This article was downloaded by: [The UC Irvine Libraries]

On: 09 June 2015, At: 10:31

Publisher: Routledge
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered
office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Journal of Ethnic and Migration Studies

Publication details, including instructions for authors and
subscription information:
http://www.tandfonline.com/loi/cjms20

Does Mandatory Integration Matter?

Effects of Civic Requirements on
Immigrant Socio-economic and Political
Outcomes
Sara Wallace Goodman & Matthew Wright
Published online: 09 Jun 2015.

Click for updates

To cite this article: Sara Wallace Goodman & Matthew Wright (2015): Does Mandatory Integration
Matter? Effects of Civic Requirements on Immigrant Socio-economic and Political Outcomes, Journal
of Ethnic and Migration Studies, DOI: 10.1080/1369183X.2015.1042434

To link to this article: http://dx.doi.org/10.1080/1369183X.2015.1042434

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the
“Content”) contained in the publications on our platform. However, Taylor & Francis,
our agents, and our licensors make no representations or warranties whatsoever as to
the accuracy, completeness, or suitability for any purpose of the Content. Any opinions
and views expressed in this publication are the opinions and views of the authors,
and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content
should not be relied upon and should be independently verified with primary sources
of information. Taylor and Francis shall not be liable for any losses, actions, claims,
proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or
howsoever caused arising directly or indirectly in connection with, in relation to or arising
out of the use of the Content.

This article may be used for research, teaching, and private study purposes. Any
substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing,
systematic supply, or distribution in any form to anyone is expressly forbidden. Terms &
 Conditions of access and use can be found at http://www.tandfonline.com/page/terms-
and-conditions
Downloaded by [The UC Irvine Libraries] at 10:31 09 June 2015
 Journal of Ethnic and Migration Studies, 2015
http://dx.doi.org/10.1080/1369183X.2015.1042434

Does Mandatory Integration Matter?

Effects of Civic Requirements on
Immigrant Socio-economic and Political
Outcomes
Downloaded by [The UC Irvine Libraries] at 10:31 09 June 2015

Sara Wallace Goodman and Matthew Wright

Several Western European countries have adopted policies of mandatory integration,

requiring immigrants to acquire country knowledge, language and values as conditions
for immigration, settlement and citizenship. The underlying concept is that—in
promoting civic skills—immigrants are better equipped to politically, socially and
economically integrate. However, the question of whether civic integration is designed to
be a real solution to repair integration problems has gone untested. This paper presents
results, across a wide range of outcomes, which finds much more support for a symbolic
narrative than a functional one. Using a unique data-set to measure civic integration
policy across six waves of the European Social Survey (2002–2012), we find little
evidence that these requirements produce tangible, long-term integration change. This
does not diminish their significance; instead, we find requirements serve a meaningful
gate-keeping role, while simultaneously repositioning the state closer to immigrant lives.

Keywords: Integration; Immigration; Europe

Introduction
Immigrant integration has proven to be a central challenge for the immigrant-
receiving societies of Western Europe. Not content to just allow processes of
acclimation and accommodation occur over time, many states have assumed a more
proactive stance by erecting a series of policies designed to facilitate, expedite, and

Sara Wallace Goodman, Department of Political Science, University of California, Irvine, 5257 Social Science
Plaza B, Mail Code: 5100, Irvine, CA 92697, USA. Correspondence to: Department of Political Science,
University of California, Irvine, 5257 Social Science Plaza B, Mail Code: 5100, Irvine, CA 92697, USA. E-mail: s.
goodman@uci.edu. Matthew Wright, Department of Government, American University, 4400 Massachusetts
Avenue, NW, Washington, DC 20016, USA. Correspondence to: Department of Government, American
University, 4400 Massachusetts Avenue, NW, Washington, DC 20016, USA. E-mail: mwright@american.edu

© 2015 Taylor & Francis

 2 S.W. Goodman & M. Wright
sometimes even troubleshoot the incorporation of a diverse group of outsiders. As
such, integration policies stand out as a core institutional feature that distinguishes
countries of immigration: while nearly every country in the world receives
immigrants, a far more limited set of states maintain policies that go beyond ‘passive
reception’—that is, regulation of entry and quotas—to tackle accommodation and
inclusion.
One integration policy that has proved to be increasingly utilised, and of great
interest to policy-makers and academics alike, is civic integration, in which
integration arises through immigrants’ acquisition of ‘citizen-like’, or civic, skills.
These might include speaking the host country language, having knowledge about the
country’s history, culture and rules, and understanding and following the liberal-
democratic values that underscore their new home. Civic integration policy promotes
Downloaded by [The UC Irvine Libraries] at 10:31 09 June 2015

these characteristics through the use of new assessment tools including integration
tests, courses and contracts. These tools are uniquely applied as conditions in the
process of obtaining citizenship and, increasingly, to non-traditional membership
statuses, including long-term/permanent residence and entry. In this respect, new
requirements both prepare an immigrant for fuller participation in a host society and
serve as gatekeeper, where the path of transitioning from an outsider to an insider is
marked by mandatory integration conditions alongside other restrictions.
Popular consensus among academics is that this new approach to immigrant
integration has arisen in the shadow of the ‘failure of multiculturalism’.1 However, it
is not quite right to view mandatory integration as a replacement for multi-
culturalism. Indeed, it has proven rhetorically popular even in states with no
experience with multicultural policy, like Germany and France.2 And in states that do
maintain multiculturalism, we see no evidence that civic integration has eroded said
policies (Banting and Kymlicka 2013). Though this premise is empirically problem-
atic as a point of departure, we can identify significant conceptual differences in the
two integration approaches. Multiculturalism offers rules and procedures that
promote accommodation along group lines, ranging from home language instruction
to funding for religious schools, while civic integration is individual-oriented (so
much so that spouses, formerly included on family visas, are independently required
to demonstrate language and country knowledge). Moreover, there are important
differences in the mechanism; civic integration compels substantive integration
through requirements, specifically course attendance, language tests, etc. By contrast,
multiculturalism maintains a set of arrangements that may produce integration along
group lines but does not comprehensively facilitate them.
Civic integration is not merely novel, it is prolific: 12 out of 15 Western European
states have taken part in this ‘civic turn’ (Mouritsen 2008), adopting robust policies of
language acquisition, civic training and liberal value commitments since 1997
(Goodman 2014). These include high-volume immigrant receiving societies which
are traditionally considered divergent in integration policy approaches, such as the
Netherlands, Germany, France and the UK. Taking a macro-perspective, Christian
Joppke (2007, 14) describes this cross-adoption as policy convergence, in which a
 Journal of Ethnic and Migration Studies 3
‘focus on “obligation” (and reverse de-emphasis on “rights”) is a shared feature of all
of them’. Yet, despite the widespread nature of this policy, its high political visibility
and policy relevance, policy-makers and academics still know very little about its
effects. Governments busily monitor these policies, but their measures of impact
(primarily reported in test pass rates and levels of acquired proficiency) do not tap
into larger questions of effects of civic integration on societal integration generally,
such as comparative levels of well-being, political efficacy or sense of belonging.
Moreover, these studies also do not get at the question of impact cross-nationally,
where contextual factors can be taken into account.
This research paper picks up these tasks by examining the effects of civic
integration policies on immigrant integration outcomes. Do mandatory integration
requirements matter? Are there conditions in which mandatory skills acquisition
Downloaded by [The UC Irvine Libraries] at 10:31 09 June 2015

improve an immigrant’s socio-political status, political inclusion or economic

performance? Do civic integration investments meaningfully alter the life chance of
an immigrant in an advanced, industrialised host society? Existing studies make use
of individual interviews and focus groups (Strik et al. 2010; van Oers 2013), or run
small case studies (e.g. Ministry of Immigration 2009) making it difficult to cross-
nationally evaluate the policy effects of mandatory integration. This paper is the first
to systematically examine these questions in a broadly comparative way.
We offer several contributions here. First, we bring evidence to bear on the
question of whether mandatory integration matters. There is a particular speculation
among scholars over the question of whether civic integration is designed to be a real
solution to repair integration problems or whether civic integration is a symbolic
gesture, to pacify native citizens that immigrants do not threaten their way of life and
collective identity or to support enactment of new immigration controls. Our results,
across a wide range of outcomes, speak to this issue directly, finding more support for
the symbolic narrative than the functional one. We observe a positive effect of civic
integration policy in political integration, but results are null for socio-economic
domains. This, when examined alongside qualitative data and pass rates of exams and
course completion, supports an interpretation of mandatory integration not as a
central instrument of long-term immigrant integration but overall an effective
mechanism of immigrant regulation—controlling the number of people who
immigrate, obtain residence and naturalise.
Additionally, we build upon a growing literature among migration and citizenship
studies that moves beyond the study of policy origins and towards a methodologically
grounded concern for policy effects. The past decade has seen a proliferation of policy
measures, first capturing dimensions of citizenship policy and, more recently,
dimensions of immigration policy (Goodman 2013). In light of massive changes
observed across Western Europe to implement civic integration policies, a systematic
examination of policy effects is warranted.
Finally, given the findings here, the paper not only sheds light on the evolving
nature of integration policy, it also, perhaps more significantly, provides evidence to
consider the newly defined and closer role of the state in the lives of individuals
 4 S.W. Goodman & M. Wright
through mandatory integration courses and exams. Stepping aside from the question
of effect of policy on individual integration, we discuss this outcome of the
repositioned state as value in of itself.

Theory and Literature: Assessing Civic Integration Policy

The empirical literature on immigrant integration is vast but suffers from theoretical
under-specification and lack of systematic analysis of effects. More problematic,
where scholarly opinion regarding the effects multicultural policies on integration
outcomes remains divided,3 civic integration policy remains largely unexplored
altogether. Therefore, in order to model effects of policy on integration outcomes, it is
critical to first examine the unique contours of civic integration policy and how other
Downloaded by [The UC Irvine Libraries] at 10:31 09 June 2015

scholars have interpreted policy design.

One of the central observations motivating the first civic integration policy (the
Dutch Newcomers Integration Act of 1998) was ‘that many immigrants arriving in
the Netherlands were not sufficiently qualified to enter the labour market on the
requirements set by a highly developed economy’ (Entzinger 2003, 76). As such, in its
initial design, civic integration encouraged immigrants to integrate as individuals
while emphasising the shared responsibility between immigrant and state by
participating in state-organised courses and orientation. As it states in the
Explanatory Statement of Newcomers Act, ‘It is the task of the local municipality
to assist immigrants with finding their way in Dutch society, at the same time [it] is
the obligation of newcomers to make use of offered opportunities’ (Klaver and Odé
2009, 62).
To promote socio-economic self-sufficiency and improvement, civic integration
uses new assessment tools such as integration tests, courses and contracts. Several
European states historically assessed language before civic integration, e.g. Germany,
the UK, but civic integration is novel because of this turn towards facilitated learning
and standardised measures with innovative content. Not only is language and country
knowledge assessed through standardised testing, often conducted by contracted
private actors, many have argued that the substance being tested is also standardised,
in which liberal values trump national differences. Joppke (2010, 148), for example,
describes this as the ‘lightening of citizenship’, in which ‘everything and everyone
flees the center’.
Related, and perhaps most significant for testing policy impact, civic integration
does not merely promote integration to this ostensibly inclusive membership
community, it requires it. Completing language and country orientation are
conditions for obtaining a variety of residency statuses, including temporary
residence, settlement and citizenship. It is in this observation that we see the policy’s
inherent irony: the state mandates individuals obtain skills to achieve autonomy.
Given these potentially contradictory policy facets—where individual autonomy is
realised through practices of state paternalism and newcomers ostensibly become
national insiders through potentially non-national content—scholarship has mainly
 Journal of Ethnic and Migration Studies 5
questioned whether conditional integration actually promotes integration or, instead,
produces exclusion (Guild et al. 2009; Kostakopoulou 2010; Böcker and Strik 2011;
Groenendijk 2011; Van Oers 2013). In one study, restrictive integration policies are
found to produce better socio-economic integration, particularly in contrast to
laissez-faire multiculturalism (Koopmans 2010). A study of Dutch integration courses
showed that ‘by learning the language and better understanding how the Dutch
society is put together, it becomes easier for [immigrants] to make contacts’ as well as
operate independently in society (Gelderloos and van Koert 2010). On the other
hand, another comprehensive study of integration programmes found the ‘connec-
tion of the integration requirements with the granting of a certain legal status
(admission, permanent residence or citizenship)’ as ‘not necessary to motivate
migrants, and it inevitably leads to the exclusion of certain groups from a secure legal
Downloaded by [The UC Irvine Libraries] at 10:31 09 June 2015

status’ (Strik et al. 2010, 118). Elsewhere, the absences of institutional barriers to
engagement are found to positively affect political integration (Helbling et al. 2015).
This perspective pervades civic integration studies, where most treat requirements
like language and country knowledge as inherently restrictive, obstructive and
culturally assimilationist.4
However, this debate reifies a false dichotomy. Policies may improve an
immigrant’s labour market mobility by improving language skills while also
prohibiting status acquisition if an immigrant cannot reach a certain level of
proficiency. In other words, it obfuscates two distinct outcomes: a migrant can
experience inclusion or exclusion in social, economic and political life (i.e.
integration) and a migrant can experience inclusion or exclusion in legal status (e.g.
obtain citizenship or not). They are certainly related (status exclusion obviously
retards political incorporation), but hindering status acquisition is only one
component of a migrant’s overall integration. Integration is comprised of four
dimensions: economic, social, cultural and political (Council of Europe 2008). And
where existing research is primarily occupied with how mandatory cultural
conditions inhibit status acquisition (Goodman 2014; Böcker and Strik 2011) or
serve as symbolic instruments of control (Permoser 2012), we know very little of their
effect on these other areas of integration at the individual level.5 In distinguishing
these outcomes of integration (effectiveness of requirements on outcomes concerned
with inclusion) versus status (effectiveness of requirements as a mechanism of
control), the empirical scope of this paper is the former.
A second problem with existing approaches to civic integration can be described as
the underspecification of immigrant expectations. Short-term expectations of civic
integration programmes may greatly differ from long-term expectations. Moreover,
immigrants are not a monolithic group and begin integration with different skill-
levels and background factors. Therefore, different expectations may correspond to
different immigrant profiles. Employing ordinal instead of temporal (short term vs.
long term) categories, we identify three increments of achievement. A first-order
achievement of taking a language course is language acquisition (measured in passing
a test or proceeding to the next proficiency level). A second-order achievement of the
 6 S.W. Goodman & M. Wright
course might be comfort in navigating the health care system, or confidence in
communicating with a child’s teacher at school. It can be distinguished from first-
order achievement in that the skill-acquirer can use newly acquired skills for
immediate functions, not just in a classroom setting or to pass an exam. She/he is not
a beginner but they are also not fluid in the host society as a native in terms of skill,
behaviour and potential. In other words, second-order achievement is a type of
transitional category, acknowledging and allowing for phases and iterations of
integration. Finally, third-order achievements are those that render the migrant close
to or indistinguishable from natives in terms of behaviour and values. Where second-
order achievement is the enablement of functional navigation (in which the
immigrant attends to immediate needs and wants), third-order achievements enables
civic navigation as a member of the polity, where one might identify with shared
Downloaded by [The UC Irvine Libraries] at 10:31 09 June 2015

values, hold a belief in one’s political efficacy or carry a sense of subjective well-being.
Figure 1 provides a visualisation of these three conceptual stages.
Third-order goals capture the full experience of immigrant socio-economic and
political integration. They are central to civic integration and often appear in
orientation coursework, but the barriers and challenges of meeting first-order goals
ensure that these ‘goods’ are invariably delayed. For more proficient speakers or
immigrants with a closer cultural background, third-order goals are more easily and
quickly obtained. Our analysis is not only the first to make this conceptual distinction
and offer this theoretical precision, it is also the first to examine effects of civic
integration policy on third-order achievements, recognising government policy
performance reports sufficiently (though unsystematically and certainly not cross-
nationally) monitor first- and second-order goals.

Third order
civic navigation/identification
with polity

Second order
functional navigation/meeting of
immediate needs

First order
language and/or knowledge
acquisition

Figure 1. Increments of civic achievement.

 Journal of Ethnic and Migration Studies 7
Finally, before we move to consider what type of impact civic integration
requirements have on participating immigrants—in other words, whether integration
tests ‘help or hinder integration’ (Strik 2013)—there is a theoretically prior question
of whether there is an impact at all. Aside from goals of integration and regulation, it
may be that civic integration is simply designed to signal to voters that the
government is ‘doing something’ about immigration. As such, political parties play
a meaningful role by aiming to garner native electoral support, reassuring citizens
that immigrants do not threaten their way of life. This draws on the competitive party
politics literature in which policies are less symbolic for their own sake and more
politically rational.6 In this sense, integration policy may be an area the government
can demonstrate competence and control where so many other aspects of migration
(e.g. border control) seem elusive, particularly where Gary Freeman (1995) observes a
strong disconnect between inclusive policy-making elites and a restrictionist public.
Downloaded by [The UC Irvine Libraries] at 10:31 09 June 2015

Integration policy may help bridge this difference, at least in voters’ minds, without
conceding too much from the elites’ side of the issue.
It may also be that civic integration produces dual outcomes: mandatory
integration could have no effect on individual-level integration but a significant
effect on status acquisition. Requiring language proficiency and host society
knowledge can work one of two ways: it can incentivise participation or it can
penalise noncompliance. Both processes yield the same outcome of performative
improvement. As a result, any change in measures of political, economic or social
integration is confirming evidence of the utility of civic integration policies to achieve
integration goals. This question, therefore, generates the central test of this paper:
whether civic integration policies appear to have an effect on (third-order) immigrant
integration. After testing the basic relationship with integration (holding aside the
question of whether it has an impact on status regulation), we can refer back to the
civic integration debate on ‘help versus hindrance’ to examine in what specific policy-
areas civic integration seems to help or hurt, namely political, economic and/or social
integration.7
In general, then, aggregate differences on integration outcomes over time or across
states requiring different levels of civic integration should tell us whether such
policies matter, and in which direction. On the latter point, mandatory integration
appearing to undercut or even retard integration would not be surprising, even where
incorporation is the goal. This may especially be the case where applicants for
citizenship are already integrated. Furthermore, from the perspective of the state, a
negative effect of policy on integration outcome may not be seen as failure per se but,
instead, indicative of an ulterior motive to policy other than integration itself
(primarily control over status).

Data and Measures

Our primary data-set is the six waves of the European Social survey (2002–2012),
with the recently released sixth wave merged into the off-the-shelf ESS Cumulative
 8 S.W. Goodman & M. Wright
file that incorporates waves 1–5. This survey contains several relevant social, political
and economic integration outcome measures, including variables on: political
integration (political interest and subjective political efficacy), economic integration
(subjective financial well-being and recent unemployment) and social integration
(generalised trust and perceived discrimination).8 And, despite its origin as a general
purpose survey not specifically targeted to generating large immigrant samples, it
contains a relatively large number of foreign-born respondents in each country and
over time.
In order to examine third-order effects of a civic integration policy ‘treatment’ on
immigrants, we use the Civic Integration Policy Index (CIVIX) (Goodman 2014).
This is an additive scale designed to record the number of membership requirements,
spanning both pre-entry (at the stage of immigration) and post-entry (requirements
for permanent residence and citizenship) criteria. Scoring is for the ‘most-difficult’
Downloaded by [The UC Irvine Libraries] at 10:31 09 June 2015

scenario for an ordinary, third-country national migrant completing all three stages.9
To summarise coding rules: first, obligatory civic requirements at entry, settlement
and citizenship receive one point per criterion. Examples include and range from the
‘civic integration test from abroad’ in the Netherlands to civic orientation courses in
Germany to the Austrian citizenship test.10 Second, there are compounding and
ameliorating factors that can make a requirement more or less arduous, respectively:
quarter-point weights are added for increasing levels of language assessment or if a
significant financial cost is incurred by the individual while deductions are factored in
where, for example, a course is recommended but not required or completion of a
requirement for permanent residence ‘double counts’ for citizenship. These simple
scoring rules yield a representative and parsimonious policy snapshots of the original
‘EU-15’ countries and in four snapshot years (1997, 2004, 2009, 2013), for a total of
60 observations. States have experienced sizable policy change over this time period,
but we can ascribe a separation between ‘high’ and ‘low’ CIVIX-practicing countries
(see dashed line in Figure 2).11 In the data analysis, these policy snapshots are
matched to their closest survey wave. Combining our ESS sample with CIVIX data
yields 130 discrete country-years for analysis.
Examining the effects of civic integration presents some unique challenges for
survey analysis. Since it is a recent requirement for recent immigrants,12 the field of
potentially susceptible immigrants has to be reduced. We draw a cut point to include
only immigrants with residence of 10 years or fewer. This choice constricts
immigrant sample size rather dramatically, since only a minority of immigrant
respondents claims to have arrived within that span.13 Moreover, it presents a
potential problem in earlier waves, where policy only gets adopted in most cases
between 2003 and 2007. However, this is a conservative cut of the eligible migrants
for later waves, where 10 years would be around the average time in which an
immigrant would apply for either residence or citizenship.
Indeed, and despite these potential drawbacks, the ESS presents a sufficient sample
of immigrants to examine third-order integration achievements, namely social,
economic and political. By design, most immigrants are susceptible to civic
 Journal of Ethnic and Migration Studies 9
Downloaded by [The UC Irvine Libraries] at 10:31 09 June 2015

Figure 2. Civic integration policy by country and year.

integration conditions. Civic integration requirements are most pervasive for third-
country nationals (immigrants born outside the Member States of the European
Union), though second-country nationals (immigrants born inside the EU and
relocating to another EU state) are required to complete these requirements for
citizenship. Thus, while limitations in the ESS make it impossible to rigorously
ascertain a given immigrant’s ‘treatment status’ with respect to civic integration
policy, we can make sufficiently educated guesses by only considering those who
arrived during the period where such policies came into vogue.
There is, of course, also the issue of sampling bias, or in other words the fact that
immigrants responding to the ESS are more likely to be ‘integrated’ (in the many
senses discussed here) than a country’s immigrant population writ large because the
questionnaire is conducted in the native language. That said, it is not altogether clear
what to make of this bias for our specific purposes here, which is to ascertain the
apparent effect of policy on attitudes in the domain of third-order integration.
Countries scoring highly on civic integration policy might be expected to produce a
disproportionate number of immigrants able and willing to answer a survey (versus
countries scoring low), and that this is somehow biasing upwards the apparent effect
on our dependent variables. But there are additional considerations: first, there is the
possibility that this bias works in the opposite direction, if the imbalance across policy
regime favours immigrants who are able to answer a survey but feel otherwise
stigmatised by the integration requirements. More broadly, however, it is also the case
that if non-response is systematically lower in high civic integration countries—which
is the minimum condition for bias of the kind that would upset our comparisons here
—this would actually strengthen our argument that these policies actually matter.
Moreover, the short time horizon since policy implementation also requires looking
 10 S.W. Goodman & M. Wright
at more proficient immigrants as they are unimpeded by first-order barriers, in other
words, the sample provided in the ESS. In sum, our approach demands caution in a
few respects, but we believe the temporal and spatial breadth of the data, which no
immigrant-specific survey can match, is a worthwhile trade.

Analytical Approach
Integration is to a substantial degree the product of individual-level factors related to
demographics (such as age, race and gender) as well as ‘human capital’-related
determinants (such as education) (Almond and Verba 1963). There is some evidence
that some of these are weaker predictors of political behaviour for immigrants (de
Rooij 2012) but the direction of measured effects remains similar to more general
Downloaded by [The UC Irvine Libraries] at 10:31 09 June 2015

models. Because socio-economic characteristics matter, the unique demographic

profile of each nation’s immigrant population complicates cross-national compar-
isons. Put another way, countries vary based in terms of their immigrant populations’
average level of education, country of origin, racial balance, age and length of
residence. These average differences may lie at the root of any apparent policy
differences we observe, thus making the latter spurious. As such, they must be
statistically controlled as far as possible.
We follow Wright and Bloemraad (2012) in estimating ‘regime-level’ effects by
using a variation on the two-step visualisation technique originally suggested by
Bowers and Drake (2005). We compare the intercept values produced when
outcomes are regressed on the predictors in a baseline individual-level model pooled
within each policy regime category (e.g. in high CIVIX countries vs. low CIVIX
countries). All respondents we analyse are immigrants to the country who have
arrived within the last 10 years. The individual-level model includes the following
variables: dummy variables tapping whether or not the respondent was born in an
EU-member state (0 = yes, 1 = no), is a citizen of the country of interview (0 = no, 1
= yes), identifies as a member of an ethnic minority group in the country (0 = yes, 1 =
no), age (5-category, 0 = 16–29), gender (0 = male, 1 = female) and formal education
(in years).
Given the way the individual-level predictors are coded, what this amounts to is
estimating, for each policy regime category, predicted outcomes for a hypothetical
immigrant who is a visible ethnic minority member, was born outside the EU,14 has
been in the country for 10 years or less, lacks citizenship, is less-educated, male and
young.15 Whenever we refer to immigrant ‘intercept’ scores, we refer to these
estimates. Given the political debate over civic integration, we believe that it is
precisely this sort of immigrant who is at the heart of integration concerns.16 Our
estimator in these equations depends upon the nature of the dependent variable:
dichotomous measures employ logistic regression, whereas all other measures (which
have four or more response values) are estimated using Ordinary Least Squares.
An important conceptual and methodological challenge in immigrant integration
research is establishing a reference point (Bloemraad, Korteweg, and Yurdakul 2008,
 Journal of Ethnic and Migration Studies 11
169–170). Different analytical strategies speak to distinct theoretical interests. We
consequently employ two comparisons: one that directly compares predicted values
for our hypothetical immigrant—in other words, levels—and a second that compares
country-level gaps between immigrant and native-born means across our policy
categories. The first comparison offers important parsimony: on the whole, do
immigrants do better in high CIVIX countries than low CIVIX countries? This type
of analysis is useful not only for its simplicity, but also because the outcome measures
we assess can be seen as ‘goods’ in and of themselves. Normatively, higher levels of
social, social and economic engagement are presumably good for immigrants, and
support a view of positive civic integration policy effects.
At the same time, the concept of ‘integration’ connotes movement towards some
attitude or behaviour consonant with the mainstream position in the host country. In
Downloaded by [The UC Irvine Libraries] at 10:31 09 June 2015

other words, examining levels across values of CIVIX does not, in and of itself,
convey whether or not immigrants in fact more ‘integrated’, since there is no
reference point. Comparing immigrant levels to native levels—deriving a series of
comparative gaps—conveys a distance to the value immigrants are putatively
integrating to, and, as such, native scores provide an ‘anchor point’ to discern how
much immigrants differ from the native-born majority. In the large-n ESS, the
outcome variables become the individual-level difference between a given immigrant’s
score on the measure of interest and the overall mean among native-born
respondents in his or her country. Throughout, our reference point in the ‘gap
analyses’ is immigrant scores with respect to the mainstream population within their
country (as opposed to, say, within all countries of a specific policy regime). Thus, we
are considering gaps between immigrants and their native-born counterparts to assess
the ‘effect’ of policy, holding constant what some might call ‘national political
culture’.
Despite efforts made in controlling for individual-level characteristics, on the one
hand, and establishing a valid reference point on the other hand, it remains the case
that our approach leaves open the possibility of contextual-level confounds, including
other policy indicators on the one hand, and broader economic and political factors
on the other hand. With so few country cases, it is difficult to credibly control for
such factors explicitly. That said, we offer two circumstantial arguments in response:
first, the fact that CIVIX stands analytically distinct from most other measures of
policy across domains of immigration is evidence that these are not likely to
spuriously cause any pattern we observe between CIVIX and our outcomes at the
bivariate level (Helbling 2013). Second, CIVIX is by-and-large unrelated to other
‘usual suspects’ at the macro-level, including measures of economic prosperity and
immigrant inflows.17 CIVIX is, however, positively correlated with two macro-
political indicators: the broader Migrant Integration Policy Index (MIPEX) (rxy =
–.46, p = .09) and a measure of extreme right-wing (ERW) party share (rxy = .33, p =
.09). Both of these relationships are straightforward, and neither undercut the
argument we wish to make here. MIPEX includes a whole spectrum of indicators,
including civic integration, and also policies ranging from labour market mobility to
 12 S.W. Goodman & M. Wright
education. It is coded in an inverse direction to CIVIX, where high scores indicate
benchmark or ‘best practice’ outcomes for immigrant integration. However, by
employing normative instead of descriptive coding rules, the relationship captured
does not convey a story about policy effects but about policy expectations (Goodman
2014, 57–59). With respect to ERW party share, we underline that the growth of
viable far-right parties has catalysed CIVIX policy in many countries, and to the
extent that this is the case any relationship we observe between the latter and
integration outcomes would not be spurious to the former per se but rather a factor
further back in the causal chain. To the extent that there is direct relationship
between ERW party strength and immigrant attitudes, we would anticipate it to be a
largely alienating one, since immigrants are likely to feel less integrated in a polity
that gives voice to anti-immigrant sentiment. What this means for us, in short, is that,
should we turn up a negative influence of CIVIX on outcomes we should worry about
Downloaded by [The UC Irvine Libraries] at 10:31 09 June 2015

a (potentially) spurious relationship with respect to right-wing party share. But,

should the relationship be null or positive, the driving force of immigrant attitudes is
unlikely to be right wing parties.

Data Analysis and Discussion

Political Integration
We examine three measures of political integration: political interest, perception of
politics as complicated or not, and difficulty in making political decisions. These three
measures all tap into the educative overtones of civic integration courses and tests:
creating an awareness of and capacity to participate in politics. We exclude political
participation for two reasons. First, it is too tied to status acquisition to convey
anything distinct about political integration. Second, as civic integration is meant
to enable participation over time, it may still be too early to discern effects of
integration programmes and tests on participation, where the sample includes
immigrant with residence of 10 years or less.18 The same argument can be made for
measures of trust.
Figure 3 presents the mean intercept value of political interest. In the top panel, we
depict immigrant scores in absolute levels on the outcome measure; in the bottom
panel, we depict immigrant scores relative to native respondents’ mean scores within
a country. The difference between the top and bottom panels lies in the outcome
measure: in the former case, it is individual-level immigrant responses on the measure
of interest; in the latter, it is the individual-level difference between a given
immigrant’s expressed attitude and the native-born mean in his or her adoptive
country. In this example, this amounts to the difference between asking ‘how
interested in politics is a hypothetical immigrant in Germany’ and ‘how interested in
politics is the same hypothetical immigrant in Germany compared to the average
native-born Germany?’
In the top panel, we see meaningful significance in interest in politics increase
among respondents in high CIVIX countries. This distinction goes away, however,
 Journal of Ethnic and Migration Studies 13
Downloaded by [The UC Irvine Libraries] at 10:31 09 June 2015

Figure 3. Political interest by CIVIX (updated).

Notes: Point estimates and confidence intervals obtained from regressing (OLS) the
dependent variable—either respondent’s score or the difference between respondent’s
score and native mean—on a set of individual-level predictors (age, ethnic minority
status, citizenship status, birth inside or outside EU, gender and education) and taking the
intercept. Sample is restricted to immigrants with 10 years or fewer of residence in ESS
country. Survey data are from the European Social Survey, Rounds 1–6.

when looking at gaps with natives in the bottom panel. This is not surprising, as
interest may be tied to perceptions of opportunity to participate (that may only come
with status acquisition). Still, increased levels of interest are consistent with the
expectation that exposure to information on the national political system and rules
increase interest in it.
Figure 4 examines two other measures that tap into political integration: whether
immigrants find politics complicated (top graph) and whether immigrants have
difficulty in making political decisions (bottom graph). Again, the idea here is that
integration requirements are, at least on paper, designed to help acquaint immigrants
with their adoptive nation’s political life; as such, it makes sense to ask whether or not
such requirements have any influence in clarifying the political landscape. The results
are basically the same across measures, and seem to affirm this expectation (albeit
modestly). In the top graph, the dependent variable is scored from 0 (‘not
complicated’) to 1 (‘complicated’). In the bottom graph, the dependent variable is
scored such that higher scores mean the respondent has an easier time making
political decisions. In that light, we see immigrants are much more ‘decisive’ about
politics in high CIVIX countries. Both cases, immigrants in high CIVIX countries do
substantially better in the absolute sense (e.g. they are less ‘confused’ and have less
difficulty making political decisions), and the non-overlapping confidence intervals
suggest a statistically significant gap at better than p < .05. And, in both cases, shifting
emphasis to gaps rather than levels sands the statistical edge off of the relationship
while leaving the direction of the gap intact.
 14 S.W. Goodman & M. Wright
Downloaded by [The UC Irvine Libraries] at 10:31 09 June 2015

Figure 4. Subjective political efficacy by CIVIX (updated).

Notes: Point estimates and confidence intervals obtained from regressing (OLS) the
dependent variable—either respondent’s score or the difference between respondent’s
score and native mean—on a set of individual-level predictors (age, ethnic minority
status, citizenship status, birth inside or outside EU, gender and education) and taking the
intercept. Sample is restricted to immigrants with 10 years or fewer of residence in ESS
country. Survey data are from the European Social Survey, Rounds 1–6.

The presumed individual-level mechanisms behind both of these finds are learning
and socialisation, where immigrants sit civic orientation classes and obtain
information on institutions and rules, find politics less complicated and may also
imbue them with confidence which increases ease in political decision-making.

Economic Integration
As one of the stated and central aims of civic integration is to better-prepare
immigrants for the labour market, we examine two measures: subjective financial
well-being, tapped by a single item assessing respondents’ feeling about how they are
getting by on their present household income (from 0 = ‘very difficult’ to 3 = ‘very
comfortable’), which is useful as a holistic indicator of whether or not a given
immigrant sees him or herself as ‘getting by’.19 We also assess whether an immigrant
reports being unemployed (either looking for work or not presently looking for work)
in the last seven days, scored 0 = No and 1 = Yes. Results for each are presented in
Figure 5, below. In approaching the latter question, where the response scale is
dichotomous, instead of showing levels and gaps we present logistic regression
intercept in the top panel, and a predicted probability that the respondent will report
unemployment for each value of CIVIX), with individual-level predictors in the
model held at zero.
 Journal of Ethnic and Migration Studies 15
Downloaded by [The UC Irvine Libraries] at 10:31 09 June 2015

Figure 5. Economic integration by CIVIX (updated).

Notes: Point estimates and confidence intervals obtained from regressing the dependent
variable on a set of individual-level predictors (age, ethnic minority status, citizenship
status, inside or outside EU, gender and education) and taking the intercept. For
subjective well-being, the estimator is OLS; for unemployment status, it is logistic
regression. Subjective well-being panel presents two separate dependent variables: levels
and gaps with native-born. Unemployment panel presents logistic regression coefficient
(intercept) and predicted probability of unemployment where all individual level variables
are scored ‘0’. Sample is restricted to immigrants with 10 years or fewer of residence in
ESS country. Survey data are from the European Social Survey, Rounds 1–6.

The findings here are less encouraging than political integration as we observe only
minor differences and generally insignificant effects in subjective financial well-being.
We see immigrants in high CIVIX countries both more likely to express satisfaction
with their economic well-being, and less likely to report being unemployed. That said,
these observations are tempered by large confidence intervals.20 As such, we cannot
infer, at least on the basis of these data, that mandatory integration has the desired
effect on labour market mobility that policy-makers putatively seek. This reflects a
number of realities. First, there is pre-existing stratification of skill-level of
immigrants in which some may benefit from skills acquisition while others see
language and civic training as time-consuming and unnecessary. Part of this division
is even reflected in policy design, where migrants can pass a language test instead of
enrolling in language classes if they can demonstrate proficiency at a certain level. Or,
in the Netherlands, for example, high-skilled migrants are exempt from the pre-entry
test under the presumption they are already prepared to enter the labour market
without the need to speak Dutch.21 In these cases, migrants would not be ‘assisted’
with integration requirements. In these cases, requirements verify instead of enhance
skills.
 16 S.W. Goodman & M. Wright
Social Integration
A final dimension examined here is social integration. This includes immigrants’
levels of social trust, indexed from 0 = Least to 10 = Most and indexed based on the
ESS’ standard ‘Rosenberg’ items, and perceived discrimination, a dichotomous
measure scored ‘1’ if respondent reports that he or she has been the victim of
discrimination along ethnic, national, linguistic or religious lines. Both tap into the
aspiration that civic integration—through the attainment of skills and knowledge—
empower migrants to feel more connected and able to participate in the common
bonds of the national political community. Both are plotted on Figure 6, below; as
with Figure 4, CIVIX-related differences on the dichotomous ‘perceived discrimina-
tion’ measure are shown via logistic regression coefficient and predicted probability,
rather than levels and gaps.
Downloaded by [The UC Irvine Libraries] at 10:31 09 June 2015

This last sphere of integration yields a similar impression as economic integration.

We do not observe any statistical significance for the of civic integration policies in
fostering trust, or minimising perceptions of discrimination. In this case, the absence
of differences between high and low CIVIX countries suggests support for the null
hypothesis: that civic integration policy has no effect on immigrant social (as well as

Figure 6. Generalised trust and perceived discrimination by CIVIX (updated).

Notes: Point estimates and confidence intervals obtained from regressing the dependent
variable on a set of individual-level predictors (age, ethnic minority status, citizenship
status, inside or outside EU, gender and education) and taking the intercept. For
subjective well-being, the estimator is OLS; for unemployment status, it is logistic
regression. Generalised trust panel presents two separate dependent variables: levels and
gaps with native-born. Perceived discrimination panel presents logistic regression
coefficient (intercept) and predicted probability of unemployment where all individual
level variables are scored ‘0’. Sample is restricted to immigrants with 10 years or fewer of
residence in ESS country. Survey data are from the European Social Survey, Rounds 1–6.
 Journal of Ethnic and Migration Studies 17
economic) integration. This is an important finding. As Perchinig (2010) reports in
his interviews with immigrants in an integration course in Austria,

The question about ‘feeling part of society’ was rejected or not well-received by many
respondents. Most of them answered that feeling part of society was not so much a
question of knowing Germany, but of a tolerant and welcoming atmosphere, which
many of them did not find in Vienna.

Similarly, in Denmark, ‘At least 50 per cent [of immigrant respondents] state that
they have integrated as a result of factors other than the requirements. They point to
their family, friends and work as factors that help them integrate’ (Ersbøll and
Gravesen 2010, 41).
One reason requirements do not promote social integration may be by design.
Downloaded by [The UC Irvine Libraries] at 10:31 09 June 2015

Social integration comes through interaction and the social aspect can be an
important motivator for course participation. However, a significant number of
immigrants qualify out of integration courses. Until 2013, the UK offered a dual-
route to fulfilling the ‘Life in the UK requirement’ for settlement and citizenship
where applicants could either complete an exam or take an English-speaker of other
languages class with civic content. Applicants for permanent residence overwhel-
mingly opted for the test route instead of the course (in 2010, 81,688 took the test vs.
17,607 for the course).22 In Germany, the figures are closer together (in 2011, 92,547
took the language exam while 73,647 finished the integration course) (BAMF 2012,
677). Still, this is over half of the applicant pool that is not experiencing the necessary
socialisation and integration experience that civic integration courses and test
preparation are designed to foster. And, among those that do take a course, a Danish
survey of municipality providers concluded that ‘the social value attributed to the
examination is absent’ (Ersbøll and Gravesen 2010, 99).
It is also possible that civic integration does not show a difference in social
integration because it washes out the different start points that first-generation
immigrants bring to a country. For example, Rahsaan Maxwell (2010) argues that
unlike native and second-generation immigrants, who have been socialised in the
host society, ‘first-generation migrants, who have gone through the disruptive process
of changing countries, will have lower expectations’. As the immigrants examined in
this study are all first-generation migrants who have resided in the host society for 10
years or less (in order to overlap with the time in which civic integration policies were
put into place), effects may be muted or premature.

Conclusion
Civic integration is quickly becoming a fixture in the immigrant experience in
Europe. Modeled off of the USA, European nation-states have gone above and
beyond relatively simple citizenship tests to design comprehensive integration
courses, exams and contract for immigrants not merely seeking citizenship but,
more crucially, seeking permanent residence. In practice, this has mandated a dense
 18 S.W. Goodman & M. Wright
integration experience for newcomers. What has been the effect of this massive
investment? In this paper, we assessed the effects of civic integration policy on
immigrants’ socio-political and economic integration, considering the extent to which
dense versus minimal requirements impact integration outcomes that range from
political awareness to economic well-being. We compared absolute levels of
immigrants’ scores on measures across policy configurations and also compared
immigrant/native-born gaps, in the event that policy differences and requirement
difficulties are a function of the ‘national culture’ into which immigrants are
integrating.
The findings are revealing. We observe little evidence that immigrant integration is
impacted by civic integration, either positively or negatively. Leaving aside methodo-
logical issues for a moment, this raises issues of policy design, where the stated,
Downloaded by [The UC Irvine Libraries] at 10:31 09 June 2015

central purpose in integration requirements is to better prepare the migrant for the
labour market and society. One dimension where we do see an effect of civic
integration requirements is in political integration, where a type of awareness and
participatory confidence increases in high CIVIX countries. Presumably, as civic
integration material is the first time an immigrant might sit down and consider
political institutions and history in a broad sense, this bump is consistent with policy
design. Academic literature tends to downplay null findings, but in this case the
remainder of our findings carries important theoretical and practical significance:
mandatory requirements may not help but they do not appear to hurt the immigrant
in terms of individual integration. And, by testing and ruling out the impact of
requirements on socio-economic integration, we are left to conclude that civic
integration is more politically strategic and rhetorically popular (particularly with a
public hostile to immigration) than it is functional and effective as an integration
policy. In fact, as national qualitative interview data overwhelmingly shows,
requirements are complicated, bureaucratic, arbitrary, and oftentimes time-consum-
ing (Strik et al. 2010).
While we do not observe high or low civic integration policies dramatically altering
an immigrant’s attitudes or behaviour here, though we cannot discount the possibility
that requirements have an impact over a longer duration of time, evidence elsewhere
shows a more immediate effect as a gatekeeper to status. The Dutch civic integration
from abroad exam, for example, show pass rates impacting entry. In 2012, the overall
test pass rate was 79% (Significant 2010, 23), though disaggregation of this figure
reveals a greater impact on older and less educated migrants. Civic integration
requirements also limit access to permanent residence, but the data can be misleading
here. Very few immigrants actually get denied permanent residence (or subsequently
deported) because of a failure to fulfil integration requirements. Instead, immigrants
are kept in a cycle of having to complete more and more language training in order to
reach an arbitrary level or pass an integration exam or, perhaps as a consequence, are
disincentivised from applying in the first place. In Germany, there was a 67.6% pass
rate for completing the integration course and language requirement among course
graduates (BAMF 2012, 677). In Austria, only 53.89% among those required to fulfil
 Journal of Ethnic and Migration Studies 19
the integration agreement in 2010, and this was the highest completion rate since the
agreements first implementation in 2003 (Permoser 2012, 193). These immigrants are
not deported, but placed in a feedback loop of conditionality. By contrast, in
Denmark we see a rather decisive gatekeeping effect of language on naturalisation,
where 40% of applicants in 2008 were refused citizenship on the grounds of Danish
language insufficiency (Ersbøll and Gravesen 2010, 33). In the UK, that number is
only 2%, though the test pass rate is at an average of 68.6%. There is variation in the
degree to which states use civic integration policies as a gatekeeper, and in
particularly whether that gate is entry, permanent residence or citizenship. But, if
we assume policy efficiency, there is no doubt that civic integration largely serves a
control function, as the evidence does not suggest it serves much of an integrative one
for the time being.
Downloaded by [The UC Irvine Libraries] at 10:31 09 June 2015

This leads to further issues. Methodologically, one issue to consider is the how the
extent of civic integration policies themselves might condition the type and number
of immigrants arriving. Conceptually, research shows these types of policies to be
more onerous for populations already defined as vulnerable, including immigrants
from less-developed countries, women, refugees, etc (Van Oers 2013). If true, such
immigrants may be tempted to go elsewhere, and civic integration policy might
ultimately ‘work’ through selection rather integration. There is no way to examine
that at the present time, and with the data at hand, but we leave this question as a
signpost for future research. This creates two unique but related problems: those who
do not need integration find mandatory requirements tedious and potentially
offensive while those who need assistance may find it insufficient or, worse, an
insurmountable barrier to secure status and, in some cases, benefits assistance (e.g.
the UK). In other words, civic integration policies appear to do little to integrate for
the vast majority of participants. It may be useful for future, qualitative case work to
examine integration effects given variation in skill-level (particularly to discern
between first- and third-order integration goals) and status, where expectation of
requirements on highly skilled workers may necessary diverge from the unskilled. For
example, research on the effects of municipal-level integration courses have shown
promising results on refugees in particular (Ministry of Immigration 2009).
Civic integration policies show no signs of letting up. If anything, existing policies
are becoming more entrenched and more states, from Europe to Australia to the
USA, are revisiting possibilities of and strategies for membership promotion. Not
only are states revisiting the idea of membership, but also they are reinventing their
role as states in that experience. Mandatory courses and tests—even when contracted
out to private providers—bring the state ever-closer to the lives of immigrants. As
such, increased requirements at increased membership gates create more ‘check-
points’. It is then, perhaps, unsurprising that civic integration policy’s most visible
function is as gatekeeper.
Therefore, the final issue at hand is asking what role the state can and should play
in promoting autonomy-enabling skills in an immigrant. Where is the line between
support and force? Whose interests are being served by this new paternalism, given
 20 S.W. Goodman & M. Wright
the results presented here? The immigrant? The state? Perhaps a wary, xenoskeptic
public? And does the state have the right to reposition itself in the name of
integration as a matter of national interest? There are no clear answers here. These
kinds of questions raise a number of interesting theoretical considerations for
scholars to take up, examining not just the limits of liberalism (and when it crosses
into ‘illiberal liberalism’ and whether this is normatively acceptable) but also state
prerogatives of control, and to what extent this can evolve under changing
circumstances. These lines of questioning, much like examining processes of
immigrant integration, draw this policy story outside of the European context and
towards a whole host of other cases, from advanced democracies that are forever
consolidating to achieve more inclusivity and participation to democracies in
transition, which are only beginning to ask questions about how to appropriately
manage issues of diversity.
Downloaded by [The UC Irvine Libraries] at 10:31 09 June 2015

Disclosure statement
No potential conflict of interest was reported by the authors.

Notes
[1] See Entzinger (2003); Joppke (2007, 2010); Also Vertovec and Wessendorf (2010); Health
and Demireva (2013).
[2] In October 2010, Angela Merkel declared a multicultural approach in Germany had ‘utterly
failed’. See ‘Merkel erklärt “Multikulti” für gescheitert’, Deutsche Welle, October 16, 2010.
In February 2011, Nicolas Sarkozy declared similar sentiments in a nationally televised
debate, arguing that such a concept enabled extremism. See ‘Nicolas Sarkozy declares
multiculturalism had failed’ France24, February 11, 2011.
[3] On this, see Banting and Kymlicka (2013); Koopmans (2010); Wright and Bloem-
raad (2012).
[4] E.g. Koopmans et al. (2012) Indicators of Citizenship Rights for Immigrants scores cultural
requirements similarly to other restrictive policies of citizenship, like residency. Howard’ s
(2009) Citizenship Policy Index, assigns a policy ‘weight’, akin to a restrictive penalty.
[5] Case studies based on interviews with a limited number of migrants reveal only marginal
integration benefits from integration courses and tests, recognising that time, stress and
money countervail added benefits (Strik et al. 2010).
[6] This would involve measuring the degree immigrants are accepted by natives, or the extent
to which there is improvement in migrant perception, after civic integration policy
adoption. Testing this requires a longer timeline than the scope of this paper provides,
though support for this hypothesis would not undermine present testing as it looks not at
immigrant behaviour but public opinion. Moreover, more than one policy rationale can
coexist.
[7] We exclude cultural integration, as it is explicitly excluded from civic integration policy
objectives. The question of whether this is the case or not in practice is important (i.e.
whether civic integration operates like assimilation), but one to be addressed elsewhere.
[8] For specific coding, see Appendix.
[9] Minors and spouses experience different permutations of requirements.
[10] Requirements can take different forms in the same category: language proficiency at
citizenship can be demonstrated through certification, testing, obtaining a waiver based on
 Journal of Ethnic and Migration Studies 21
education or completing an interview. Coding these instruments equally avoids interjecting
subjectivity in the scoring, for what is easy for one migrant, say, certification of language,
might be impossible for another.
[11] This cut-point is 2.5, where countries at or below this score are considered ‘low CIVIX
countries’. This is consistent with qualitative evidence on policy design, where these states
assess language but through traditional and subjective interviews and do not use the
language of civic integration in policy-making. See Goodman (2014).
[12] An exception is the Dutch model, which requires ‘oudkomers’ (settled immigrants) to pass
the integration exam for continued residence.
[13] Among country-years for which we have CIVIX data, the exact figure is 4482 out of 13,467,
or 33%.
[14] EU nationals automatically acquire the right of permanence residence in another EU
country if they have lived there for at least five years continuously. As a result, we include a
dummy variable scored ‘0’ for those born outside the EU and ‘1’ for those born inside.
In our ESS sample, ‘Non-Minority’ is a dummy indicating that the respondent does not
Downloaded by [The UC Irvine Libraries] at 10:31 09 June 2015

[15]
claim to be a member of an ethnic minority in the country. ‘Citizenship’ is a dummy
indicating that the respondent has citizenship in the host country. Age and formal
education are both measured as years.
[16] Public opinion surveys capture only the most integrated people, minimising the kind of
variance reflective of actual course and test participation. Still, ESS respondents represent
‘typical’ cases for testing third-order effects.
[17] Data on file with authors. Available upon request.
[18] Furthermore, it should be said that the goal of civic integration is not necessarily that
people participate, but rather that they have the skills to do so (i.e. act autonomously).
[19] We choose this measure as it provides context for interpreting well-being, unlike a strict
measure of income.
[20] The more relevant figure in the unemployment analysis is the confidence interval in the top
panel, e.g. around the logistic regression estimate for the intercept. The tiny confidence
intervals around the predicted probability reflect low simulation error, not low estimation
error.
[21] The entry exam also does not apply to Turkish nationals and family members (since 2011),
migrants with paid employment (except for spiritual leaders), migrants seeking temporary
residence (e.g. students) or migrants with educational qualifications (e.g. Dutch language
diploma).
[22] FOI #20784. 13 December 2011. On file with authors.

References
Almond, Gabriel, and Sidney Verba. 1963. The Civic Culture: Political Attitudes and Democracy in
Five Nations. [In English]. Princeton, NJ: Princeton University Press.
BAMF. July 2012. 9. Lagebericht Der Beauftragten Der Bundesregierung Für Migration, Flüchtlinge
Und Integration Über Die Lage Der Ausländerinnen Und Ausländer in Deutschland [Report
of the Federal Government Commissioner for Migration, Refugees and Integration on the
Situation of Foreigners in Germany]. BAMF. Berlin.
Banting, Keith, and Will Kymlicka. 2013. “Is There Really a Retreat from Multiculturalism Policies?
New Evidence from the Multiculturalism Policy Index.” Comparative European Politics 11
(5): 577–598. doi:10.1057/cep.2013.12.
Bloemraad, Irene, Anna Korteweg, and Gokce Yurdakul. 2008. “Citizenship and Immigration:
Multiculturalism, Assimilation, and Challenges to the Nation-state.” Annual Review of
Sociology 34 (1): 153–179. doi:10.1146/annurev.soc.34.040507.134608.
 22 S.W. Goodman & M. Wright
Böcker, Anita, and Tineke Strik. 2011. “Language and Knowledge Tests for Permanent Residence
Rights: Help or Hindrance for Integration?” European Journal of Migration and Law 13:
157–184.
Bowers, Jake, and Katherine W. Drake. 2005. “Eda for Hlm: Visualization When Probabilistic
Inference Fails.” Political Analysis 13 (4): 301–326.
Council of Europe. 2008. White Paper on Intercultural Dialogue: Living Together as Equals in Dignity.
Strasbourg: Council of Europe. http://www.coe.int/t/dg4/intercultural/source/white%20paper_
final_revised_en.pdf
de Rooij, Eline A. 2012. “Patterns of Immigrant Political Participation: Explaining Differences in
Types of Political Participation between Immigrants and the Majority Population in Western
Europe.” European Sociological Review 28, (4): 455–481. doi:10.1093/esr/jcr010.
Entzinger, Han. 2003. “The Rise and Fall of Multiculturalism: The Case of the Netherlands.” In
Toward Assimilation and Citizenship: Immigrants in Liberal Nation-States, edited by
Christian Joppke and Ewa T. Morawska, 59–86. New York: Palgrave Macmillan.
Ersbøll, Eva, and Laura Katrine Gravesen. 2010. “Country Report Denmark.” In The INTEC Project:
Downloaded by [The UC Irvine Libraries] at 10:31 09 June 2015

Integration and Naturalisation Tests: The New Way to European Citizenship. Nijmegen:
Centre for Migration Law, Radboud University.
Freeman, Gary P. 1995. “Modes of Immigration Politics in Liberal Democratic States.” International
Migration Review 29 (4): 881–902. doi:10.2307/2547729.
Gelderloos, Wout, and Jessica van Koert. 2010. Inburgeren En Participeren [Integration and
Participation]. Verslag Van Vier Kwalitatieve Casestudies Naar De Effecten Van Inburgering
Op Participatie [Report of Four Qualitative Case Studies into the Effects of Integration on
Participation]. Den Haag: B&A Consulting.
Goodman, Sara Wallace. 2013. “‘Concluding Remarks: What We Can Learn from Citizenship
Policy Indices,’ in ‘How to Measure Immigration Policies.’” APSA Newsletter for Migration
and Citizenship Section, Issue 2 (Summer): 48–54.
Goodman, Sara Wallace. 2014. Immigration and Membership Politics in Western Europe. New York:
Cambridge University Press.
Groenendijk, Kees. 2011. “Pre-departure Integration Strategies in the European Union: Integration
or Immigration Policy?” European Journal of Migration and Law 13 (1): 1–30. doi:10.1163/
157181611X558191.
Guild, Elspeth, Kees Groenendijk, and Sergio Carrera, eds. 2009. Illiberal Liberal States:
Immigration, Citizenship, and Integration in the Eu. Farnham: Ashgate.
Heath, Anthony, and Neli Demireva. “Has Multiculturalism Failed in Britain?” Ethnic and Racial
Studies 37: 161–180.
Helbling, Marc. 2013. “Validating Integration and Citizenship Policy Indices.” Comparative
European Politics 11: 555–576.
Helbling, Marc, Tim Reeskens, Cameron Stark, Dietlind Stolle, and Matthew Wright. 2015.
“Enabling Immigrants to Participate: Redirecting Our Attention to the Role of Integration
Regimes.” In The Political Immigrant: A Comparative Perspective, edited by Antoine
Bilodeau. Toronto: University of Toronto Press.
Howard, Marc Morjé. 2009. The Politics of Citizenship in Europe. New York: Cambridge University
Press.
Janoski, Thomas. 2010. The Ironies of Citizenship. Naturalization and Integration in Industrialized
Countries. New York: Cambridge University Press.
Joppke, Christian. 2007. “Transformation of Immigrant Integration: Civic Integration and
Antidiscrimination in the Netherlands, France, and Germany.” World Politics 59 (2): 243–
273. doi:10.1353/wp.2007.0022.
Joppke, Christian. 2010. Citizenship and Immigration (Immigration & Society series). Cambridge;
Malden, MA: Polity.
 Journal of Ethnic and Migration Studies 23
Joppke, Christian, and Ewa T. Morawska. 2003. Toward Assimilation and Citizenship: Immigrants
in Liberal Nation-states. Migration, Minorities, and Citizenship. New York: Palgrave
Macmillan.
Klaver, Jeanine, and A. W. M. Odé. 2009. Civic Integration and Modern Citizenship: The
Netherlands in Perspective. Groningen: Europa Law Publishing.
Koopmans, Ruud. 2010. “Trade-offs between Equality and Difference: Immigrant Integration,
Multiculturalism and the Welfare State in Cross-national Perspective.” Journal of Ethnic and
Migration Studies 36 (1): 1–26. doi:10.1080/13691830903250881.
Koopmans, Ruud, Ines Michalowski, and Stine Waibel. 2012. “Citizenship Rights for Immigrants:
National Political Processes and Cross-national Convergence in Western Europe, 1980–
2008.” American Journal of Sociology 117 (4): 1202–1245. doi:10.1086/662707.
Kostakopoulou, D. 2010. “Matters of Control: Integration Tests, Naturalisation Reform and
Probationary Citizenship in the United Kingdom.” Journal of Ethnic and Migration Studies
36 (5): 829–846.
Maxwell, Rahsaan. 2010. “Political Participation in France among Non-European-origin Migrants:
Downloaded by [The UC Irvine Libraries] at 10:31 09 June 2015

Segregation or Integration?” Journal of Ethnic and Migration Studies 36 (3): 425–443.

Ministry of Immigration. 2009. Effectiveness Measurement of Danish Municipalities Integration
Policies from 1999 to 2007. Copenhagen: Edited by Refugee and Integration Affairs
(Denmark) The Ministry for Immigration.
Mouritsen, Per. 2008. “Political Responses to Cultural Conflict: Reflections on the Ambiguities of
the Civic Turn.” In Constituting Communities. Political Solutions to Cultural Conflict, edited
by Per Mouritsen and Knud Erik Jørgensen, 1–30. London: Palgrave.
Perchinig, Bernhard. 2010. “Country Report Austria.” In The INTEC Project: Integration and
Naturalisation Tests: The New Way to European Citizenship. Nijmegen: Centre for Migration
Law, Radboud University.
Permoser, Julia Mourão. 2012. “Civic Integration as Symbolic Politics: Insights from Austria.”
European Journal of Migration and Law 14 (2): 173–198. doi:10.1163/157181612X642367.
Seveker, Marina, and Anne Walter. 2010. “Country Report Germany.” In The INTEC Project:
Integration and Naturalisation Tests: The New Way to European Citizenship. Nijmegen:
Centre for Migration Law, Radboud University.
Significant. 2010. “Evaluatierapport Inburgering in Nederland [Evaluation Report Integration in the
Netherlands].” Barneveld, the Netherlands: Significant.
Strik, Tineke. 2013. Integration Tests: Helping or Hindering Integration? Strasbourg: Edited by
Council of Europe Parliamentary Assembly.
Strik, Tineke, Anita Böcker, Maaike Luiten, and Ricky van Oers. 2010. “Synthesis Report.” In The
INTEC project: Integration and Naturalisation Tests: The New Way to European Citizenship.
Nijmegen: Centre for Migration Law, Radboud University.
van Oers, Ricky. 2013. Deserving Citizenship: Citizenship Tests in Germany, the Netherlands and the
United Kingdom. Leiden, The Netherlands: Martinus Nijhoff Publishers.
Vertovec, Steven, and Susanne Wessendorf. 2010. The Multiculturalism Backlash: European
Discourses, Policies and Practices. New York, NY: Routledge.
Wright, Matthew, and Irene Bloemraad. 2012. “Is There a Trade-off between Multiculturalism and
Socio-political Integration? Policy Regimes and Immigrant Incorporation in Comparative
Perspective.” Perspectives on Politics 10 (1): 77–95. doi:10.1017/S1537592711004919.

Appendix
All variables listed below are referenced to original measures in downloaded ESS data. For more
detailed wordings, please see documentation available at http://ess.nsd.uib.no/downloadwizard/#.
 24 S.W. Goodman & M. Wright
Outcomes
Political interest: single item (from raw 4-category measure ‘polintr’), recoded from 0 = ‘not at all
interested’ to 1 = ‘very interested’.
Politics too complicated: single item (from raw 5-category measure ‘polcmpl’), recoded from 0 =
‘never’ to 4 = ‘frequently’.
Difficulty making political decisions: single item (from raw 5-category measure ‘poldcs’), recoded
from 0 = ‘very difficult’ to 4 = ‘very easy’.
Subjective economic well-being: single item (from raw 4-category measure ‘ ‘), recoded from 0 =
‘very difficult’ to 4 = ‘very easy’.
Unemployed, last seven days: dummy variable based on a ‘yes’ response to either raw variable
‘uempla’ or ‘uempli’.
Generalised trust: additive index of three 11-point ‘Rosenberg’ items (originally ‘ppltrst’, ‘pplfair’
and ‘pplhlp.’ Index is scored from 0 = least trusting to 1 = most trusting.
Downloaded by [The UC Irvine Libraries] at 10:31 09 June 2015

Perceived discrimination: whether respondent perceives him/herself as part of a minority that has
been discriminated against the country on the basis of either ‘color or race’, ‘nationality’, ‘religion’,
‘language’ or ‘ethnic group’. From raw variables ‘dscrrce’, ‘dscrntn’, ‘dscrrlg’, ‘dscrlng’ and ‘dscretn’,
respectively. The measure employed is coded 1 if the response any of these way ‘yes’ and 0 if the
response to all of them was ‘no’.
Predictors
Citizenship status: R holds [country] citizenship (from raw measure ‘ctzcntr’), recoded 0 = ‘no’ and
1 = ‘yes’.
Length of residence: how long R has lived in country (from raw 5-category measure ‘livecntr’), re-
scored from 0 = ‘within the last year’ to 1 = ‘more than 20 years ago’. Minority status: R identifies
self as belonging to a minority ethnic group in country (from raw measure ‘blgetmg’), recoded 0 =
‘no’ and 1 = ‘yes’.
Gender: (from raw measure ‘gndr’), recoded 0 = ‘male’ and 1 = ‘female’.
Age: (from raw measure ‘age’), recoded into 5-category measure such that 0 = ‘18–29’, .25 = ‘30–39’,
.50 = ‘40–49’, .75 = ‘50–64’ and 1 = ‘65+’.
Education: (from raw measure ‘eduyrs’), simply years of formal education completed.