14.3 12.0

1986 11.7 23.1 7.7 15.6

1987 13.0 23.6 7.4 15.2 13.5

1988 12.2 24.8 7.5 14.5 12.3

1989 12.0 24.7 7.0 13.6 11.4

1990 11.7 22.7 5.8 10.2 10.4

1991 11.2 21.1 4.5 8.6 9.5

1992 12.3 25.3 4.3 9.4 7.5

1993 13.0 23.5 5.4 13.1 7.4

1994 13.7 28.3

Source: EUROSTAT: Community labour force survey (for F, D, NL) SCB (Central Statistical Office): Labour force sample survey, annual averag

164
Table 4: Female activity rates of nationals and foreigners by age groups1993, 1991 (1983)

France
age group nationals foreigners
total EC-nationals
1993 1991 1983 1993 1991 1983 1993 1991
15 – 24 37 35 46 31 29 34 41 39
25 – 49 79 77 70 56 51 46 67 69
50 - 64 39 39 39 37 32 33 42 39

Germany (West)
age group nationals foreigners
total EC-nationals
1993 1991 1983 1993 1991 1983 1993 1991
15 – 24 56 54 48 42 37 37 48 43
25 – 49 75 69 58 59 55 57 64 63
50 - 64 42 39 33 50 44 50 49 47

Netherlands
age group nationals foreigners
Total EC-nationals
1993 1991 1983 1993 1991 1983 1993 1991
15 - 24 61 58 45 38 37 30 50 57
25 - 49 67 63 46 43 41 39 56 63
50 - 64 27 25 18 17

165
 Sweden
age group nationals foreigners
Total EC-nationals
1993 1991 1983 1993 1991 1983 1993 1991
16 - 24 53 65 64 36 56 52
25 - 49 89 92 89 65 73 77
50 - 64 64 68 63 36 46 45
1993 refers to total Germany
Source: Eurostat (Community Labour Force Survey) and SCB (Central Statistical Office): Labour Force Survey, annual averages (Sweden) and

Heinz Werner

166
 Experience in measuring refugees' integration in finland
Elina Ekholm

Background

The population of Finland is ethnically more homogenous than the population of

most European countries. The largest minority group are the Swedish speaking
Finns, differentiated from the majority population on linguistic rather than ethnic
grounds. The "old" ethnic minorities in Finland are Sámi and Romany populations,
and the smaller Jewish and Tatar groups.

The number of migrants in the total population is 1.2 per cent, which is the lowest in
Western Europe. However, Finland has rapidly changed from a country of emigration
into a country of immigration as a result of the inflow of refugees, an increase in
mixed marriages and, most important of all, the immigration of ethnic Finns from the
former Soviet Union.

A marked increase in the number of immigrants and refugees at the beginning of the
1990s coincided with the worst economic recession in Finland since World War II. In
1995, the unemployment rate for Finns is 19 per cent and for the migrant population
over 50 per cent. Occupational downgrading is noticeable among those migrants
who have managed to get jobs. Due to the economic crisis and mass unemployment,
the political parties and leading politicians have been reluctant to consider
immigration and refugee issues. However, as part of its Programme the Government
of Prime Minister Paavo Lipponen will develop an aliens and refugees policy.

Finland had considerable experience of organizing the reception of refugees in the

1920s and 1930s. During the three decades after World War II, Finland received
very few refugees. The regular reception of refugees only started in the 1980s. The
annual refugee quota is nowadays 500. The Advisory Board on Migrant and Refugee
Matters (a cross-administrative body dealing with migration matters) has proposed
raising Finland´s quota gradually to one thousand. Finland has received within the
framework of the quota Chileans, Vietnamese, Iranians, Iraqis, Kurds, refugees from
the former Yugoslavia and other groups. Over 120 municipalities out of 455 have

167
received refugees. The reception of refugees has become a permanent part of the
functions of the State and the municipalities.

The annual number of asylum seekers increased notably at the beginning of the
1990s. This increase brought about a completely new situation in the reception of
refugees. New reception units were established at a rapid pace. In the first phase it
was most important to provide the asylum seekers with accommodation, basic
income security and health care. Later on the reception policy has also been outlined
and the labour division between the State, the Finnish Red Cross, municipalities and
federations of municipalities has been agreed upon.

As the reception of refugees and asylum seekers has increased rapidly, less
attention has been paid to evaluation. In the middle of the 1980s the Research
Department of the Ministry of Social Affairs and Health initiated a study about the
position of Vietnamese and Kampuchean refugees in Finland. The idea was to
repeat a study of the same type at regular intervals, but the study was dropped for
many years. The Academy of Finland, the Ministry of Social Affairs and Health, the
National Agency for Welfare and Health and the Ministry of Labour have financed
various studies dealing with the reception and welfare of refugees, but monitoring of
the reception of refugees and evaluation of measures was not done by the
authorities. The question of compiling statistics on refugees was also in many
respects unsolved. Therefore, the Ministry of Social Affairs and Health and the
municipal central organizations initiated a planning project in November 1992 in
order to create a system of evaluating the reception of refugees. The study
"Integration or Isolation - Life of Refugees in Finland" was finished in October 1994.

Integration or isolation - a study conducted on refugees´ living conditions

The purpose of the integration study was to conduct an overall survey on the
refugees' living conditions and well-being in Finland to serve as the basis for
planning and development. Another goal was to assess how successful the integra-
tion measures have been.

The theoretical starting point for the study was the concept of integration. Integration
was divided into structural integration and internal integration within the ethnic
community. The definition of structural integration is ethnic equality, i.e. that
immigrants could participate in the economic, social and political life of the majority

168
community, taking into account their own cultural and ethnic background. Internal
integration focuses on developing and maintaining the ethnic community's own
culture and mutual support.

The integration process has been evaluated by means of the following indicators:

Employment, income, housing standard and internal migration, knowledge of the

mother tongue and Finnish/ Swedish, education level, use of the mass media, voting,
applying for nationality, contacts with Finns: friends and xenophobia towards
refugees, the significance of the ethnic community, culture and religion, ethnic
organization and willingness to repatriate.

The data consisted of statistics on refugees socio-economic status, a postal

questionnaire to municipal refugee workers and interviews with a hundred refugees
(Iranians, Kurds, Somalians and Vietnamese).

The statistics on refugees' socio-economic situation were difficult to come by since in

Finland, as in many other countries, refugee status does not appear in the population
register. The question was solved by tracing the social security numbers of the
refugees in the material of the Central Statistical Office. The procedure was
approved by the Data Protection Ombudsman.

The aim of the questionnaire was to survey the opinions of social workers in daily
contact with refugees'. The purpose was, in addition to obtaining an overall view of
the well-being of refugees, to study the differences between small and big munici-
palities.

The most important information came from the interviews of refugees. One hundred
refugees were contacted with the help of social and employment authorities and
refugees organisations. Refugees of different age, sex and occupational background
were interviewed, both in the metropolitan area and in the small rural municipalities.
The sample was not, however, statistically representative.

An important principle is that studies concerning minorities should not be carried out
without the participation of the target group, both at the planning and the

169
implementation stage. This principle was adhered to by arranging a meeting with
representatives of associations of refugees in the beginning of the project. Various
methodological and practical questions were discussed at the meeting. These
included, for example, whether when interviewing refugees, it would be better to use
research assistants recruited from among refugees or a Finn assisted by an
interpreter. Refugees preferred the latter alternative. (The interpreter was actually
present in two thirds of the interviews.) The representatives of refugees' associations
participated also in the planning of the questionnaire.

Results of the survey

Structural integration

Full participation in the economic, social and political life of the Finnish society was
realized only in the case of few refugees. Among the interviewed refugees there
were a couple of real winners and some losers, but the great majority tried to get on
with their lives while waiting for the employment situation to get better. The
youngsters who have arrived in Finland as unaccompanied minors had the most
hopeless situation.

On the basis of the results, the greatest obstacle to integration was unemployment.
Compared with that of Finnish nationals, the rate of unemployment in 1992 was
threefold among the refugees. The unemployment rate of refugees was 10 per cent
in 1990 and 61 per cent in 1992. Since then the unemployment situation has
deteriorated further. Almost only vacancies for refugees in the 1990s have been jobs
within the reception system (interpreter, assistant, etc).

Owing to unemployment, refugees are left without economic independence and

contacts with Finns. Unemployment has also increased the Finns' negative attitudes
towards refugees. Many refugees felt isolated and lonely. One third of those
interviewed had experience of xenophobia or open racism. However, half of those
interviewed told that they had Finnish friends.

Refugees have not, so far, been very active in Finnish politics. Over 80 per cent was
going to vote when they are eligible to do so. Participation requires information about
the surrounding society. Many refugees felt that due to the language difficulties they

170
did not have enough knowledge of the global and national events.

Internal integration

Integration of refugees into Finnish society is facilitated by the support from their own
ethnic community. The internal integration within their own group was hampered by
the political and religious controversies, events in the home country and the small
size of the communities in Finland. (With the exclusion of the biggest cities in
Finland, the refugee communities in municipalities are very small.) Their own ethnic
community was important to all the groups of refugees participating in the study, but
it was particularly important to the Vietnamese refugees.

Except Somalis most refugee groups kept their own culture and religion within their
family and own community. The Somali refugees wished, for example, to have Koran
schools and muslin churchyards subsidized by the state.

Service system and coping strategies

According to refugees and refugee workers, the service system has functioned
relatively well. Refugees wished to get more instruction in Finnish for adults and
more instruction in their mother tongue for their children, job opportunities and
greater ethnic communities. They also wanted more information on the status of
refugees in the eyes of the general public in order to overcome prejudices among
Finns. Social workers were of the opinion that the greatest hindrances to integration
were unemployment, prejudices, lack of Finnish/Swedish language skills and the
resettling of too small ethnic communities into too small Finnish municipalities.

However, the refugees have their own means of managing, the most important of
which are work (for those few refugees, who found jobs), studies, family and
concentrating on securing the children's well-being and future. Religion was
parlicularly important coping strategy among Somalis.

Future measures

The integration study gave a voice to refugees' own opinions. The reception system

171
has concentrated too much upon the social and welfare services, and the importance
of the ethnic community and refugees´ own resources has not been used enough.
The Ministry of Social Affairs and Health had initiated training projects in order to find
new methods of furthering the integration work in municipalities. Partly as a result of
the project, the Advisory Board of Migrant and Refugee Matters established in 1994
a multicultural section. Its function was to serve as a forum between authorities and
immigrants' and refugees' organisations.

Discussions have taken place as to whether the integration study should be repeated
at three yearly intervals. The first study was an overall survey. The next study might
concentrate on a specific theme, for example, the living conditions of families and
children. The data basis collected for the first study will also be used in the future, as
Statistics Finland (the Central Statistical Office) will compile yearly statistics
concerning the socio-economic and demographic situation of refugees and
immigrants.

172
 Using indicators of integration for the implementation of
integration policies
Summary and conclusions
prepared by
Manfred Werth, Silke Delfs, Willy Stevens
(Isoplan Institute)

1. Background and aims of the Conference

The issue of measuring the integration of immigrants into their respective European
host societies completes a series of thematic meetings of the Council of Europe's
Specialist Group dealing with equal opportunities for migrants. According to its terms
of reference, the Specialist Group was requested to consider the labour market for
immigrants, their housing situation, their access to and use of social welfare benefits,
their personal security, their cultural rights and their participation in decision-making
processes of their host country. Although many practical aspects were compared
during these meetings, it was felt that the effectiveness of the measures carried out
could not be evaluated and that a tool was missing to compare the experiences
made in different countries.

This is where indicators of integration came into play. The shared impression was
that common indicators of integration used in all member States of the Council of
Europe could help to better monitor the equality (or inequality) of opportunities for
immigrants. It was hoped that the results could be of help when planning political
and/or administrative measures.

The aims of the meeting now coming to a conclusion were twofold: Firstly, the
papers, the presentations and the comments made in the discussions evaluated the
possibilities of making the integration of migrants in the member States of the
Council of Europe measurable by using indicators of integration as a tool of
measurement. Using this evaluation as a basis the meeting's second aim was to find
common ground for developing guidelines for indicators of integration and integration
policies.

173
2. Suggestions and guidelines for using indicators when measuring
integration

1. Definition of target groups

Before discussing the aspects of measuring integration, one already faces

considerable problems when it comes to defining the basic terms giving answer to
the two fundamental questions of whose integration is to be examined and what is
precisely meant when talking about migrants' integration. Answering these questions
is by no means a semantic exercise, but has a bearing on finding common standards
allowing to compare the situation of migrants in different countries:

The national and comparative presentations have shown a good deal of different
understandings as to who is the object of integration policies. The term "migrant" is a
very broad one and yet at the same time insufficient, when it excludes certain groups
who still are target groups for integration measures, such as, for example, the
second generation. On the other hand, there are categories of migrants, eg (not
formally recognised) refugees, which are not considered target groups for integration
policies by some governments. Although the Member States are likely to stick to
their national definitions of "immigrants" for practical reasons, common ground is in
reach when we define "migrants"/"minorities" - in a very pragmatic way - as persons
who have come from abroad or whose parents and grand-parents have come from
another country.

A common definition of "integration" is even more difficult to find. Although there is

some consensus that integration means the incorporation of an individual into the
host country and its society, different concepts of integration are used by policy-
makers in different countries.

These different concepts cover a broad spectrum ranging from ideas close to
assimilation on the one end and multiculturalism on the other. The best consensus in
sight seems to be the notion of equality of opportunities. This definition views
integration to be a process as well as a policy goal and a state.

2. Dimensions of integration

174
Whatever definition of integration we use the conference agrees that the
phenomenon of integration is multidimensional. One can identify at least two
different dimensions: an individual one (private domain) and a societal one (public
domain). The individual dimension refers to the migrants' situation on the labour
market, their housing situation, their educational level or their dependence on social
welfare benefits, whereas by "societal dimension" we mean the legal and social
environment in which the migrant is located. This dimension covers the juridical
provisions for migrants as well as dominant social values in a given country and it
very much influences the migrant's prospects for "success" in integration.

Another way of categorizing dimensions of integration is, for example, the

differentiation of an economic, a social, a political and an internal dimension.

Whatever categorization we choose, we have to consider two main questions: The

first question concerns the consistency of the definition of integration and the
measurement of integration. This means that we have to make a clear decision
concerning our fields of measurement. The second aspect concerns the relevance of
individual dimensions of integration when it comes to establishing of a hierarchy of
indicators. Consensus seems to be reachable that certain indicators such as
education or the employment situation (including their income etc.) of migrants can
be described as key indicators.

Apart from the fact that it would be desirable to analyze all facets of the broad
spectrum of dimensions of integration by the means of further scientific research, a
pragmatic decision should be taken to identify a simple, usable minimum set of
indicators.

3. Typology of indicators

As a result of the meeting we recommend to categorize indicators of integration into

three different types:

a) The first type of indicators - which we would like to call indicators of accessibility -
deals with the legal framework in which migrants live. Legal provisions reflect a
country's political will and determine in how far a migrant has access to crucial

175
sectors of a country, for example, its labour market. A country's legislation also
defines the migrants' status of residence or their protection against discrimination
and thus it needs to be included into the assessment of the migrants' situation in a
given country. Measuring this sort of indicator is admittedly difficult, because of the
different legal traditions in Europe and because the legal framework does not give
any hint as to the implementation of legal rules. But efforts have still to be made to
find ways to include the legal framework into the evaluation of integration. Maybe we
have to give up some of our reservation concerning the handling of data when
dealing with this type of indicator, in order to be able grasp it at all.

b) The second type sets out to describe the actual situation of migrants in their host
country. In this category, the "classical" indicators such as employment situation,
education, dependence on welfare of housing are examined. The basis for this
assessment is mostly provided for by the national statistical offices. When taking a
closer look at this category one in fact evaluates how the legal framework mentioned
above is put into practice, i.e. how the legal rules concerning migrants are
implemented.

c) The third set of indicators takes a closer look at the attitudes of the migrants
towards their host country, their personal expectations concerning their situation and
their participation in social and political processes. But since integration also involves
and concerns the host society, one also needs to examine attitudes of the majority
(indigenous) population. The basis for both analyses are opinion polls and surveys.
They, too, have their flaws, but they are indispensable.

What seems to be important is that all three types are necessary to obtain a
complete picture of the migrants' situation in a specific country.

4. Availability of statistical information

Sufficiently differentiated and comparable statistical information on the migrants'

situation is the precondition for the measurement of integration by using indicators.
Besides it is the background for the political discussion of integration processes.
Therefore, the conference underlines the need for regularly collected and valid data.

From a researches's point of view there can hardly be too much information. And no

176
doubt: there still remains much to be done in order to meet even minimum
expectations. A lot of information is gathered, but not made available to researchers.
This should be changed.

But with all good reason for complaining one should not forget that we can already
formulate a good deal of hypotheses on the progress of migrants' integration with the
data available. When asking for more statistical information we should keep in mind
that the quality of data is more important than quantity alone. We would like to
suggest that first of all some efforts be made in the international field, because when
it comes to comparing different countries in Europe it becomes clear that we still
have a long way to go in order to obtain standardized and comparable data.
Therefore, one recommendation concerning statistics is to take the steps necessary
to make some progress in harmonizing international and European migration
statistics.

177
5. Interpretation of data: The need for comparison

Statistical information alone can only do little when it comes to assessing migrants'
integration. Data need to be interpreted. And when doing that, we have to keep in
mind that the integration of migrants reflects both a social, economic, political and
individual process and a social, economic, individual and political status quo. This
implies that data need to be compared either over time (longitudinal studies) - in
order to determine whether a change has taken place - or to another set of data. This
second set of data serves as a point of reference. Usually, the data of the migrants
are compared with those of the majority population. This does not seem to be the
best method in all cases, because it may tend to neglect certain characteristics of the
migrants and structural characteristics of the host society at the same time. On the
other hand it can be risky to choose a reference group with comparable features,
because we never know for sure whether we really picked the right characteristics
when selecting the members of the reference group.

If a comparison to the average population or some other reference group does not
seem advisable an alternative way could be to compare the migrants' conditions to
legal, economic or social standards and values concerning the general living and
working conditions in the country concerned.

Comparing the situation in different countries also has some pitfalls. Apart from the
problem of comparable statistics we have to be careful when comparing countries
with different legal systems and/or migratory histories. Again, the necessity to look at
all three types of indicators as they were described above becomes clear.

6. Including migrant groups into the evaluation of integration

With all caution being used when interpreting and explaining statistics it has become
clear that statistics alone can by no means satisfy all information needs we have
when it comes to assessing migrants' integration. Therefore, we have to take a look
at migrants themselves in order to get the full picture.

As we have mentioned above; migrants' attitudes and the attitudes of the host
society form an important category of indicators. If we want to grasp that sort of
qualitative information on integration we need to gather opinions of the migrants as

178
well as the opinions of the indigenous population with which they are living. Most of
the time we will have to carry out opinion surveys. This type of instrument is also
most helpful when we want to know more about the participation of migrants in
political processes in the broadest sense (ranging from the use of mass media to
elections) or when we wish to take a look at their social life (leisure time contacts
with the indigenous population).

7. Influence of indicators on public discussion and integration politics

Using indicators for measuring migrants' integration is undoubtably necessary as a

basis for decisions to be taken by policy makers, because they facilitate the
understanding of a very complex phenomenon. At the same time one cannot deny
the implicit danger of indicators contribution to the stigmatization of migrants or to
reinforcing existing stereotypes on migrants. This counter-productive effect can only
be avoided when indicators are used in a very cautions manner. Bearing this in mind,
we nevertheless can conclude that the relevant data collected and the indicators of
integration used to measure integration can also prove that existing sterotypes are
wrong. Furthermore, they can serve as a "tool" for migrants' organisations when
advancing their cause for equality of opportunities.

8. Need for more research

Based on the seven points mentioned above, the experts present at the conference
suggest that more comparative research be undertaken in order to obtain a more
complete picture of the status and process - of migrants' integration in Europe.

179
 LIST OF PARTICIPANTS

Chairman: Mrs Mary COUSSEY,

The Lodge,
Clare College,
GB - Cambridge CD2 ITL.

Mrs Elisabeth SEM CHRISTENSEN,

Directorate of Immigration,
Department of Reception and Integration,
PO Box 8108 Dep,
N-0032 Oslo.

Consultants:

Mr Manfred WERTH, Director, ISOPLAN, Martin-Luther Str. 20, D - 66125

Saarbrücken, GERMANY

Mr Willy STEVENS, ISOPLAN, Martin-Luther Str. 20, D - 66125 Saarbrücken,

GERMANY

Ms Silke DELFS, ISOPLAN, Martin-Luther Str. 20, D - 66125 Saarbrücken,

GERMANY

180
Project organisers:

Ms Karin ALFENAAR, Nederlands Centrum Buitenlanders, Kanaalweg 84B, NL -

3533 HG Utrecht, NETHERLANDS

Ms Berit BERG, Researcher, SINTEF IFIM, N - 7034 Trondheim, NORWAY

Mr Richard BERTHOUD, Head of Ethnic Equality and Diversity Group, Policy

Studies Institute, 100 Park Village East, London GB - NWI 3SR., UNITED
KINGDOM

Mr Raimondo CAGIANO de AZEVEDO, Universita "La Sapienza", Department of

Statistics and Faculty of Economics, Via del Castro Laurenziano 9, I - 00161 Rome,
ITALY

Mr J. DAGEVOS, Institute of Social and Economic Research (ISEO), Erasmus

University, Rotterdam NL - 3000, NETHERLANDS

Ms Elina EKHOLM, Immigration Training, Helsinki Institute of Social and Welfare

Studies, Ruutikatu 4 C 30, F - 02600 Espo, FINLAND

Ms Marian FITZGERALD, Home Office, Research and Planning Unit, 50 Queen's

Anne's Gate, London GB - SWIH 9A, UNITED KINGDOM

Ms Eva HAAGENSEN, Head of Research and Planning, Ministry of Local

Government and Labour, Postbox 8112 DEP, N-0032 Oslo, NORWAY

Mr Chistoph HOFINGER, Institut für Höhere Studien, Stumpergasse 56, A - 1060

Vienna, AUSTRIA

Mr Philip MUUS, Department of Human Geography, University of Amsterdam,

Institute of Social Demography, Nieuve Prinsengracht 130, NL - 1018 VZ
Amsterdam, NETHERLANDS

181
Mr Lars ØSTBY, EUROSTAT, Unit E1, Bâtiment "Jean Monnet", C3192, Kirchberg,
L - 2920, Luxembourg, LUXEMBOURG

Mme Marie POINSOT, Agence pour le développement des relations interculturelles

(ADRI)), 4, rue René Villermé, F - 75011 Paris, FRANCE

Mr Günther SCHULTZE, Friedrich Ebert-Stiftung, D - 53170 Bonn, GERMANY

Mr Andy TEAGUE, Office of Population Censuses and Surveys, Segensworth Road,

Titchfield, Fareham, Hants GB - PO15 5RR, UNITED KINGDOM

Mme Michèle TRIBALAT, Institut National des Etudes Démographiques - (INED),

27 rue du Commandeur, F - 75675 Paris, FRANCE

Mr Heinz WERNER, Institute for Employment Research (IAB) of the Federal

Employment Services, Regensburger Strasse 104, D - 90327 Nürnberg, GERMANY

182