ABSTRACT
What determines the confidence of immigrants in public institutions? Using pooled data from the European Social Survey of 26 countries, the study examines whether processes of social exclusion and discrimination can account for migrants’ confidence in public institutions. Specifically, it examines the role of the quality of public institutions and of the migrant integration policies of the host country and how they interact with migrant status and proxies for experienced or potential discrimination in shaping institutional trust. Results show that the performance of public institutions matters less for the confidence of immigrants than that of natives, in particular for those who see themselves as an ethnic minority or members of a group that has faced discrimination. Second-generation migrants exhibit less trust than first-generation migrants. However, discriminatory processes appear to be of less importance than the expectations carried from the home country or acculturation processes.
1. Introduction
An increasing part of Europe's population is made up of migrants from diverse backgrounds, and a crucial question is whether European countries can succeed in integrating these immigrants. Studies on such political integration have focussed largely on naturalisation, voting and non-electoral political participation. In this paper, we address a complementary question: How do migrants relate to the social and political institutions in their countries of residence? While much of the existing literature on political integration is driven by concerns that migrants’ ‘divided loyalties’ prevent their integration (Black 1987; Huntington 2004) or that lack of political skills, democratic values and trust in public institutions require ‘political re-socialisation’ (White et al.2008), we examine whether processes of social exclusion and discrimination shape immigrants’ confidence in public institutions.
Recent research has thrown up puzzling and contradictory pieces of evidence on social exclusion of and discrimination against immigrants and their confidence and trust in public institutions. On the one side, there is large concern about discrimination and social exclusion of immigrants and ethnic minorities in Europe. Audit studies, reports about perceived discrimination and statistical analyses of outcomes leave no doubt that immigrants in Europe have a much higher risk of being discriminated against than natives (OECD 2008; Safi 2010). The recent European Union Minorities and Discrimination Survey has underscored the fact that public services and institutions play an important role in the discrimination and social exclusion of immigrants (FRA 2009). However, this does not appear to be reflected in migrants’ confidence in these institutions. National surveys and pan-European studies indicate that migrants have no less confidence in public institutions than natives. Similar findings are reported for the US (Michelson 2003; Wenzel 2006) and Canada (Bilodeau and Nevitte 2003). The ‘optimistic evaluations’ (Maxwell 2010) of migrants whose perceptions are affected by ‘dual frames of reference’ (Röder and Mühlau 2010) are argued to account for this phenomenon, as migrants compare the host country institutions along standards developed in their countries of origin. But is there a strong undercurrent of political disengagement and alienation from the host country's public institutions?
In order to evaluate whether immigrants’ confidence in the host country institutions is shaped by the processes of discrimination and social exclusion, we proceed from the baseline model that confidence in public institutions is based on an evaluation of the quality of public good provisions. Subsequently, we argue that a key assumption of this baseline model, non-exclusiveness of good provisions, may not hold for immigrants. Immigrants and their offspring may not benefit from societal institutions to the same degree as natives due to social exclusion and discrimination, which may affect particular groups of immigrants more than others. The hypotheses derived from this idea are tested using pooled European Social Survey (ESS) data linked to country information regarding quality of governance and immigrant integration policies.
2. Theory and hypotheses
The ‘institutional performance model’ of political trust (Newton and Norris 2000; Mishler and Rose 2005) holds that citizens’ confidence in institutions reflects an evaluation of the economic and political performance of these institutions. According to this view, public institutions promoting economic and social development and guaranteeing public safety, rule of law and individual liberties in a reliable and effective manner positively deserve the confidence of citizens. The evaluation of institutional performance has both instrumental and relational aspects to it. When in direct contact with institutions, citizens read the way they are treated as signals of their valuation as members of the society and procedural fairness concerns are salient in their judgement (Tyler 2006). More generally, an important, although frequently neglected aspect of institutional performance, is their capacity to create a fair and inclusive society. ‘Fairness-generating institutions’ ‘must enshrine concepts such as fairness, justice, incorruptibility, non-partisanship, truthfulness, or even transparency as the core norms of communal living, … they are universally oriented and provide their citizens with equal opportunities’ (Freitag and Bühlmann 2009: 8–9).
Newton and Norris (2000) have emphasised that the goods produced by public institutions have the character of public goods: All citizen groups benefit to the same degree from well-performing institutions and failures of the political systems affect all citizens in a similar fashion. Moreover, an implicit assumption is that these public goods are evaluated according to standards that are shared among the citizens. In line with this argument, Newton and Norris (2000) quote in support of this theory the fact that political trust is largely unrelated to demographic factors such as education, income, age, religion and gender. This public good view of confidence in public institutions suggests that the quality of public institutions is of utmost importance in keeping up the citizens’ confidence in them and that natives and immigrants differ little in this respect.
The public good view of institutional performance, however, raises the question of whether immigrants are not to some degree excluded from the public goods provided by institutions of host countries and whether this, in turn, is reflected in lower levels of confidence in the political institutions. Many immigrants have ‘less than full’ citizenship status and face substantial hurdles for naturalisation. If this group is excluded from the collective good provided by otherwise well-performing societal institutions, they may have good reason to evaluate the performance of institutions less favourably. Moreover, lack of political and economic rights is only one aspect of immigrants’ and minorities’ experiences with host country institutions and their representatives. For example, US research shows that, in general, Black Americans have less trust than Whites in political institutions (Abramson 1983; Howell and Fagan 1988), the police (Frank et al.2006) or the healthcare system (Boulware et al.2003). It is typically assumed that this trust gap reflects the ‘political reality’ (Abramson 1983) characterised by a legacy of underrepresentation, racial discrimination, persistent disadvantage and social exclusion. In Europe, compared to natives, immigrants stand a much higher risk of being discriminated against (Andre et al.2008) with serious consequences for their well-being: Safi (2010) claims that discrimination explains the disparities in life satisfaction between immigrants and natives in Europe.
The European Union Minorities and Discrimination Survey (FRA 2009) has highlighted the degree to which public services and institutions such as the police discriminate against immigrant and minority groups. Such institutional discrimination is likely to affect the confidence in institutions directly: Treatment with honesty and respect by authorities is one important aspect in determining satisfaction with and trust in institutions (Tyler 2006). Most incidents of discrimination involve bad treatment by employers and by the general public (e.g. McGinnity et al.2006). This experience of discrimination by third parties is also likely to affect confidence in public institutions, as immigrants may regard political institutions as performing poorly if they do not prevent these acts of discrimination by third parties (e.g. Kääriäinen 2007).
If these processes are ongoing, immigrants are expected to benefit less from the performance of institutions than natives and also more likely to experience discrimination by the authorities representing the institutions. Hence, variations in the general ‘quality of institutions’ are expected to be less salient for immigrants’ confidence in these institutions:
H1: The relationship between the quality of public institutions and confidence in the institutions is weaker for immigrants than for natives.
Two lines of the social exclusion argument can be distinguished: a formal-legal one that relates social exclusion to the legal status of immigrants, and an informal one that relates social exclusion to processes of discrimination, which does not derive from (but may coincide with) the legal status. The formal-legal line relates to the political, economic and social rights of immigrants. Typically, only citizens of a state have a full set of rights. Within the European Union (EU), the status of citizens of other EU countries is similar (with the exception of full political rights) to that of citizens of the focal state. If ‘less than full citizen status’ is associated with exclusion from the public goods produced by public institutions, we expect that these immigrants will be less responsive to variations in the quality of public institutions:
H2: The relationship between the quality of institutions and confidence in institutions is weaker for immigrants without citizenship status than for immigrants who are citizens of their country of residence, and weaker for immigrants from non-EU countries than for immigrants from EU countries.
The informal line refers to discrimination processes as the driving forces of social exclusion, i.e. unequal treatment, which is not legitimised by unequal rights due to the migrant status of the immigrants. The discrimination argument holds that discrimination is a mechanism that excludes migrants from the public goods produced by public institutions. This holds true for the instrumental aspects of the public goods, such as housing or health services, but more so for the relational aspects, such as fair and equal treatment. Consequently, we expect that immigrants who are discriminated against are less likely to respond to changes in the quality of institutions with increased confidence in these institutions. This implies that the relationship between the quality of institutions and confidence is expected to be weaker for groups that perceive themselves to be discriminated against. However, it is also likely that people who have low trust in public institutions are more likely to attribute negative experiences to discrimination. Further credit for the assumed direction of causality would therefore be gained if the same would hold for members of groups that are at higher risk of being discriminated against, regardless of whether an individual perceives having experienced the discrimination himself or herself. A vast amount of literature indicates that people who differ in their appearance or behaviour from members of the mainstream society are more likely to experience discrimination (e.g. Andre et al.2008). Hence we expect:
H3: The relationship between quality of institutions and confidence is weaker for people who report having experienced group discrimination and for groups that differ from mainstream society with regard to religion, language usage, ethnic minority and non-White background.
Countries differ in the degree to which they have adopted formal policies aiming to enhance the opportunities of immigrants by granting them equal access to services and rights similar to citizens or by easing the road to acquire those rights. These integration policies mainly protect the status and rights of immigrants from outside the EU. Immigrants in countries with strong legal and policy frameworks aiming to aid migrant integration are expected to exhibit more trust and confidence in public institutions than immigrants residing in countries that only weakly subscribe to such policies. As such frameworks may indicate a good policy environment and efficient governance in general, it may be also positively related to the trust of natives, but the relationship should be stronger for immigrants, in particular those from non-EU countries:
H4: The relationship between the quality of immigrant integration policies and confidence in institutions will be positive for immigrants, particularly those from non-EU countries, and stronger than for natives.
If processes of social exclusion and discrimination are the dominant mechanisms shaping confidence in public institutions, it is also expected that immigrants are, similar to ethnic minorities such as Afro-Americans in the US, less confident in and trusting of political and social institutions in general. However, earlier research has indicated that for immigrants, the potential effects of social exclusion and discrimination may be suppressed by their ‘optimistic evaluations’ (Maxwell 2010). This appears to be largely due to their ‘dual frames of reference’ (Röder and Mühlau 2010), but the ‘friendly indifference’ of temporary migrants to the public institutions of their country of residence may also play a role. However, children of immigrants born in the country of residence have little or no contact with the home country of their parents and are likely to take their lead from their peers rather than their parents. Second-generation migrants are, therefore, expected to be far less affected by the quality of institutions in the country of their parents’ origin and how they compare to the quality of the institutions in their country of residence. They are also less likely to see themselves as transient migrants. Consequently, we expect that for second-generation migrants, a lack of confidence induced by social exclusion and discrimination is not suppressed by ‘optimistic evaluations’. And once proper controls are put in place for the expectations of migrants, it is expected that the negative effects of discrimination and social exclusion will become visible. Hence, our final hypothesis is:
H5: Second-generation immigrants exhibit less confidence in public institutions than natives and the confidence of first-generation migrants becomes lower than that of natives once controls for ‘optimistic expectations’ are put in place.
3. Data and methods
Data were extracted from the first three rounds of the ESS and collected between 2002 and 2006. The ESS is particularly useful for a cross-national study because not only does it cover a wide range of countries, but it is also designed to allow cross-national analyses by ensuring that the questions are understood in the same way by respondents in different countries using different languages. It is a high quality dataset, where the sampling design approximates a simple random sample and has a relatively high response rate. However, it is not an immigrant-specific survey. Respondents must be registered in their host country in order to be included in the sample and be able to communicate in the host country language to participate.
In the dataset for this analysis, we included natives, defined as individuals whose parents and who themselves were born in a country; first-generation migrants, defined as respondents whose parents and who themselves were born abroad; and second-generation migrants, who were born in their country of residence, but whose parents were both born abroad. As the first round of the ESS does not include detailed information for parents’ country of origin, we excluded second-generation migrants from this round. We also excluded any cases that did not belong in any of the three categories, such as return migrants or individuals with only one foreign-born parent. These criteria were used to construct the three categories as clearly as possible and exclude any ambiguous cases such as return migrants.
Individuals are seen as nested within their country of residence and their ‘community’. For first-generation migrants and natives, the ‘community’ is defined by their country of origin and their country of residence. For second-generation migrants, the mother's country of origin and country of residence are used to define the community to which the respondent belongs. First- and second-generation migrants were put in separate communities even if their country of origin was identical, which is further explained in the modelling section below. A respondent was excluded if there was no information about the country of origin.
We have used data from 26 of the 29 countries covered by the first three rounds of the ESS. These include 15 countries from the EU and the European Free Trade Association (EFTA) states of Iceland, Norway and Switzerland. Eight new member states of the EU, Bulgaria, Cyprus, Czech Republic, Estonia, Hungary, Poland, Slovenia and Slovakia, are also included. We have excluded Turkey, the Ukraine and Israel from the analysis.
3.1. Measures
3.1.1. Dependent variables
Confidence in public institutions is measured in the ESS with a range of variables on a zero to 10 scale of how much people trust the different public institutions of the surveyed country. Four of these variables are present in all three rounds of the ESS: how much people trust their country's parliament, the legal system, the police and politicians. The reliability of this four-item scale was confirmed with Cronbach's Alpha for individual survey countries at between 0.655 and 0.785, and a sum score was calculated from the four items, with higher scores indicating more trust. Cases with missing values were excluded.
3.1.2. Independent variables
3.1.2.1. Quality of governance
To determine the institutional quality of the host country, we used the Worldwide Governance Indicators (Quality of Governance) published by the World Bank (Kaufmann et al.2009), which covers all the countries in the survey here. This is a composite indicator incorporating six dimensions: voice and accountability, political stability and absence of violence, government effectiveness, regulatory quality, rule of law and control of corruption, all based on a large number of sources. Figures from 2002 to 2005 inclusively were used to compute averages for the six dimensions. A factor score of these indicators was computed based on the values of the host countries included in the dataset. This factor explains 89.69 percent of the variance of the six dimensions.
3.1.2.2. Migrant integration policies
The degree to which countries adopt a legal and policy framework aiming to provide equal opportunities and rights to immigrants was measured by the Migration and Integration Policy Index (MIPEX) developed by the British Council and the Migration Policy Group (Niessen et al.2007). This index compares policies for the integration of migrants with best practices in 28 countries, including the EU member states. It is based on a wider range of indicators that are divided into six domains (long-term residence, access to nationality, anti-discrimination, family reunion, political participation and labour market access). These six domains are highly correlated and unidimensional. We use the factor scores of a principal component analysis of these domains as the variable, MIPEX.
3.1.2.3. Legal status
The legal status of the migrants is measured by two variables: EU status and citizenship. The variable, EU status, has three categories, EU15, New EU and Non-EU, referring to the country of origin for first-generation migrants. Second-generation migrants are categorised as EU15 if at least one parent was born in an EU15 country, as New EU if neither parent was born in an EU15 country and at least one in a new member state country, and Non-EU if both parents were born in a non-EU country. The variable, citizenship, indicates that a migrant is a citizen of the country of residence. The very few non-migrants reporting not being citizens of their country of residence were excluded from the analysis.
3.1.2.4. Discrimination
Group discrimination is measured by several items asking respondents whether they belonged to a group that experienced discrimination on grounds of colour or race, nationality, religion, language, ethnic group membership, age, gender, sexuality, disability or other grounds. It is a binary indicator between those who do not report discrimination and those who report it along at least one of these criteria. This indicator relies on people perceiving and reporting discrimination, which may be influenced by a variety of factors, such as the prominence of discrimination discourse in the host country, political awareness and, indeed, generally low levels of trust in people and institutions. Therefore, potential discrimination was further measured by other indicators, such as ethnic minority, language spoken and affiliation to a non-Christian religion. Ethnic minority status was assessed using self-reported belonging to an ethnic minority group. A further variable to measure ethnic minority status was computed based on the non-White majority in the origin country (UN 2006). Whilst this is only a rough indicator of the ethnicity of the actual respondent, it may be more reliable than self-reported ethnicity. Additionally, skin colour has been shown to be of importance in terms of likelihood of discrimination of an individual (Rooth 2002), and origin country ethnicity is the closest possible indicator we have for this factor. Non-Christian religion was computed as a binary variable for respondents who indicated belonging to a religion other than Christianity, as opposed to those who were Christian or who belonged to no religion. The variable, different language, indicates that a migrant speaks a language at home that is not an official language of the country of residence and is considered to be an indicator for belonging to a group that is potentially discriminated against (see, for example, André et al. 2008).
3.1.3. Individual level controls
In addition to the age and gender of the respondents, we controlled for social position as well. The social position of the respondents was measured by their education, their household income and their main source of household income. Education was measured in years of education for the highest completed credential. Income was measured by the relative income position of the household in the country of residence. Income source was coded into three categories, depending on the main source of income declared: The first includes income from employment and self-employment (reference category), the second indicates income from welfare and unemployment benefit (Benefits) and the third covers pensions and income from savings and other sources (Other Income).
3.1.4. Community and country level controls
The quality of governance is highly correlated with the economic performance of a country, such as economic growth or wealth (e.g. Kurtz and Schrank 2007) and economic performance, in turn, is seen as a major determinant of confidence in institutions (e.g. Newton and Norris 2000). In order to prevent confounding economic performance and institutional quality, we controlled, therefore, for Gross Domestic Product per Capita (GDP) at the country level and its differential effects on natives and immigrants. Moreover, previous research indicated that experiences and evaluative standards derived from the country of origin influenced the confidence of immigrants in the institutions of the host countries (e.g. Maxwell 2010; Röder and Mühlau 2010). In order to capture the potential effects of ‘dual frames of reference’, we further controlled for the difference between the quality of home and host country institutions (▵ Quality of Governance), calculated as the difference between the 2002–05 country averages of the quality of governance scores and the difference in GDP per Capita (▵ GDP). Both variables served also as proxies for economic motivations (▵ GDP) and political motivations (▵ Quality of Governance) to migrate to the host country.
3.1.5. Missing values
Missing values for independent variables were substituted, in hierarchical order, by the mean of the community group (migrant status×country of origin×country of residence), the country of origin mean (migrant status×country of origin) and the country of residence mean (migrant status×country of residence), and a dummy was included for missing values to account for any differences. All quantitative variables were centred on the sample mean with the exception of the quality of governance indicator. Household income was centred on the country mean.
3.2. Modelling
The data were modelled as hierarchical linear models (multilevel models) using MLwiN 2.21. The specification of the models was motivated by the nesting of the data and the structure of the hypotheses that we aimed to test in order to evaluate whether the data supported the core ideas of our theoretical sketch. Data were collected as random surveys of individuals in a number of European countries. The hypotheses required a direct comparison of how variables measured at the level of the surveyed countries affected migrants relative to natives. It was expected that these effects would differ for first- and second-generation migrants and sub-groups thereof. Moreover, earlier research has shown that attitudinal and positional outcomes of migrants do not only depend on the circumstances of the residence and origin countries, but also how they work together and on the relation between the residence and origin countries (e.g. Van Tubergen et al.2004). This was, among others, reflected in the fact that important control variables were measured as distances (differences) between survey countries and countries of origin. Accordingly, the data of natives and of first- and second-generation migrants were analysed in one model with three levels. Three levels were specified: Individuals at the lowest (n=100,538 of which 6,013 were first-generation migrants and 1,464 second-generation migrants) and countries of residence at the highest level (n=26). The second level was formed by ‘communities’. Communities were defined as (country of residence×country of origin×immigrant status) cells. An implication of this modelling strategy was that natives form a separate cluster of ‘communities’.
As there were no second level units containing both natives and immigrants, the differences between natives and immigrants were measured at the second level of communities and not at the level of individuals. Consequently, we chose to separate first- and second-generation migrants from the same country of origin in the same country of residence in different units because this simplified the model substantially. An alternative model with random slopes for first- and second-generation immigrants at the second level (country of residence×country of origin) and at the third level (country of origin) proved to be too complex to be estimated as were models with country of origin and country of residence cross-classifications. An implication of this way of modelling was that dummies for first- and second-generation immigrants were variables that varied at the community level and not an individual level attribute. Thus, the regression coefficients for these dummy variables represented not the average difference between a first-generation or second-generation migrant and a native individual within a country, but between the average native and the average first- and second-generation migrant community; product terms had to be interpreted accordingly. In total, there were 1,173 level 2 units. Twenty-six units comprised natives, 892 units comprised first-generation immigrant communities from 92 countries of origin and 255 units comprised second-generation immigrant communities from 68 countries of origin. Note that the model contains no cross-classified level of country of origin. Random intercepts were specified for the higher levels and random slopes for variables that were interacted with higher level predictor variables. For reasons of simplicity, only the variances of random intercepts and slopes were estimated, not the co-variances. The confidence in institutions exhibited much more variation for immigrants, in particular second-generation immigrants, than for natives. In order to take this ‘heteroscedasticity’ into account, additional residual variance components for immigrants were specified at the individual level (‘complex variance modelling’; see Browne et al.2002). In the tables, we report only the main variance component at the respective level. All models were estimated using full maximum likelihood estimation as implemented as iterated general least square (IGLS) in MLwiN.
4. Results
Model 1 reports the differences between natives and first- and second-generation migrants controlled for demographic and social position variables, host country context as well as host and origin country differentials in economic performance and quality of governance. Demographic variables all go in the expected direction. Older and better educated individuals and men are more confident than younger and poorer educated people and women, respectively. Low household income and dependence on social benefits are also related to low confidence in public institutions. The outcome variable is unrelated to the GDP per capita for natives. Neither GDP per capita nor GDP differences between residence and origin country matter for first- or second-generation migrants. The difference between the quality of residence and origin country institutions is positive and significant for first-generation migrants, revealing that a strong contrast between host and home country institutions exerts a positive influence on the evaluation of host country institutions. In line with the ‘dual frame of reference theory’, the influence of the relative quality of country of origin institutions is much smaller and more insignificant for second-generation immigrants. A strongly significant average relationship between the quality of host country institutions on the respondents’ confidence in public institutions underscores the importance of institutional performance for the evaluation and trust of residents. In contrast, the average effect of migrant integration policies is negative, even insignificant, indicating that, once controlled for the general quality of institutions, strong pro-immigrant policies may not impress the general public.
In Model 2, the differential impact of host country quality of governance is analysed and, as hypothesised, the positive effect of good governance on trust levels is weaker for immigrants. When taken separately for first- and second-generation migrants, these differences in the strength of the relationship remain below conventional significance levels, but are significant if both generations are taken together. This is in line with hypothesis H1. Further support for H1 is gained by the results of Model 3. This model contains a number of variables aiming to measure the propensity of being discriminated against, such as self-reported group discrimination, and variables measuring ‘otherness’ as indicators of being members of a group with a discrimination risk. Respondents who report belonging to a group that experiences discrimination are far less confident than others (b=–3.22) and immigrants are far more likely to have experienced discrimination (17 percent for first-generation immigrants and 19 percent for second-generation immigrants vs 5 percent for natives, see Table 1). However, ethnic minorities and those belonging to a religion other than Christianity are not less confident in public institutions than others, and immigrants speaking a language other than the official host country language at home are even more confident than immigrants speaking the official language.
. | Total . | Native . | Generation 1 . | Generation 2 . |
---|---|---|---|---|
Confidence | 19.8 | 19.7 | 21.4** | 18.6** |
Gender (% female) | 52.1 | 52.1 | 52.3 | 49.5* |
Age | 47.0 | 47.2 | 46.0*** | 40.7*** |
Education in years | 11.9 | 11.9 | 12.2*** | 12.1 |
Income | 6.13 | 6.12 | 6.39*** | 5.68*** |
Income source (% from wages/employment) | 65.8 | 65.6 | 68.2*** | 68.4*** |
Income source (% from welfare/unemployment) | 4.3 | 4.1 | 6.5*** | 5.9*** |
Income source (% from other sources) | 29.9 | 30.3 | 25.3*** | 25.7*** |
Not citizen of country of residence (%) | n/a | n/a | 52.9 | 21.8 |
Ethnic minority (%) | 3.6 | 1.7 | 27.6*** | 24.2*** |
Non-Christian religion (%) | 1.4 | 0.5 | 14.1*** | 12.1*** |
Different language spoken (%) | 6.4 | 3.7 | 40.7*** | 28.8*** |
Discrimination (%) | 5.6 | 4.8 | 16.2*** | 17.6*** |
Non-White minority (%) | n/a | n/a | 30.5 | 26.8 |
EU status (% from new member states) | n/a | n/a | 10.5 | 11.5 |
EU status (% from non-EU states) | n/a | n/a | 55.9 | 55.7 |
Quality of governance | 0.18 | 0.17 | 0.38*** | 0.10** |
▵ Quality of governance | n/a | n/a | 2.08 | 1.89 |
MIPEX | 0.00 | 0.01 | −0.06*** | −0.20*** |
GDP/10,000 | 3.51 | 3.47 | 4.22 | 3.39 |
▵ GDP | n/a | n/a | 2.61 | 1.85 |
Number of cases | 100,538 | 93,055 | 6,019 | 1,464 |
. | Total . | Native . | Generation 1 . | Generation 2 . |
---|---|---|---|---|
Confidence | 19.8 | 19.7 | 21.4** | 18.6** |
Gender (% female) | 52.1 | 52.1 | 52.3 | 49.5* |
Age | 47.0 | 47.2 | 46.0*** | 40.7*** |
Education in years | 11.9 | 11.9 | 12.2*** | 12.1 |
Income | 6.13 | 6.12 | 6.39*** | 5.68*** |
Income source (% from wages/employment) | 65.8 | 65.6 | 68.2*** | 68.4*** |
Income source (% from welfare/unemployment) | 4.3 | 4.1 | 6.5*** | 5.9*** |
Income source (% from other sources) | 29.9 | 30.3 | 25.3*** | 25.7*** |
Not citizen of country of residence (%) | n/a | n/a | 52.9 | 21.8 |
Ethnic minority (%) | 3.6 | 1.7 | 27.6*** | 24.2*** |
Non-Christian religion (%) | 1.4 | 0.5 | 14.1*** | 12.1*** |
Different language spoken (%) | 6.4 | 3.7 | 40.7*** | 28.8*** |
Discrimination (%) | 5.6 | 4.8 | 16.2*** | 17.6*** |
Non-White minority (%) | n/a | n/a | 30.5 | 26.8 |
EU status (% from new member states) | n/a | n/a | 10.5 | 11.5 |
EU status (% from non-EU states) | n/a | n/a | 55.9 | 55.7 |
Quality of governance | 0.18 | 0.17 | 0.38*** | 0.10** |
▵ Quality of governance | n/a | n/a | 2.08 | 1.89 |
MIPEX | 0.00 | 0.01 | −0.06*** | −0.20*** |
GDP/10,000 | 3.51 | 3.47 | 4.22 | 3.39 |
▵ GDP | n/a | n/a | 2.61 | 1.85 |
Number of cases | 100,538 | 93,055 | 6,019 | 1,464 |
Note: Significance levels: *p<.05, **p<.01, ***p<.001 (different from natives).
In spite of these controls, the relationship between quality of institutions and confidence in institutions becomes relatively weaker for immigrants, and the difference relative to natives in this specification is not only significant for the combined group of immigrants, but also for first-generation migrants separately. The larger gap between the effects of good governance for natives and for immigrants indicates that good governance is also associated with less discrimination against immigrants. However, these results show that immigrants are less sensitive to variations in the quality of public institutions even if a direct relationship between institutional quality and the extent of discrimination against immigrants is taken into account. This supports the argument that migrants do not partake as much as natives of the benefits of high-quality residence country institutions because exclusion and discrimination are a general feature of public institutions in European societies, and not only a characteristic of countries with low-performing public institutions.
Models 2.1 to 2.3 include the effects of quality of governance in the host country on immigrants, differentiated by those who hold citizenship and those who come from countries outside of the EU15. The interaction effects are best understood as separate slopes for each group, made up of the main slope for natives plus the effects for the relevant group to which a migrant belongs. In Model 2.2, for example, the slope for quality of governance for first-generation immigrants from an EU15 country, who does not hold citizenship of the host country, would have to be calculated as the sum of the main effect (quality of governance), plus the effect for being a first-generation migrant (quality of governance×generation 1) plus the effect of being a non-citizen first-generation migrant (quality of governance×non-citizen×generation 1). Here, however, the main point of interest is whether the effect is weaker depending on citizenship status and EU membership of the origin country, and we find that immigrants with citizenship experience a weaker effect of host country quality of governance than those without citizenship of the residence country. This is in contrast to the hypothesis that the relationship between quality of governance and confidence in institutions should be weaker for immigrants without citizenship status. It is also not the case that the relationship between quality of governance and confidence is weaker for immigrants from non-EU countries. The interaction effects for non-EU migrants is very small, indicating that there is no difference between migrants from countries outside the EU and migrants from established EU countries. Hence, there is no support for hypothesis H2.
Models 3.1 to 3.5 include the product terms between the quality of governance in the host country and the discrimination variables introduced in Model 3. Again, the interaction effects in these models can be understood basically as separate effects for subgroups. When these indicators of discrimination are interacted with the quality of governance in the host country (Models 3.1 to 3.5), we see that quality of governance has a weaker effect on individuals who are different or who report belonging to a group that is discriminated against or that is an ethnic minority. The effects are significant only for the latter two, but together, this does provide evidence to support hypothesis H3.
Immigrant integration policies do not have a differential impact on immigrants’ confidence levels (see Model 2). Equally, the citizenship of immigrants or the EU status of their origin country does not alter the impact of these policies, as illustrated in Model 2.3. Therefore, there is no support for hypothesis H4 that integration policies are positively associated with more confidence in public institutions amongst immigrants, especially from non-EU countries.
As indicated in the average raw scores (Table 1), second-generation migrants are less confident than natives. Similarly, significant differences between second-generation immigrants and natives can be found for multilevel models without controls (b=–0.94; s.e.=0.36) or in models with demographical and social position controls (b=–0.86; s.e.=0.37; both not reported in the table). Additional controls at the country and community level increase slightly the estimated difference between natives and second-generation migrants (Table 2, Model 1: b=–0.95; s.e.=0.49), but the coefficient becomes insignificant as its standard error is increased by the additional controls. The data thus provide considerable support for the part of hypothesis H5 pertaining to second-generation migrants.
. | Model 1 . | Model 2 . | Model 3 . | |||
---|---|---|---|---|---|---|
. | Coeff . | SE . | Coeff . | SE . | Coeff . | SE . |
Individual level | ||||||
Intercept | −4.15 | 6.53 | −4.28 | 6.53 | −6.56 | 6.49 |
Background | ||||||
Gender (female) | −0.10* | 0.05 | −0.09* | 0.05 | −0.10* | 0.05 |
Age/10 | 0.13*** | 0.02 | 0.13*** | 0.02 | 0.12*** | 0.02 |
Social Position | ||||||
Education | 0.15*** | 0.01 | 0.15*** | 0.01 | 0.16*** | 0.01 |
Income | 0.25*** | 0.02 | 0.25*** | 0.02 | 0.24*** | 0.01 |
Benefits | −1.49*** | 0.12 | −1.49*** | 0.12 | −1.26*** | 0.12 |
Other source | 0.24** | 0.07 | 0.24** | 0.07 | 0.26*** | 0.07 |
Discrimination | ||||||
Ethnic minority | −0.05 | 0.15 | ||||
Non-Christian religion | 0.34 | 0.23 | ||||
Different language spoken | 0.70*** | 0.13 | ||||
Discrimination | −3.22*** | 0.10 | ||||
Community level | ||||||
Migrant status | ||||||
Generation 1 | −0.22 | 0.34 | 0.01 | 0.35 | 0.07 | 0.34 |
Generation 2 | −0.95 | 0.49 | −0.72 | 0.51 | −0.73 | 0.48 |
Discrimination | ||||||
Non-White majority | −0.23 | 0.32 | ||||
Country of origin | ||||||
▵ GDP/10,000 X Generation 1 | 0.11 | 0.15 | 0.09 | 0.15 | 0.13 | 0.15 |
▵ GDP/10,000 X Generation 2 | −0.12 | 0.28 | −0.24 | 0.30 | −0.23 | 0.29 |
▵ Quality of governance X Gen. 1 | 0.35** | 0.14 | 0.38** | 0.14 | 0.44** | 0.15 |
▵ Quality of governance X Gen. 2 | 0.17 | 0.28 | 0.30 | 0.29 | 0.45 | 0.30 |
Country of residence | ||||||
GDP/10,000 | 0.50 | 0.37 | 0.31 | 0.38 | 0.19 | 0.38 |
X Generation 1 | −0.07 | 0.18 | 0.32 | 0.29 | 0.31 | 0.27 |
X Generation 2 | 0.13 | 0.27 | 0.68 | 0.43 | 0.62 | 0.42 |
Quality of governance | 2.91 | 0.95 | 3.52** | 0.99 | 3.81*** | 0.99 |
X Generation 1 | −1.36 | 0.74 | −1.52* | 0.70 | ||
X Generation 2 | −1.64 | 1.05 | −1.40 | 0.95 | ||
(X Immigrant) | −1.39* | 0.68 | −1.49* | 0.66) | ||
MIPEX | −0.68 | 0.44 | −0.73 | 0.47 | −0.74 | 0.47 |
X Generation 1 | 0.25 | 0.30 | 0.25 | 0.29 | ||
X Generation 2 | −0.10 | 0.38 | −0.06 | 0.39 | ||
Variance components (selectively) | ||||||
Individual level | 51.65*** | 0.24 | 51.65*** | 0.24 | 51.17*** | 0.24 |
Community level | 1.07*** | 0.29 | 0.98*** | 0.27 | 0.95*** | 0.26 |
Country level | 3.65*** | 1.14 | 3.58*** | 1.12 | 3.73*** | 1.15 |
Number of Cases | n1=100,538 | n2=1,173 | n3=26 |
. | Model 1 . | Model 2 . | Model 3 . | |||
---|---|---|---|---|---|---|
. | Coeff . | SE . | Coeff . | SE . | Coeff . | SE . |
Individual level | ||||||
Intercept | −4.15 | 6.53 | −4.28 | 6.53 | −6.56 | 6.49 |
Background | ||||||
Gender (female) | −0.10* | 0.05 | −0.09* | 0.05 | −0.10* | 0.05 |
Age/10 | 0.13*** | 0.02 | 0.13*** | 0.02 | 0.12*** | 0.02 |
Social Position | ||||||
Education | 0.15*** | 0.01 | 0.15*** | 0.01 | 0.16*** | 0.01 |
Income | 0.25*** | 0.02 | 0.25*** | 0.02 | 0.24*** | 0.01 |
Benefits | −1.49*** | 0.12 | −1.49*** | 0.12 | −1.26*** | 0.12 |
Other source | 0.24** | 0.07 | 0.24** | 0.07 | 0.26*** | 0.07 |
Discrimination | ||||||
Ethnic minority | −0.05 | 0.15 | ||||
Non-Christian religion | 0.34 | 0.23 | ||||
Different language spoken | 0.70*** | 0.13 | ||||
Discrimination | −3.22*** | 0.10 | ||||
Community level | ||||||
Migrant status | ||||||
Generation 1 | −0.22 | 0.34 | 0.01 | 0.35 | 0.07 | 0.34 |
Generation 2 | −0.95 | 0.49 | −0.72 | 0.51 | −0.73 | 0.48 |
Discrimination | ||||||
Non-White majority | −0.23 | 0.32 | ||||
Country of origin | ||||||
▵ GDP/10,000 X Generation 1 | 0.11 | 0.15 | 0.09 | 0.15 | 0.13 | 0.15 |
▵ GDP/10,000 X Generation 2 | −0.12 | 0.28 | −0.24 | 0.30 | −0.23 | 0.29 |
▵ Quality of governance X Gen. 1 | 0.35** | 0.14 | 0.38** | 0.14 | 0.44** | 0.15 |
▵ Quality of governance X Gen. 2 | 0.17 | 0.28 | 0.30 | 0.29 | 0.45 | 0.30 |
Country of residence | ||||||
GDP/10,000 | 0.50 | 0.37 | 0.31 | 0.38 | 0.19 | 0.38 |
X Generation 1 | −0.07 | 0.18 | 0.32 | 0.29 | 0.31 | 0.27 |
X Generation 2 | 0.13 | 0.27 | 0.68 | 0.43 | 0.62 | 0.42 |
Quality of governance | 2.91 | 0.95 | 3.52** | 0.99 | 3.81*** | 0.99 |
X Generation 1 | −1.36 | 0.74 | −1.52* | 0.70 | ||
X Generation 2 | −1.64 | 1.05 | −1.40 | 0.95 | ||
(X Immigrant) | −1.39* | 0.68 | −1.49* | 0.66) | ||
MIPEX | −0.68 | 0.44 | −0.73 | 0.47 | −0.74 | 0.47 |
X Generation 1 | 0.25 | 0.30 | 0.25 | 0.29 | ||
X Generation 2 | −0.10 | 0.38 | −0.06 | 0.39 | ||
Variance components (selectively) | ||||||
Individual level | 51.65*** | 0.24 | 51.65*** | 0.24 | 51.17*** | 0.24 |
Community level | 1.07*** | 0.29 | 0.98*** | 0.27 | 0.95*** | 0.26 |
Country level | 3.65*** | 1.14 | 3.58*** | 1.12 | 3.73*** | 1.15 |
Number of Cases | n1=100,538 | n2=1,173 | n3=26 |
Note: Significance levels: *p<.05, **p<.01, ***p<.001.
The situation is less clear-cut for first-generation migrants. Raw scores (Table 1) and multilevel models without controls (b=0.76; s.e.=0.30) and with demographical and social position controls (b=0.84; s.e.=0.31, both not reported in the table) all indicate significantly higher confidence levels among first-generation migrants than natives. In Model 1, this coefficient turns negative (b=–0.22; s.e.=0.34), but the coefficient is small and insignificant. As Model 1 shows, the higher confidence of first-generation migrants can be accounted for by the economic and institutional differences between the origin and host countries. It is largely the gap in quality of governance between the host and home countries that explains the higher confidence of first-generation migrants. Controlled for the gap, first-generation migrants exhibit less confidence than natives in the sample, but support for H5 is limited by the fact that the difference is far from being significant.
. | Model 2.1 . | Model 2.2 . | Model 2.3 . | |||
---|---|---|---|---|---|---|
. | Coeff . | SE . | Coeff . | SE . | Coeff . | SE . |
Individual level | ||||||
Legal status | ||||||
Non-citizen Generation 1 | 0.66* | 0.26 | 0.14 | 0.27 | 0.66* | 0.25 |
Non-citizen Generation 2 | −1.06 | 0.57 | −1.57* | 0.62 | −0.95 | 0.60 |
Non-EU | −0.14 | 0.45 | −0.08 | 0.51 | −0.18 | 0.46 |
New EU | −0.40 | 0.51 | −0.22 | 0.57 | −0.47 | 0.52 |
Community level | ||||||
Migrant status | ||||||
Generation 1 | 0.34 | 0.37 | −0.16 | 0.44 | 0.36 | 0.37 |
Generation 2 | −1.56* | 0.66 | −2.22*** | 0.74 | −1.46* | 0.67 |
Country of residence | ||||||
Quality of governance | 3.64*** | 0.95 | 3.64*** | 0.96 | 3.64*** | 0.96 |
X Generation 1 | −1.24 | 0.73 | −0.26 | 0.85 | −1.25 | 0.73 |
X Generation 2 | −1.20 | 0.96 | 0.49 | 1.40 | −1.11 | 0.99 |
X Non-citizen generation 1 | 1.19*** | 0.38 | ||||
X Non-citizen generation 2 | 1.48 | 0.96 | ||||
X Non-EU | −0.15 | 0.47 | ||||
X New EU | −0.66 | 0.65 | ||||
MIPEX | −0.76 | 0.45 | −0.76 | 0.46 | −0.76 | 0.46 |
X Generation 1 | 0.28 | 0.30 | 0.27 | 0.30 | 0.16 | 0.36 |
X Generation 2 | −0.28 | 0.38 | −0.26 | 0.36 | −0.19 | 0.68 |
X Non-citizen generation 1 | 0.33 | 0.24 | ||||
X Non-citizen generation 2 | 0.37 | 0.68 | ||||
X Non-EU | 0.38 | 0.25 | ||||
X New EU | 0.25 | 0.39 | ||||
Variance components (selectively) | ||||||
Individual level | 51.65*** | 0.24 | 51.65*** | 0.24 | 51.65*** | 0.24 |
Community level | 0.82*** | 0.26 | 0.76*** | 0.24 | 0.87*** | 0.26 |
Country level | 3.55*** | 1.10 | 3.71*** | 1.14 | 3.57** | 1.11 |
Number of cases | n1=100,538, n2=1,173, n3=26 |
. | Model 2.1 . | Model 2.2 . | Model 2.3 . | |||
---|---|---|---|---|---|---|
. | Coeff . | SE . | Coeff . | SE . | Coeff . | SE . |
Individual level | ||||||
Legal status | ||||||
Non-citizen Generation 1 | 0.66* | 0.26 | 0.14 | 0.27 | 0.66* | 0.25 |
Non-citizen Generation 2 | −1.06 | 0.57 | −1.57* | 0.62 | −0.95 | 0.60 |
Non-EU | −0.14 | 0.45 | −0.08 | 0.51 | −0.18 | 0.46 |
New EU | −0.40 | 0.51 | −0.22 | 0.57 | −0.47 | 0.52 |
Community level | ||||||
Migrant status | ||||||
Generation 1 | 0.34 | 0.37 | −0.16 | 0.44 | 0.36 | 0.37 |
Generation 2 | −1.56* | 0.66 | −2.22*** | 0.74 | −1.46* | 0.67 |
Country of residence | ||||||
Quality of governance | 3.64*** | 0.95 | 3.64*** | 0.96 | 3.64*** | 0.96 |
X Generation 1 | −1.24 | 0.73 | −0.26 | 0.85 | −1.25 | 0.73 |
X Generation 2 | −1.20 | 0.96 | 0.49 | 1.40 | −1.11 | 0.99 |
X Non-citizen generation 1 | 1.19*** | 0.38 | ||||
X Non-citizen generation 2 | 1.48 | 0.96 | ||||
X Non-EU | −0.15 | 0.47 | ||||
X New EU | −0.66 | 0.65 | ||||
MIPEX | −0.76 | 0.45 | −0.76 | 0.46 | −0.76 | 0.46 |
X Generation 1 | 0.28 | 0.30 | 0.27 | 0.30 | 0.16 | 0.36 |
X Generation 2 | −0.28 | 0.38 | −0.26 | 0.36 | −0.19 | 0.68 |
X Non-citizen generation 1 | 0.33 | 0.24 | ||||
X Non-citizen generation 2 | 0.37 | 0.68 | ||||
X Non-EU | 0.38 | 0.25 | ||||
X New EU | 0.25 | 0.39 | ||||
Variance components (selectively) | ||||||
Individual level | 51.65*** | 0.24 | 51.65*** | 0.24 | 51.65*** | 0.24 |
Community level | 0.82*** | 0.26 | 0.76*** | 0.24 | 0.87*** | 0.26 |
Country level | 3.55*** | 1.10 | 3.71*** | 1.14 | 3.57** | 1.11 |
Number of cases | n1=100,538, n2=1,173, n3=26 |
Notes: Significance levels: *p<.05, **p<.01, ***p<.001.
Controls as in Model 2.
5. Conclusion
In this paper, we explored whether immigrants experience institutions differently from mainstream members of society, reflecting processes of discrimination and social exclusion. In general, it is more likely that migrants feel they belong to a group that is discriminated against, and this perceived discrimination, in turn, is associated with lower confidence in public institutions. We argued that systematic processes of discrimination and social exclusion would imply that immigrants do benefit less from public service provisions than natives. In line with this expectation, we found that the performance of public institutions of the country of residence, measured by a quality of governance indicator, which is a strong predictor of the confidence in institutions of natives, has less effect on the confidence of migrants. This holds for both first- and second-generation migrants to the same extent. Moreover, the interpretation that the weaker relationship between performance of and confidence in public institutions for immigrants reflects discriminatory processes is further strengthened by the finding that a weaker relationship also holds for those belonging to an ethnic minority and who see themselves as members of a discriminated group.
. | Coeff. (SE) Model 3.1 . | Coeff. (SE) Model 3.2 . | Coeff. (SE) Model 3.3 . | Coeff. (SE) Model 3.4 . | Coeff. (SE) Model 3.5 . |
---|---|---|---|---|---|
Quality of governance | 3.83*** (0.97) | 3.88*** (0.97) | 3.90*** (0.96) | 3.80*** (0.97) | 3.81*** (0.98) |
X Generation 1 | −1.42* (0.70) | −1.43* (0.69) | −1.34 (0.72) | −1.37* (0.67) | −1.32 (0.69) |
X Generation 2 | −1.37 (0.94) | −1.36 (0.93) | −1.58 (0.90) | −1.44 (0.92) | −1.19 (0.95) |
X Discrimination | −0.42* (0.19) | ||||
X Ethnic minority | −0.99** (0.35) | ||||
X Different language | −0.41 (0.44) | ||||
X Non-Christian | −0.32 (0.64) | ||||
X Non-White majority | −0.76 (0.55) |
. | Coeff. (SE) Model 3.1 . | Coeff. (SE) Model 3.2 . | Coeff. (SE) Model 3.3 . | Coeff. (SE) Model 3.4 . | Coeff. (SE) Model 3.5 . |
---|---|---|---|---|---|
Quality of governance | 3.83*** (0.97) | 3.88*** (0.97) | 3.90*** (0.96) | 3.80*** (0.97) | 3.81*** (0.98) |
X Generation 1 | −1.42* (0.70) | −1.43* (0.69) | −1.34 (0.72) | −1.37* (0.67) | −1.32 (0.69) |
X Generation 2 | −1.37 (0.94) | −1.36 (0.93) | −1.58 (0.90) | −1.44 (0.92) | −1.19 (0.95) |
X Discrimination | −0.42* (0.19) | ||||
X Ethnic minority | −0.99** (0.35) | ||||
X Different language | −0.41 (0.44) | ||||
X Non-Christian | −0.32 (0.64) | ||||
X Non-White majority | −0.76 (0.55) |
Note: Controls as in Model 3.
Further in line with our expectations, we find that second-generation migrants exhibit less confidence in public institutions than natives and first-generation migrants. This difference is robust regardless of the control and potentially mediating variables included in the models. Whilst first-generation migrants are to some extent influenced by discrimination as discussed above, their more optimistic outlook, based on a favourable comparison to their country of origin, seems to decrease the negative impact on confidence in institutions to some extent. The second generation, on the other hand, does not compare the institutions to the origin country, but, having grown up in the host country, will have developed expectations similar to those of the natives. Discrimination and exclusion, perceived or real, will therefore have a stronger impact on their judgements of public institutions, resulting in a constellation of high expectations and limited opportunities in conjunction with feeling less valued and accepted by mainstream society (see, e.g., Gans 1992; Crul and Vermeulen 2003; Heath et al.2008).
Unlike second-generation migrants, first-generation immigrants are not less confident overall in public institutions. This reflects previous findings by Bilodeau and Nevitte (2003), Maxwell (2010) and Röder and Mühlau (2010) that negative past experiences with institutions lead to overconfidence when immigrants arrive in countries with more favourable conditions, which then fades over time as expectations rise and evaluations become more in line with those of the native population.
The dominant role of expectations formed in the country of origin may also be the key for understanding the puzzling effects of the variables that measure legal status. While second-generation migrants holding citizenship of the residence country are more trusting than second-generation migrants who are not citizens of the country of residence, naturalised first-generation migrants are less trusting than those without citizenship. Moreover, the performance of public institutions has a significantly stronger effect on non-citizens than citizens. In interpreting this, however, it is important to note that citizenship is not only an indicator of legal rights, but also one of acculturation to the host country. This effect is, therefore, better explained by the idea that ‘frames of reference’ are of importance for first-generation immigrants’ confidence, with more acculturated migrants being less confident in general as they begin to evaluate the country less with a comparison to the origin country in their minds. This effect has, for example, been shown also by Michelson (2001, 2003) and Wenzel (2006) for Mexican Americans living in the United States. The idea of ‘dual frames of reference’ is further supported by the observation that groups that are most likely socially excluded and discriminated against (migrants from outside Europe, affiliated to non-Christian religions or speaking a different language than the official language of the country of residence at home) tend to exhibit more, rather than less, confidence in public institutions than natives or migrants from other groups. The difference in quality of public institutions between the origin and host countries also has a strong effect on first-generation migrants, with migrants coming from countries with low institutional quality and arriving in a comparatively well-governed host country exhibiting higher confidence.
Migrant integration policies as far as they are reflected in the Migration Policy Index, in contrast, display a negative relationship with confidence in public institutions of natives and migrants alike, once controlled for quality of governance. This is in line with findings that migrant integration policies are unrelated to the perceived discrimination (Andre et al.2008), scholastic achievement (De Heus and Dronkers 2009) or labour market integration (Fleischmann and Dronkers 2010) of migrants. The consistency of the findings across various indicators of integration calls for a debate on how well the index reflects effective integration policies.
Taken together, the data provide some support for the idea that discrimination shapes the confidence of immigrants’ in public institutions, but we do not find that lack of entitlements are important for immigrants’ confidence.
In this study, we evaluated confidence in the four public institutions that were included in all of the first three rounds of the European Social Survey (parliament, the legal system, the police and politicians). This is a severe limitation because other institutions such as social, housing, educational and health services may be more important for the daily experiences of migrants and their well-being. Within these limitations, the study indicates that migrants’ confidence in public institutions is only partially driven by processes of discrimination and social exclusion. For first-generation immigrants, different expectations based on previous experiences in the origin country appear to be of importance, and this seems to be more important than discrimination as a mechanism to explain their levels of confidence. The second generation, without these dual frames of reference, is somewhat less confident, but these effects are not sufficiently strong to speak of an overall disengagement from institutions in the host country.
References
Antje Röder is a PhD student in the Department of Sociology and the Institute for International Integration Studies at Trinity College Dublin. Her main research interest is how the societal contexts of the origin country and the country of residence shape the socio-economic and socio-cultural integration of immigrants in Europe.
Peter Mühlau is a lecturer in the Department of Sociology at Trinity College Dublin. His main research interests are employment and migration studies.