Does immigration hollow out state legitimacy in times of economic crisis?

ABSTRACT

This research examines the linkage between immigration and legitimacy by using comparative data and methods. Two approaches will be used to test the assumption that there is a connection between immigration and state legitimacy. First, the cross sectional approach compares the attitudes of groups of citizens on immigration and compares them with their attitude towards the democratic functioning and performance of the state. Second, the longitudinal approach investigates whether the rise of immigration coincides simultaneously with a decline of state legitimacy. It will be shown that immigration matters for state legitimacy, but that there is no general increase as a result of the economic recession. The main exception to this is Southern Europe where the structural state support dropped dramatically. In addition, a large policy gap lowers the trust of citizens in the state and its capacity to control the borders. The number of immigrants hardly matters for anti-immigration attitudes. Although the crisis-effect is weak, the satisfaction with the economy plays an important role. A multi-level analysis shows that it is an important determinant of the support of citizens of the state. The same goes for the indicator of anti-immigration attitudes which is among the strongest determinants of state legitimacy.

KEYWORDS:

• Immigration
• state legitimacy
• survey research
• citizen perceptions
• economic recession

1. Introduction

European states are confronted with increasing immigration flows. This is widely discussed in the media and large parts of the public seem very concerned with this topic. If European states are not willing or unable to control the flows of immigration, this may go at the cost of their legitimacy because it touches upon one of the primary state functions: to define those who are members of its society and to control entry to and exit from its territory (Dryzek & Dunleavy, 2009). An associated characteristic of the state is that its authority is accepted by a large majority of the citizens. If states are not able to fulfil one or more of their primary functions and/or if a large majority of the citizens feels that the state is not able to do so, a state may (partly) loose its legitimacy. As a consequence, citizens lose their trust in state institutions. This response may be strengthened by an economic recession.

The goal of this research is to examine the linkage between immigration and legitimacy by using comparative data and methods. Two approaches will be used to test the assumption that there is a connection between immigration and state legitimacy. First, the cross sectional approach compares the attitudes of groups of citizens on immigration and compares them with their attitude towards the democratic functioning and performance of the state. Second, the longitudinal approach investigates whether the rise of immigration coincides simultaneously with a decline of state legitimacy. A combination of these approaches offers a broader scope than is usually offered in existing research on the variations between immigration and state legitimacy by the inclusion of five European regions and six time points. This design allows for a pan-European and longitudinal analysis of the variations in the attitudes towards immigration and state legitimacy.

Whether the attitudes of citizens towards immigration matter for state legitimacy depends on how much they trust the state to be able to control its borders and to protect the national identity from disruptive external influences. These are rather abstract relationships that are often not part the questions that are asked in surveys and therefore hard to research in an empirical way (Gilley, 2009). For this reason, we opt for an approach in which we compare the group of citizens that is opposed to immigration with the group that is positive about immigration on several indicators that are related to democratic performance of states. In this way we can examine how strong (the perception of) immigration is related to the support of citizens of the state system in which they live. It will be shown that immigration matters for state legitimacy, but that this relationship is not strongly affected by the economic recession.

2. Theories and hypotheses

Legitimacy is the right of an authority to exercise power which is accepted by the citizens (Beetham, 2013). Immigration is only one challenge for the modern state. It is part of a wider range of factors that accompany globalisation and that might weaken the state (Gilley, 2009; Hampshire, 2013). However, there is no consensus in the literature on whether this means that the state has actually been weakened. One group of scholars argues that globalisation has hollowed out the state functions (and therefore its legitimacy) in several ways (Holton, 2011; Vries, 2001). Hay (2002) has argued that globalisation itself does not force states into any direction. Instead the rise of neo-liberal ideas lead to restrictive fiscal policies that may widen the gap between what states have to offer and what citizens expect from them. Immigration may be used by anti-establishment groups like populist parties to highlight the problems that states have with securing the national identity and borders (Kriesi, Grande, Lachat, Dolezal, & Bornschier, 2008; Mudde, 2004). Another group of scholars argues that the role of the state has been strengthened due to globalisation. The reason is that it brings problems to the front door like illegal immigration, environmental degradation and economic crises that can only be solved by means of cooperation between states (Vries, 2001). Since state legitimacy is strongly based on public consent with state activities, analysing survey data on individual attitudes may reveal how it has been affected by immigration and the economic crisis.

This research focuses on immigration in general and not on more specific groups like illegal immigrants or asylum seekers because most questions in Survey research refer to the broadest category. The research will incorporate both individual and contextual factors that impact on the attitudes of citizens towards immigration and how this affects their view on the capacity of states to fulfil their basic functions. The focus will be on five European regions (North, South, Anglo-Saxon, West and East). By regions we do not mean subnational regions but clusters of geographically nearby countries that have cultural, economic and historical ties and similarities. It is expected that these differences are relevant for the perception of immigration. This is not to deny that there may be important differences between countries belonging to the same region. However, these intra-regional differences are less pronounced than the differences between the regions regarding immigration and economic welfare.

Apart from the debate on the effects of globalisation on the state and works on the ethical aspects of immigration there are no established theories on the relationship between immigration and state legitimacy (see for an overview of existing insights and assumptions: McLaren, 2015). However, there is much research on the factors that impact on the attitudes towards immigration. Existing research has shown that some individual variables like the personal income have a strong impact on the opposition to immigration (McLaren, 2015). The impact of contextual variables like the past experience with immigration, far-right party popularity, government policy towards migrants and the type of political institutions are often shown to be low to moderate. This research seeks to combine both types of variables by examining how groups with different attitudes towards immigration perceive different aspects of the democratic functioning of states. In order to examine this relationship, a number of hypotheses are formulated on the conditions in which immigration might hollow out state legitimacy.

The support hypothesis specifies the conditions in which state legitimacy is weakened. Central are indicators that measure the degree to which citizens trust and support their state. Two of these indicators have been developed by Easton (1975). Diffuse support of state institutions is based of affective attachment which is stable and enduring and stems from the socialisation process. Specific support relates mainly to the incumbent authorities and may fluctuate depending on the perceived performance of governments. We expect a weak relationship between immigration and diffuse support (except in times of crisis) and a strong relationship in case of specific support. The crisis may offer a context in which the negative attitudes towards immigration are strengthened to such a degree that it has not only a negative impact on the specific support, but also on the diffuse (affective) support of state institutions.

The distrust hypothesis expects that the more citizens feel threatened by immigration, the more they distrust state institutions. If such a connection exists, it shows that state legitimacy is affected by concerns about immigration. It is assumed that this relationship differs per European region. In regions that are economically and socially pressurised due to the economic crisis and/or due to a low level of socio-economic development, the relationship is expected to be stronger even if the influx of immigrants is low.

The influx hypothesis focuses on the number of immigrants. Some authors argue that countries that are confronted with significant immigration flows for a longer time run a higher chance that their legitimacy is being weakened due to immigration (McLaren, 2015). Other researchers have shown that the impact of the number of immigrants is weak and instead the perception of immigration matters (Parsons & Smeeding, 2006). This perception may be influenced by the nationality of immigrants, the national political culture and degree of openness of a country towards external influences. Since many factors may impact on how immigration affects public opinion, we hypothesise that the perception of immigration and not the number of immigrants matters for how citizens perceive the state.

The gap hypothesis explores how immigration policies matter. If states adopt restrictive policies, there is less chance that parts of the population will distrust the state capacity to control immigration, but at the same time it might anger those parts of the population that feel positive about immigration. According to the gap hypothesis it is likely that the size of the gap between public opinion regarding immigration and immigration policy affects the (dis)content with immigration. Since countries have different past histories regarding immigration and different national identities, it is likely that this gap and the response of the public differs per country. Some have found that the levels of concern about immigration were very high in the early 1990s in several of the longer term immigration countries (McLaren, 2015). If there is a gap between what the government does and the feelings of large groups of citizens about immigrants, this may strengthen the perception of government failure in handling this policy area especially in times of economic crisis. We hypothesise that a large policy gap will lower the trust of citizens in the state and its capacity to control the borders.

The crisis hypothesis stresses the role of the global economic recession which has fuelled the discussion on the cost of immigration for domestic economies (Dancygier & Donnelly, 2014). The crisis may have strengthened the framing of immigration as a primarily economic issue. Such economic concerns may be driven by evaluations of citizens’ personal economic situation and/or a nation’s economic health. Existing research has found that, especially in times of economic recession, citizens are inclined towards unfavourable attitudes towards immigration because foreigners are perceived as competitors on the labour market and the housing market (Burns & Gimpel, 2000). In general, it is expected that the stronger the impact of the economic crisis is, the more citizens are inclined towards anti-immigration attitudes and distrust in government.

The populism hypothesis focuses on how the growth of populist parties has contributed to the politicisation of the immigration issue. Politicisation can take several forms. One is the rise of a cultural conflict dimension next to the traditional economic one (Kriesi et al., 2008). This may lead to a divide within left parties between a pro-immigration and an anti-immigration camp. Those who are against may leave the left party and join the anti-immigrant parties. The populist challenge of the established parties will in that case strengthen the process of the erosion of state legitimacy. However, in other circumstances the rise of populism may strengthen the trust in state institutions because it may weaken the role of corruption, as in the case in (parts of) Eastern Europe (Gherghina, Miscoiu, & Soare, 2013; Kriesi, 2014). Given this diversity, we hypothesise that populism is weakly related to immigration within Europe as a whole.

These six hypotheses will be tested using the European Social Survey (ESS, http://www.europeansocialsurvey.org). In particular the sixth wave (2012/2013) and the longitudinal ESS data (2002–2014) will be used in order to examine the relationship between immigration and legitimacy.

3. Exploring regional diversity

We expect regional variations in the degree to which immigration has hollowed out state legitimacy. We opt for a comparative approach by means of the quantification of indicators of the politics of immigration per European region. The degree to which immigration hollows out state legitimacy strongly depends on the national context and the moment in time. For this reason, an inclusive research design is preferred that incorporates several time points and many countries in the main European regions.¹ The design is both variable oriented and case oriented in order to capture the full range of spatial and temporal variations and to avoid generalisations which are based on too few instances. In these respects this research offers a broader scope of the connection between immigration and state legitimacy than is offered by many other studies on this topic.

Table 1 compares the politics of immigration of European regions which is indicated by a selection of variables that follow from the hypotheses just discussed. A high percentage in the first column indicates that more countries in the region are confronted with high levels of immigration. In the next columns a high percentage means more negative attitudes towards immigration, more restrictive policies, a larger gap or more electoral support for populist parties. The divisions between high versus low etc. are based on the overall median score. The net immigration in this table is calculated as the percentage of countries within an European region in which the number of immigrants is above the median. It turns out that there are important differences in this regard between the regions. The Anglo-Saxon and Western European countries are most pressed, the Nordic and Southern countries take an intermediate position and the Eastern countries are hardly confronted with any immigration. These differences make it interesting to examine to which extent these regions are confronted with negative attitudes towards immigration that may affect state legitimacy. We have earlier referred to the statement made by some authors that large immigration flows coincide with more negative attitudes towards immigration. Our analysis yields a low correlation (r = 0.04, p = .000) between these variables. One inference that can be made on the basis of this finding is that, if the real immigration flows do hardly matter, the perception of the immigration is more decisive for trust and support than the number of immigrants.

Table 1. Basic information about the politics of immigration per European region.

Region	Per cent countries with high levels of immigration (net_immigration)	Per cent citizens with negative attitudes towards immigration (imm_index)	Per cent countries with restrictive policies (MIPEX)	Per cent countries with a larger policy gap	Per cent countries with a large electoral support for populist parties (Populism)
Nordic	56.8	30.8	21.8	36.5	0
Anglo-Saxon	100.0	55.6	100.0	44.4	0
Western	91.0	42.9	38.9	39.8	72.2
Southern	39.0	47.6	52.1	48.2	100.0
Eastern	1.8	46.9	100.0	53.1	87.8
Total	62.8	45.8	59.9	44.4	59.8

Notes: Explanation: Reading example: In the Nordic region 56.8 per cent of the countries has a net immigration that is higher than the overall median. See Appendix for an explanation of the variables.

The immigration index (imm_index) measures the positive or negative attitudes of citizens on immigration. It is based on the mean of three indictors on culture, economy and living conditions (see Appendix). Table 1 shows the percentage of respondents which finds that immigrants have a negative impact (the division between positive and negative is based on the median score). The results show that the citizens are most negative in the Anglo-Saxon countries, more positive in the Nordic countries and they take an intermediate position in the other three regions. The differences are remarkable small since around half of the population in most countries is negative about immigration, irrespective the different conditions and circumstances in the regions. This is a remarkable outcome which suggests that variables at the individual level have more effect on attitudes towards immigration than contextual variables.

The column in Table 1 labelled MIPEX-index² indicates how many countries in the region have implemented restrictive immigration policies. The correlation between the net immigration and the immigration index is rather weak which indicates that the restrictiveness of states is not strongly related to the number of immigrants. This outcome underlines that state policies are the result of path dependencies, long term developments and international treaties.

The policy gap is derived from a comparison between the attitudes on immigration with the migrant integration policy (the MIPEX index). The gap is measured as the difference between the dichotomised scores of MIPEX and the attitudes on immigration (imm_index). This measure is relevant because it signals the distance between state policies and the perception of the citizens. Such a gap indicates whether there is a potential legitimacy problem. A gap does exist if citizens are against immigration and the policy is non-restrictive or in case citizens are pro immigrants and the policy is restrictive. On average such a gap does exist in case of 44 per cent of the respondents. This gap is considerable and would be worrisome if most of these respondents would be negative about the state performance. Additional analysis (not shown here, but part of Table 3) indicates that this is not the case. On average about half of the respondents that experiences a gap is negative about performance indicators like satisfaction with democracy and government. This is about the same in case of respondents that do not experience a gap. The scores are similar in all five regions (they vary between 53.1 per cent in the East and 36.5 per cent in the Nordic countries) which indicates that all regions are confronted with this gap to a similar degree. The gap seems to be more or less independent from the conditions and circumstances in the regions.

Finally, the variable populism presents the electoral success of populist parties (both left and right ones) that often take an anti-immigration position in the most recent national elections relative to 2015 (Mudde, 2015). The percentage is based on the number of countries in the region in which such parties are successful (based on the median score). Again, it turns out that there is hardly any relationship with the other variables. The results show that the impact of populism on the support of immigration and democracy is low to moderate. The correlation between populism and democracy satisfaction (How satisfied with the way democracy works in country) is −0.23 and the correlation between populism and the immigration index is 0.04. Populism is therefore stronger related to the satisfaction with democracy than with anti-immigration attitudes.

These patterned variations indicate that negative attitudes towards immigration do not result from the number of immigrants, government policies or populism. If variables at the national or regional level are less important for attitudes towards immigration, it is likely that factors related the individual socio-economic situation of citizens are more important. This is in line with large parts of the literature that emphasise the role of individual characteristics and intergroup relations (Jackson, 1993; McLaren, 2015).

 Table 2 shows the dissatisfaction and distrust of two groups, namely the group that is negative about immigration and group that is positive about immigration (based on the median score of the immigration index presented in Table 1). This way of structuring the data helps to shed light on the degree that immigration impacts on the legitimacy that states have.

Table 2. Types of dissatisfaction and distrust of groups that are negative or positive about immigration per region.

		Dissatisfied with:
Negative or positive attitude towards immigration per region		Democracy (stfdem)	Government (stfgov)	Economy (stfeco)	Life (stflife)	No institutional trust	Democracy is unimportant (imp.dem)	Democraticness is low
North	Negative	31.5	43.1	25.6	28.0	27.1	2.9	49.9
	Positive	12.0	22.7	14.8	23.1	12.1	.7	28.5
	Total	18.0	28.9	18.1	24.6	16.7	1.4	35.0
Anglo-Saxon	Negative	59.2	64.2	70.9	51.1	51.0	6.6	70.8
	Positive	32.8	45.6	56.7	37.8	28.2	2.3	48.1
	Total	47.2	56.0	64.7	45.2	40.9	4.7	60.5
West	Negative	58.8	59.3	57.8	52.7	53.9	5.8	65.1
	Positive	26.7	33.9	31.1	36.4	25.0	1.2	41.4
	Total	40.5	44.9	42.6	43.4	37.4	3.1	51.5
South	Negative	80.7	80.9	86.3	62.6	69.8	6.5	85.1
	Positive	65.7	70.6	78.9	52.0	52.2	2.0	75.9
	Total	72.8	75.5	82.4	57.1	60.6	4.2	80.2
East	Negative	69.7	70.7	67.8	65.0	76.3	7.7	79.3
	Positive	55.6	59.0	52.5	50.8	67.9	4.4	72.0
	Total	62.2	64.5	59.7	57.4	71.9	5.9	75.4
Total	Negative	65.3	66.9	67.9	56.0	59.9	6.3	72.9
	Positive	40.1	47.1	47.8	41.5	37.2	1.9	54.1
	Total	51.6	56.2	57.0	48.1	47.6	3.9	62.6

The results show that around 65 per cent of the respondents with a negative view on immigration is dissatisfied with democracy, government and the economy. In case of citizens with a positive view on immigration, 40 per cent to 48 per cent is dissatisfied. In all regions, except the Nordic countries, the majority of the citizens that is anti-immigration is also dissatisfied with democracy. A minority of the citizens that is pro-immigration is dissatisfied with democracy, except in Southern Europe where the Great Recession has eroded both the specific and diffuse support for the state, as will be documented later on in more detail.

The dissatisfaction with government follows a similar pattern as dissatisfaction with democracy, including the regional variations just discussed, but the disaffection in this case is somewhat higher. Whereas overall 51.6 per cent is dissatisfied with democracy, 56.2 per cent is dissatisfied with government.

Also important is the satisfaction with the economy. Table 2 shows that 68 per cent of the citizens that is negative about immigration is also negative on the economy. On average, the degree of dissatisfaction with the economy is similar to the degree of dissatisfaction with government, but there are some important regional differences. In the Nordic countries the disaffection is low, whereas in the Southern countries it is very high. In the Eastern countries the disaffection with the economy is close to the overall average which may result from the rather different levels of socio-economic development of countries in Eastern Europe (Binelli, Loveless, & Whitefield, 2015). Several Eastern countries did realise economic growth and/or the prospects of realising growth in these countries are relative good. In the Anglo-Saxon countries the degree of dissatisfaction with the economy is slightly above average. These countries are part of the second European region that has been hit hard by the crisis and which effects may potentially weaken state legitimacy, be it less radically than in case of Southern Europe.

The dissatisfaction with life is lower than the dissatisfaction with government and economy and more or less equal to dissatisfaction with democracy. The differences between the two groups are small, meaning that dissatisfaction with immigration is not strongly related to dissatisfaction with life (see also Table 4).

Political trust is an alternative indicator of state legitimacy that focuses on diffuse support of the state. Our trust indicator is an index of three institutional variables (see Appendix). The table shows that trust and immigration are related. In case of institutional trust there are significant differences between the groups that are pro- and anti-immigration. Negative attitudes towards immigration often coincide with a lack of trust in institutions. 60 per cent of the citizens that has a negative view on immigration has no trust in political institutions. There are remarkable regional differences. In the Northern, Anglo-Saxon and Western European countries a majority of the respondents that is positive about immigration does also trust national institutions, whereas this is not the case in Southern and Eastern regions. In case of the Southern region, the severity of the economic recession may play a part in this. In case of the Eastern region, the corruption in some of the countries may invoke less diffuse support and more distrust than in other regions (Kriesi, 2014; McLaren, 2012).

The degree of democraticness indicates the extent to which citizens characterise their national political system as being democratic (Keman, 2014). Citizens in Nordic, Anglo-Saxon and Western regions that are opposed to immigration are more negative about the democraticness of their country than citizens that are positive about immigration. In Eastern and Southern Europe the differences between both groups are much smaller. In most Southern countries this may be caused by the impact of the economic crisis which has hit different socio-economic status groups more or less equally. In some Eastern countries the cause may lay in the corruption that affects the life of all status groups (Berglund, Ekman, Deegan-Krause, & Knutsen, 2013; Kriesi, 2014).

What can we learn from the patterned variations in Table 2? First of all, the attitudes towards immigration are related to the satisfaction with the key components of the ‘good life’ (democracy and welfare). Secondly, there are regional differences in how strong immigration and satisfaction are related. The table shows that the citizens in the countries in each region show a similar degree of dissatisfaction, that is, low levels in Scandinavia, in-between levels in Western Europe and high levels of dissatisfaction in Eastern and Southern Europe. The attitudes of citizens are not only affected by net immigration. Also important is how the perceived level of economic welfare (i.e. the satisfaction with the economy) impacts on the fear of citizens that immigrants might worsen the individual welfare. If citizens have the idea that governments are not able to secure their welfare they become more dissatisfied with government and democracy. This dissatisfaction also spreads out to immigrants, irrespective whether there is a large influx or not. This preliminary result suggests that negative attitudes towards immigration may not be the cause but the result of dissatisfaction with government and democracy. This causal connection cannot be proven empirically, but the pattern in the empirical results suggests that this is a plausible interpretation.

In addition to the cross sectional analysis, a longitudinal analysis may help to find patterned variations that shed light on the relationship between immigration and state legitimacy. Given the importance of the satisfaction with the economy in the explorative results, it is relevant to assess the impact of economic crisis on immigration attitudes.

 Table 3 shows the change in the percentage of citizens that are both anti-immigration and dissatisfied or distrusting national political institutions. The results indicate a minor crisis-effect: before the crisis 60 per cent of the citizens that is negative about immigration is also not satisfied with government whereas during the crisis this percentage is 65 per cent. In case of satisfaction with democracy this percentage rises from 43.4 per cent to 47.4 per cent. Hence, the Great Recession does overall not lead to a significant increase of citizens that are dissatisfied with democracy. This is in line with previous research that has shown that contextual variables, in particular economic ones, have hardly had any impact on anti-immigration attitudes (Schlueter, Meuleman, & Davidov, 2013).

Table 3. The change in the degree to which two groups of citizens before and after the Great Recession are not satisfied or do not trust national political institutions.

		Dissatisfied with:
Region	Negative or positive attitudes towards immigration	Democracy (stfdem)	Government (stfgov)	Economy (stfeco)	Life (stflife)	No institutional trust
North	Neg	−2.8	2.3	1.7	−0.5	−2.5
	Pos	−5.3	−0.7	1.0	−1.4	−3.2
Anglo-Saxon	Neg	−1.7	7.0	35 . 2	1.0	−0.2
	Pos	−2.8	6.5	37.3	−1.4	−2.2
West	Neg	0.9	−1.2	−3.5	−1.6	0.1
	Pos	−3.3	−6.3	−9.4	−3.6	−4.1
South	Neg	15.3	24.0	30.8	0.5	15.4
	Pos	15.6	22.9	31.9	−1.6	12.8
East	Neg	−6.4	−2.7	−3.0	−6.1	−0.5
	Pos	−8.0	−6.4	−7.0	−7.7	−2.6
Total	Neg	2.4	5.2	9.8	−1.3	3.4
	Pos	−0.6	1.3	5.3	−3.7	−0.3

Note: Reading example: In the Nordic countries the percentage of citizens that is negative about immigration before and during the crisis has dropped with 2.8 per cent. The bold numbers indicate the largest changes and they are all significant at the level of p=0.000.

Southern Europe and to a lesser extent the Anglo-Saxon countries are two important exceptions to this. In these regions the dissatisfaction with the economy has risen sharply. In case of Southern Europe the distrust of national political institutions has also risen significantly in case of both groups, which has not happened in case of the other European regions. These results indicate that the support hypothesis is partly confirmed: immigration is related to state legitimacy, but it only affects the diffuse support in countries that are hit hard by the economic crisis. Although there are more fluctuations throughout the years covered by the ESS, an inspection of the trends on a year-by-year basis (not shown here) shows that a rise or decline did often not start with the Great Recession but well before that. All in all, the opposition to immigration does not increase significantly. One possible explanation is that the cultural objections to immigration, which are more or less independent from the economic crisis, overdetermine the economic objections. Additional analysis (not shown here) shows that citizens which think negative about the cultural effects of immigration nearly always also think negative about the economic effects of immigration. If we reverse this, we find a much weaker relationship.

All in all we can conclude that immigration is related to state legitimacy in Europe. This is mainly due to variables at the individual level. If the socio-economic position of a country is strong, citizens are more likely to welcome immigration and to be satisfied with their national democracy. The economic crisis does not lead to an increase in the dissatisfaction with immigration, except in those regions that are most affected by the crisis.

4. Multi-level analysis

The explorative analysis has shown that there is no overall increase in the opposition against immigration in the period investigated. The way immigration matters for the perceived state legitimacy depends mainly on the socio-economic position of citizens. Factors at the macro-level seem to be less important. Since the explorative analysis has examined the relevant variables one by one, the logical next step is to combine them into multi-level models in order to assess their relative causal impact.

Most existing approaches predict that on the demand side, individual characteristics (such as education, age and life satisfaction) do affect the way that citizens perceive the democratic performance of states. Education may enhance a positive attitude towards immigration because it invokes a broader knowledge, higher reflexivity, a critical stance, higher socio-economic status and security, more exposure to foreign cultures and membership of cosmopolitan social networks (Coenders & Scheepers, 2003). If we analyse the relationship between education and anti-immigration in Europe we find considerable regional differences. In the North Western part of Europe the attitudes towards immigration are relatively positive and the differences in anti-immigration attitudes between lower and higher educated citizens are small. In South Eastern Europe the attitudes towards immigration are less positive and the difference between lower and higher educated citizens is large (Coenders & Scheepers, 2003). This pattern indicates that the longer the democratic tradition and the higher the level of education are, the more positive citizens are towards immigration and the less impact the immigration issue has on satisfaction with democracy.

In addition, on the supply side, several macro-level factors such as policies and immigration flows may have an effect on these orientations (Callens, 2015). The models that are discussed below include two distinct levels: A representative sample of individual respondents (level 1) is nested within regional contexts (level 2). Given the use of multi-level data, multi-level regression analysis is an appropriate technique for analysis.³ The independent variables were treated as fixed components, reflecting the weighted average for the slope across all groups, while region was treated as a random component, capturing the regional variability in the slope. The strength of the beta coefficients (slopes) can be interpreted intuitively as how much change in the dependent variable is generated by a one percent change in each independent variable)

 Table 4 shows the relative impact of the relevant explanatory variables in three models with different dependent variables that are related to legitimacy and are central in the explorative analysis. Level 1 consists of the individual respondents and level 2 of the five European regions. The demographical variables age, sex and education are omitted because they are not significant. The variable with the strongest causal impact is satisfaction with economy (stfeco) except in the model with the dependent variable called implvdm (‘How important for you to live in democratically governed country’). This indicates that the perception of the national economic situation has a relatively strong impact on how citizens think about their democracy. The variables at the national level that are related to the politics of immigration (see Table 1 and Appendix) are not significant or have a small causal impact, except the immigration index that measures the overall attitude of citizens towards immigration in their country. This means that immigration forms a relevant dimension that should be taken into account when explaining the attitudes of citizens towards the political system in which they live. This finding is in line with the explorative analysis and also confirms the results by others that immigration matters for the attitudes towards democratic performance of states (see for an overview: McLaren, 2015). The crisis hypothesis is not confirmed as the depth of the economic crisis has had hardly any impact on the dependent variables, except in the Southern region. This outcome confirms that regional differences should be taken into account as the impact of the crisis is not the same across Europe.

Table 4. Multi-level analysis of the factors that impact on the perception of the democratic performance of states.

	How satisfied with the way democracy works in country (stfdem).		How satisfied with the national government (stfgov).		How important for you to live in democratically governed country (implvdm).		Institutional Trust
Parameter	Estimate	Sig.	Estimate	Sig.	Estimate	Sig.	Estimate	Sig.
Intercept	0.819	00.093	1.511	0.012	4.86	.000	1.611	.003
Individual level
copeinc	0.086	0.011	0.095	0.004	0.142	0.000	0.133	0.000
imm_index	0.175	0.000	0.096	0.000	0.213	0.000	0.188	0.000
lrscalec	0.068	0.000	0.141	0.000	−0.007	0.210	0.026	0.000
stflife	0.129	0.000	0.053	0.000	0.107	0.000	0.076	0.000
stfeco	0.424	00.000	0.559	00.000	−0.006	0.326	0.301	00.000
Country level
Policy gap	0.040	0.129	0.123	0.000	−0.094	0.000	−0.075	0.002
MIPEX index	0.016	0.017	−0.012	0.072	0.012	0.070	0.008	0.202
Net immigration	−0.004	0.825	−0.028	0.094	0.176	0.000	0.051	0.001
Depth of economic crisis	0.005	0.060	0.004	0.114	0.039	0.000	0.024	0.000
Intraclass correlation	0.05		0.17		0.10		0.08

Source: ESS. See the Appendix for the operationalisation of the variables.

The Intraclass Correlation indicates how much of the total variance in the dependent variable is explained by the variance between the regions. This correlation is highest in case of the satisfaction with government. This indicates that the citizens in different regions make different assessments of the performance of their government. The correlation is somewhat lower in case of the satisfaction with democracy, probably because all included countries are developed or developing democracies.

5. Conclusions

During the last 10 years Europe has experienced the effects of an enduring economic recession plus an increase in (the public attention for) immigration. In this research we have analysed the relationship between the attitudes of citizens towards immigration in relation to variables that constitute legitimacy. The findings indicate that the attitudes towards immigration are related to the satisfaction of citizens with the performance of the state. Citizens that are negative about immigration are often also negative about the functioning of democracy and the performance of governments and they distrust political institutions.

How do these results relate to the hypotheses? We expected that the immigration issue has weakened the specific support and not the diffuse support unless a country has been hit hard by the economic crisis. This hypothesis is confirmed. The diffuse support has dropped in Southern Europe but not in the other European regions. It was also expected that anti-immigration attitudes fuel distrust in state institutions. The analysis shows that such a connection does exist as feelings about of state legitimacy are affected by concerns about immigration. In regions that are hit hard by the economic crisis this relationship is strongest. We also assumed that the number of immigrants does hardly matter for anti-immigration attitudes. This is confirmed by the analysis which shows that not the number of immigrants, but immigration attitudes matter for how citizens perceive state performance. We hypothesised that a large policy gap will lower the trust of citizens in the state and its capacity to control the borders. This is confirmed by the multi-level analysis, but the estimated effect is relatively small. The crisis hypothesis assumes that during the economic recession, citizens are inclined towards unfavourable attitudes towards immigration because foreigners are perceived as competitors on the labour market and housing market. This is not confirmed by the analysis, except in Southern Europe. In this region the attitudes towards immigration are less positive and the difference between lower and higher educated citizens is large. We also assumed that populism is weakly related to immigration within Europe as a whole. This is confirmed by the analysis which does not show a significant effect.

All in all, the effect of the economic crisis on the degree to which anti-immigration attitudes weaken the specific and diffuse support of states is weak, unless the economic crisis has a strong impact on the national economy and the lives of citizens. Although the crisis-effect is weak, the satisfaction with the economy plays a significant role as it is an important determinant of the support of citizens of the state. The anti-immigration attitudes turn out to be relevant too since they are among the strongest determinants in the multi-level analysis.

Notes

1 The five regions are Scandinavia (Norway, Finland, Denmark, Sweden), Anglo-Saxon Europe (Ireland, UK), Western Europe (Switzerland, Belgium, Netherlands, France, Germany), Southern Europe (Portugal, Spain, Italy) and Eastern Europe (Estonia, Slovenia, Slovakia, Bulgaria, Hungary, Czech Republic, Poland). Excluded are the smallest countries (Iceland, Cyprus, Luxembourg) and the most peripheral countries (Turkey, Ukraine, Turkey). Greece is part of the longitudinal analysis, but not of the cross sectional analysis because it did not participate in the sixth wave of the ESS.

2 The MIPEX-index measures policies to integrate migrants in all EU Member States. Retrieved from: http://www.mipex.eu/ (assessed 10 May 2016).

3 The models in this study use restricted maximum likelihood techniques to estimate direct and cross level effects for hierarchical data. All independent variables were centered, by subtracting the grand mean (which becomes zero).

Appendix

Table 1.

Per cent countries with high levels of immigration (Net_immigration). The division between high and low is based on the median value.

Per cent citizens with negative attitudes towards immigration (imm_index). In ESS (pooled waves 2002 to 2014) three items assess attitudes towards migrants: ‘Would you say it is generally bad or good for [country]’s economy that people come to live here from other countries’, ‘Would you say that [country] cultural life is generally undermined or enriched by people coming to live here from other countries’ and ‘Is [country] made a worse or a better place to live by people coming to live here from other countries’. The mean of three scales for each country or region is taken as an indicator of the anti-immigrant climate.

Per cent countries with restrictive policies (MIPEX). The division between restrictive and non-restrictive is based on the median value.

Per cent countries with a larger policy gap (gap). The division between a large and small gap is based on the median value.

Per cent countries with a large electoral support for populist parties (Populism). The division between a large and small support is based on the median value.

Tables 2, 3 and 4.

Negative or positive attitude towards immigration per region. The division between negative and positive is based on the median value.

Dissatisfied with democracy (stfdem). And on the whole, how satisfied are you with the way democracy works in [country]?

Dissatisfied with government (stfgov). Now thinking about the [country] government, how satisfied are you with the way it is doing its job?

Dissatisfied with economy (stfeco). On the whole how satisfied are you with the present state of the economy in [country]?

Dissatisfied with life (stflife). All things considered, how satisfied are you with your life as a whole nowadays?

Democracy is unimportant (implvdm): How important is it for you to live in a country that is governed democratically?

No institutional trust. Institutional Trust is measured by the mean of the values on three variables. Using this card, please tell me on a score of 0–10 how much you personally trust each of the institutions I read out: Trust in Parliament, Trust in the legal system and trust in the police.

Democraticness is low (dmcntov). How democratic do you think [country] is overall?