Abstract
This article examines the role of party system polarization in shaping immigrants’ party loyalties in their host country. It suggests that foreign-born individuals are more likely to become partisans when political parties take more distinct policy positions on immigration control. Moreover, this relationship is more pronounced among foreign-born non-citizens than foreign-born citizens. Using individual-level public opinion data from eight rounds of the European Social Survey (ESS), 2002–2017, and measures of party system polarization constructed using party policy positions from the Chapel Hill Expert Survey (CHES) in 17 West European democracies, the analyses confirm these expectations. The findings presented here suggest that party polarization on immigration control enhances, rather than undermines, immigrant political integration in contemporary democracies.
Introduction
International migration has transformed the political landscape of West European democracies. Specifically, perceptions of foreigners as constituting a major threat to their host country’s culture and security, as well as intensifying competition for jobs, wages, and social welfare, have triggered the rise of far-right parties and resulted in more anti-immigrant party systems (van Spanje 2010; Alonso and da Fonseca 2012). Moreover, growing politization around immigration (van der Brug et al. 2015; Urso 2018; Dancygier and Margalit 2019) has made it a central issue of electoral politics in Western Europe in recent decades (van der Brug and van Spanje 2009; Kriesi et al. 2012; Hooghe and Marks 2018).
While much scholarly attention has focused on the patterns of party positions on immigration and their determinants (e.g., Green-Pedersen and Odmalm 2008; Green-Pedersen and Krogstrup 2008; van Spanje 2010; Alonso and da Fonseca 2012; Helbling 2014; Pardos-Prado, Lancee, and Sagarzazu 2014; van der Brug et al. 2015; Abou-Chadi 2016; Westlake 2018; Dancygier and Margalit 2019), the consequences of these positions remain little understood. We are still in the dark, for example, concerning how party system polarization on immigration control influences immigrant political orientations and behavior. Given that immigration control policies directly target new arrivals, one key question is to what extent does the distinctiveness of party positions on these policies affect the development of immigrants’ party allegiances in their adopted homeland? Responding to this question is important not only because the numbers of foreign-born individuals in Western Europe are on the rise (Czaika and de Haas 2015, 305) but also because, once adopted, party loyalties are usually durable and inherited by later generations (Campbell et al. 1960; Franklin and Jackson 1983).
This article examines how party system polarization on immigration control influences partisanship acquisition among first-generation immigrants in Western Europe. I expect that foreign-born individuals are more likely to become partisans 1 — that is, to self-identify with a political party — when parties take more distinct positions on immigration control. I suggest that by clarifying choices on these policies, party polarization both enables and motivates immigrants to become partisans in their host country. Conversely, where parties take more similar policy stances, foreign-born residents have more difficulty in identifying a preferred party and developing an allegiance to it.
I expect that this positive relationship between party polarization on immigration control and partisanship acquisition among foreign-born individuals is also moderated by an individual’s citizenship status. Since citizenship reduces immigrants’ vulnerability to restrictive immigration policies (Pantoja, Ramirez, and Segura 2001; Gilbertson and Singer 2003; de Rooij 2012; Aptekar 2016), foreign-born citizens should be less responsive to party polarization on immigration control than foreign-born non-citizens in developing their party loyalties. To test these expectations, I first construct measures of party polarization on immigration control, using information on party policy positions from the Chapel Hill Expert Survey (CHES) data. I then combine these macro-level indicators with individual-level information from eight rounds of the European Social Survey (ESS), 2002–2017, in 17 West European democracies.
This article adds to existing scholarship on partisanship, party polarization, and immigrant political engagement in several ways. First, on a general level, it enhances our understanding of how people develop party allegiances. Although this question has been at the heart of scholarship on people’s political behavior since the mid-20th century (Campbell et al. 1960), we still have a limited understanding of how individuals who resettle to another country acquire party attachments. Existing research suggests that since foreign-born residents usually have a limited understanding of their host country’s politics and lack parental socialization into its political parties, they are less likely to become partisans than native-born individuals (Abrajano and Alvarez 2010; Hajnal and Lee 2011). Yet, the absence of such socialization also means that first-generation immigrants may be highly sensitive to party policies and electoral appeals in developing their party attachments (Alvarez and García Bedolla 2003; McCann and Nishikawa Chávez 2016). This article adds to this scholarship by showing that party polarization on immigration control plays an important role in stimulating partisanship acquisition among foreign-born individuals, especially among non-citizens.
Second, this article contributes to a growing body of research on the consequences of party system polarization (e.g., Hetherington 2001; Dalton 2008; Lachat 2008; Hellwig, Mikulska, and Gezgor 2010; Dahlberg 2013; Lupu 2013, 2015; Moral 2016, 2017). Elite polarization has usually been considered damaging to democratic governance due to its tendency to generate legislative gridlock, reduce economic growth, and weaken regime legitimacy (Sartori 1976; King 1997; Jones 2001; Frye 2002). However, recent studies show that party polarization often encourages people’s political engagement in ways that benefit democratic governance (e.g., Hetherington 2001; Lachat 2008; Dalton 2008; Hellwig, Mikulska, and Gezgor 2010; Dahlberg 2013; Lupu 2013, 2015; Moral 2016, 2017). This article enhances this literature by demonstrating that the distinctiveness of party positions on immigration control has a positive effect on the development of partisanship among first-generation immigrants and, thus, facilitates, rather than obstructs, their political integration in contemporary democracies.
Finally, much current research on party attachments or vote choice among immigrants is based on isolated country cases and examines a limited number of immigrant or ethnic minority groups (e.g., Wong 2000; Wüst 2000, 2004; Alvarez and García Bedolla 2003; Dancygier and Saunders 2006; Teney et al. 2010; Bergh and Björklund 2011; Bird, Saalfeld, and Wüst 2011; Hajnal and Lee 2011; Heath et al. 2011; Sanders et al. 2014; Strijbis, 2014; Huddy, Mason, and Horwitz 2016; McCann and Nishikawa Chávez 2016; Sears, Danbold, and Zavala 2016). By considering a wider set of West European democracies with diverse immigrant populations, this article offers a more generalizable cross-national perspective on the determinants of immigrant party attachments than previous research.
To develop these contributions, the article proceeds as follows. The next section presents my theoretical argument on how and why party polarization on immigration control matters in shaping partisanship acquisition among foreign-born residents and why we can expect citizenship to moderate this relationship. I then introduce the data, indicators, and statistical techniques used in the empirical analyses. The subsequent section presents my results, while the concluding section discusses the implications of this analysis and offers suggestions for future research.
Party Polarization and its Consequences for Partisanship
Scholars of political parties and electoral competition have long been interested in the extent to which parties present distinct policy positions (e.g., Downs 1957; Carmines and Stimson 1989). More diverse electoral alternatives are said to provide more meaningful choices to the electorate and thereby enhance the democratic process (Dalton 2008; Wessels and Schmitt 2008; Dahlberg 2013). In adopting policy stances, parties usually react to constraints and incentives created by political institutions, such as electoral systems, as well as to policy choices of rival parties (Sartori 1976; Cox 1990; Adams, Merril, and Grofman 2005). Parties also respond to structural conditions, such as social cleavages (Lipset and Rokkan 1967), economic circumstances (Sigelman and Yough 1978), and political uncertainty (Lupu and Riedl 2013).
While there is considerable research on the determinants of party system polarization (e.g., Sartori 1976; Sigelman and Yough 1978; Curini and Hino 2012), our understanding of its consequences is more limited and mixed. Some scholars view elite polarization with concern, expecting more policy inaction, government instability, and citizen alienation from their governments when political elites are polarized (Taylor and Herman 1971; Dionne 1991; King 1997; Jones 2001; Hetherington 2007). This perspective is consistent with a view that people generally dislike party confrontation and disagreement, and they prefer public officials to cooperate with one another in solving problems of primary importance to their country (Hibbing and Theiss-Morse 1995, 2002).
However, recent scholarship has questioned whether elite polarization always leads to negative outcomes for democratic governance. Studies find, for example, that more distinct party choices motivate people to become partisans (Hetherington 2001; Lupu 2013, 2015), vote in elections (Brockington 2009; Dalton 2008; Hellwig, Mikulska, and Gezgor 2010; Moral 2017), and engage in other forms of political action (Hellwig, Mikulska, and Gezgor 2010). There is also evidence that party polarization enhances people’s satisfaction with democracy in some circumstances (Harding 2011, but see Dassonneville and McAllister 2020) as well as generates more accurate perceptions of party positions (Dahlberg 2013), more consistent ideological voting (Ensley 2007; Dalton 2008; Lachat 2008), and fewer split-ticket and invalid ballots (Hetherington 2001; Jacobson 2005; Moral 2016). In short, then, there is growing evidence that party polarization boosts some forms of mass political behavior that benefit, rather than undermine, democracy.
Building on the insights of these studies, I posit that party polarization also matters for immigrant identification with political parties in their host country. Specifically, I expect that the distinctiveness of party positions on policies that directly target immigrants — that is, immigration control — contributes positively to partisanship acquisition among foreign-born residents. Existing research suggests that party polarization clarifies choices for ordinary individuals and thereby facilitates and motivates their partisanship acquisition (Layman, Carsey, and Horowitz 2006, 102; Lupu 2013, 2015). Specifically, the further apart the political parties, the more people can distinguish among them, identify a party that is closer to their views, and eventually develop an allegiance to it. In contrast, where or when parties take similar issue stances and, therefore, exhibit agreement rather than conflict over policies, parties may become irrelevant to people (Hetherington 2001, 623; Holmberg 1994, 110), reducing their incentives to develop party loyalties.
This expectation does not depend on whether partisanship is conceptualized as a social identity (Campbell et al. 1960) or a cognitive cue (Fiorina 1981). According to a social identity perspective that considers partisanship as an individual’s affective attachment to a party resulting from early socialization and persisting over one’s lifetime (Campbell et al. 1960; Green, Palmquist, and Schickler 2002), party polarization enhances partisanship acquisition because it allows individuals to better distinguish between party brands (Lupu 2015, 332). In comparison, a rational actor perspective viewing partisanship as a cognitive cue — a kind of “running tally” of party performance (Fiorina 1981) — similarly suggests that when parties adopt more distinct policy stances, people derive more benefit from choosing one party over another and are, thus, more likely to become partisan.
While not directly linked to the literature on partisanship, existing research on vote choice based on spatial (proximity) models (e.g., Downs 1957; Davis, Hinich, and Ordeshook 1970; Enelow and Hinich 1984) or directional models (e.g., Rabinowitz and Macdonald 1989; Macdonald, Listhaug, and Rabinowitz 1991) also concludes that elite polarization may have positive electoral outcomes. According to this literature, party system polarization is often rewarded by voters because it clarifies party stances, provides more meaningful alternatives to the electorate (Cho and Endersby 2003; Dinas and Pardos-Prado 2012), and offers guidance in opinion formation (Iversen 1994). Moreover, party polarization enables citizens to pull policies in a desired direction where the policies of individual parties are watered down by multi-party bargaining and power-sharing institutions (Kedar 2005).
The capacity of party polarization to reduce uncertainty about party policy positions and identify a preferred party means that polarization has an instrumental value to people. However, the distinctiveness of party offerings may boost partisanship also because there is an intrinsic value in having choice. Some scholars find that individuals derive more satisfaction when they select an option from among alternatives as opposed to having only one option, even when this option is exactly what they want (Sen 1988; Przeworski 2003; Harding 2011, 223). Experimental research confirms that people value having choices, particularly in Western societies (Iyengar and Lepper 1999) and when these choices do not overwhelm people’s abilities to evaluate them (Iyengar and Lepper 2000). In short, whether for instrumental or intrinsic reasons, party policy distinctiveness can be expected to have a positive effect on people’s partisanship acquisition.
Party Polarization on Immigration Control and Partisanship among Immigrants
While most research on party polarization focuses on a unidimensional left-right continuum (e.g., Dalton 2008; Lupu 2015), one particularly relevant policy dimension for foreign-born individuals is immigration control. Besides immigrant integration policies, immigration control directly targets immigrants (Givens and Luedke 2005; Duncan and van Hecke 2008). Although this article focuses on foreign-born individuals already residing in their host country, immigration control policies are not irrelevant to them (Schildkraut 2013; Street, Zepeda-Millán, and Jones-Correa 2015; McCann and Nishikawa Chávez 2016; Yang and de la Garza 2017). These policies may determine such residents’ long-term ability to remain in their host country or, if they have left their families behind, influence the prospects of their loved ones joining them in the future. Moreover, having migrated themselves, foreign-born individuals may develop a sense of kinship or solidarity with other newcomers (Just and Anderson 2015; Street, Zepeda-Millán, and Jones-Correa 2015). There is also evidence that while some immigrants are concerned that continued migration might result in their status degradation and greater negative stereotyping about immigrants (Jiménez 2007; Strijbis and Polavieja 2018), others welcome new arrivals because growing immigrant communities may create new economic opportunities for their group members and may enhance their political and cultural influence in the adopted homeland (Jiménez 2007). In short, while immigrants may vary in their support for open borders in their host country, they have strong reasons to pay attention to its immigration control policies.
Consistent with this perspective, existing research shows that immigrant identity is deeply salient to foreign-born individuals, providing a critical lens through which they view and evaluate their receiving nation’s politics. In a study of Latino immigrants in the United States based on focus groups, Schildkraut (2013) found that when thinking about democracy in their adopted homeland, foreign-born residents consistently invoked their identity and experiences as immigrants rather than as members of their national-origin group or ethnic minorities, and did so even when questions made no mention of immigration (35–37). Moreover, many first-generation immigrants noted that they cared greatly whether their representatives paid attention to immigrants’ needs and policy preferences and that these policies were not generic, such as education, crime, or the economy, but immigration specific, such as amnesty and deportation (Schildkraut 2013, 38; see also Pantoja, Ramirez, and Segura 2001; Bowler, Nicholson, and Segura 2006; Abrajano and Alvarez 2010, 37; Collingwood, Barreto, and Garcia-Rios 2014; Street, Zepeda-Millán, and Jones-Correa 2015).
Thus, if immigrants indeed care about immigration control, they are likely to know how their host country’s parties align along this policy dimension. Extant research suggests that although party systems in Western Europe have become more anti-immigrant since the mid-1970s, parties vary in their positions on immigration (Alonso and da Fonseca 2012; van der Brug et al. 2015). Compared to right-wing parties, left-of-center parties generally favor more liberal immigration policies (Helbling 2014; Urso 2018), particularly in countries with substantial ethnic minorities and single-member-district electoral systems (Westlake 2018). However, there is also evidence that left-wing parties often moderate their positions in competitive elections to avoid losing votes to other parties (Pardos-Prado, Lancee, and Sagarzazu 2014; Abou-Chadi 2016; Westlake 2018) or in their effort to join coalition governments (Green-Pedersen and Odmalm 2008; Green-Pedersen and Krogstrup 2008; Urso 2018).
As parties adjust their policy positions on immigration control in response to their vote- and office-seeking motivations, they may produce varied levels of party polarization on this issue across countries and over time. This polarization may in turn provide different opportunities and incentives for first-generation immigrants to develop party loyalties in their adopted homeland. Since more polarization on immigration control means more distinct and, thus, more meaningful choices on policies that directly target new arrivals, foreign-born individuals are more likely to become partisans when political parties are more polarized. Conversely, when parties take similar positions on immigration control, foreign-born residents should be less likely to develop identification with their host country’s parties.
Contingent Effects of Party Polarization: The Role of Citizenship
Besides examining the direct effects of party polarization on partisanship acquisition among first-generation immigrants, I am also interested in the extent to which this relationship is moderated by an individual’s citizenship status. I expect that being formally integrated into one’s host country as a citizen reduces the extent to which foreign-born residents respond to party polarization on immigration control in adopting their party identification. In other words, party polarization’s positive effect on partisanship acquisition should be less pronounced among foreign-born citizens than foreign-born non-citizens.
These expectations are based on several findings in existing research. By differentiating between outsiders and insiders, citizenship provides an instrumental and symbolic value to individuals who acquire it (Joppke 2007; Bloemraad, Korteweg, and Yurdakul 2008). Instrumentally, citizenship offers immigrants formal privileges, such as legal protection against deportation, easier international travel and family reunification, the right to vote in national elections, and access to wider welfare benefits and job opportunities (Balistreri and Van Hook 2004; Birkvad 2019; Harpaz and Mateos 2019). Since these entitlements reduce immigrants’ vulnerability to restrictive immigration policies (Pantoja, Ramirez, and Segura 2001; Gilbertson and Singer 2003; de Rooij 2012; Aptekar 2016), the extent to which parties offer distinct choices on immigration control should matter less to citizens than to non-citizens among foreign-born residents. 2
Citizenship may also moderate the relationship between party polarization and partisanship acquisition for more symbolic reasons. Naturalization is sometimes considered an expression of belonging or loyalty to a country (e.g., Birkvad 2019; Pogonyi 2019; Donnaloja 2020). In line with this view, some studies suggest that naturalization motivates immigrants to become more attentive to their adopted homeland’s political and socio-economic realities in forming their political orientations and policy preferences (Röder and Mühlau 2011; Just and Anderson 2015). Thus, because citizenship transforms immigrants into full political members of their host society and provides them with a sense of having a stake in it, foreign-born citizens may consider a wider range of issues than just immigration control in forming their party allegiances. In short, for both instrumental and symbolic reasons, citizenship should moderate the extent to which party polarization on immigration control influences party loyalties among first-generation immigrants. Specifically, party polarization on immigration control should be more strongly linked to partisanship acquisition among foreign-born non-citizens than foreign-born citizens.
Data and Methods
To test my expectations, I use individual-level data from the ESS 2002–2017 (1–8 rounds) — a project that is well known due to its exemplary standards of cross-national data collection and survey design (Kittilson 2009). The survey includes questions related to respondents’ foreign-born status, citizenship, and duration of stay in their host country, along with standard measures of partisanship. It employs nationally representative samples obtained using strict random sampling of individuals regardless of nationality, citizenship, or language. 3 Existing research shows that shares of foreign-born populations and their composition with respect to origin region in the ESS data closely mirror their host country’s official statistics (de Rooij 2012, 461; Just and Anderson 2014, 945). 4 My sample contains 17 West European democracies with relatively large immigrant populations: Austria, Belgium, Cyprus, Denmark, Finland, France, Great Britain, Germany, Greece, Ireland, Italy, the Netherlands, Norway, Portugal, Spain, Sweden, and Switzerland. Pooling data across countries and survey rounds generates 79 macro-level observations (country-rounds) for which information on all relevant variables was available.
Dependent Variable
Party identification is based on the following survey question: “Is there a particular political party you feel closer to than other parties?” This item is widely used in existing cross-national research on partisanship acquisition (e.g., Budge, Crewe, and Farlie 1976; Holmberg 1994, 2003). Individual answers were coded dichotomously, with one indicating a positive response and zero, a negative one. The data show that 39.5% of foreign-born individuals aged 16 years or older identified with a political party in their host country (compared to 53.2% among native-born respondents).
Key Independent Variables
Party polarization on immigration control was constructed using information on party policy positions from the CHES project (Bakker et al. 2015). The CHES is the longest-running, most extensive expert survey of political parties in European democracies that employs multiple experts in each country to evaluate their national parties (Polk et al. 2017). 5 Expert surveys provide valid and reliable estimates of where political parties stand on various policy issues (Marks et al. 2007; Steenbergen and Marks 2007; Hooghe et al. 2010). The reliability and validity of CHES measures have also been confirmed with respect to party positions on immigration (Ruedin and Morales 2019).
To construct my key variable of interest — party polarization on immigration control — I relied on the CHES “Trend File” covering the national elections in European Union (EU) countries from 1999 to 2014. Moreover, I was able to include several recent elections, using the 2017 CHES survey, and two non-EU countries (Norway and Switzerland), using the 2014 survey. The CHES project measures party positions on immigration control on a scale from 0 to 10, where 0 indicates that a party strongly opposes a tough (restrictive) policy while 10 denotes an orientation of strongly favoring such a policy. I employed party positions on this item in national legislative elections preceding ESS interviews in each country-round. Party polarization on immigration control was computed following Dalton’s (2008) formula: Polarization=SQRT{Σ((party vote share
i
)*([party policy score
i
– party system average policy score]/5)2)},
where i represents individual parties. This formula takes into account party policy positions and party relevance in a party system by weighting party positions by their vote shares, since a large party at an extreme end of a policy dimension indicates greater polarization of a party system than a splinter party in the same position. 6 The lowest possible value on this variable is 0, when all parties occupy the same position on the policy scale, and the highest value is 10, when all parties are split between the scale’s two extremes (Dalton 2008, 906). Looking at the data reveals that party polarization on immigration control ranges from 0.96 in Ireland (2011) to 5.28 in Belgium (2010), with a mean value of 3.79 (for details, see Table A1 in the Supplemental Appendix).
To identify foreign-born individuals, I relied on the ESS question: “Were you born in this country?” Individual answers were coded dichotomously, with a value of 1 indicating a positive answer, and 0, a negative one. To ensure that only foreign-born individuals of immigrant background are included in my sample, I additionally employed survey items measuring whether both respondent’s parents are native-born and excluded foreign-born respondents with two native-born parents from the sample. Finally, citizenship was measured using the survey question “Are you a citizen of this country?” and coded dichotomously, with 1 indicating a positive response and 0 – otherwise.
Control Variables
To better isolate the effects of my key variables of interest, the models presented here control for party system mean on immigration control along with party system left-right polarization and party system left-right mean (each computed using party positions weighted by party vote shares). The left-right scale is a commonly used summary indicator of the cleavages and issues that shape political contestation in each country (e.g., Fuchs and Klingemann 1989; Huber and Inglehart 1995). Furthermore, since party system fragmentation may play an independent role in shaping people’s motivations to become partisans (Lupu 2015), my models account for the effective number of electoral parties (Laakso and Taagepera 1979). To facilitate the interpretation of results, all macro-level variables measuring host-country characteristics are centered around their means.
My individual-level controls include respondents’ socio-economic status in the form of education, income, unemployment, and manual skills, as higher socio-economic status generally enables and motivates people’s political engagement (Verba, Schlozman, and Brady 1995; Verba, Nie, and Kim 1971). 7 I also account for respondents’ gender because women are sometimes less politically active due to socialization into traditional gender roles, family and household responsibilities, and more limited socio-economic resources available to them in comparison to men (Jennings 1983; Burns, Schlozman, and Verba 1997; Kam, Zechmeister, and Wilking 2008). Age is another relevant control because political participation usually increases with age, although it weakens and even reverses among particularly old individuals (e.g., Niemi et al. 1985).
Beyond standard socio-demographic characteristics, political engagement tends to be higher among those who are more socially connected (Cain, Kiewiet, and Uhlaner 1991; Giugni and Grasso 2020). I, therefore, control for respondents’ marital status, union membership, and frequency of meeting with others. Attending religious services can similarly stimulate political action, as it offers opportunities for social interaction and exposure to political messages in religious institutions (Calhoun-Brown 1996; Jones-Correa and Leal 2001; Just, Sandovici, and Listhaug 2014). Differences across religious groups are additionally accounted for using dummy variables for being a Christian, Muslim, Jewish, or a believer of some other faith.
Because individuals with left-wing orientations are more politically active than those with right-wing views across a wide range of countries (Dalton, Van Sickle, and Weldon 2010), I include respondents’ left-right self-placement. Furthermore, social grievances are captured using perceived discrimination (Cain, Kiewiet, and Uhlaner 1994) and being a crime victim (Bateson 2012). Finally, among immigrant-specific characteristics, I control for duration of stay and proficiency in at least one of the host country’s official languages (Wong 2000; White et al. 2008; de Rooij 2012). Pre-migration experiences are further captured using measures of democracy, economic development, and autocracy type in the origin country before immigrants’ arrival (Just 2019), while host country’s pro-immigrant opinion climate has been shown to be a relevant feature of post-migration environment (Just and Anderson 2014). (Descriptive statistics for all variables are available in Table A3 in the Supplemental Appendix.)
Analysis and Results
Figure 1 plots the shares of foreign-born partisans by the level of party polarization on immigration control across all country-elections in my sample. The graph reveals a positive and statistically significant relationship between the two variables, with a Pearson’s correlation of .34, statistically significant at 0.05 level (two-tailed). Thus, the higher the level of party polarization on immigration control, the larger the share of foreign-born residents who identify with a party in their host country. This aggregate pattern is consistent with the idea that party distinctiveness on immigration policies is positively linked to partisanship acquisition among first-generation immigrants and, thus, provides preliminary evidence in support of my expectations.

Party Identification among Foreign-Born Individuals and Party Polarization on Immigration Control in 43 West European Country-Elections, 2002–2017.
To test this relationship at the individual level, I combine the micro-level survey data with macro-level information on the host country’s national legislative elections in each country-round. To avoid statistical problems associated with such data structure, such as non-constant variance, clustering, and incorrect standard errors (Steenbergen and Jones 2002), I employ multi-level statistical techniques. Since my dependent variable is dichotomous, the results presented here are multi-level (random intercept) logistic regression estimates (reported as log odds with their standard errors in parentheses and odds ratios in italics).
The results, presented in Table 1, reveal that party polarization on immigration control is consistently positive and statistically significant in predicting partisanship acquisition among first-generation immigrants. The polarization coefficient remains largely the same whether we look at foreign-born respondents with at least one foreign-born parent (Model I) or focus more narrowly on foreign-born individuals with two foreign-born parents (Model II). Hence, the results confirm that newcomers are more likely to identify with a political party if their host country’s parties take more distinct policy positions on immigration control. However, there is also evidence that this relationship is moderated by an individual’s citizenship status (Model III). Specifically, the interaction term between polarization and citizenship is negative and statistically significant, indicating that foreign-born citizens respond to party polarization on immigration control less strongly than foreign-born non-citizens in developing their party attachments.
Party System Polarization on Immigration Control, Citizenship, and Partisanship Acquisition among Foreign-Born Individuals in 17 West European Democracies, 2002–2017.
Note: These are multilevel (random-intercept) logistic regression estimates obtained using the Stata’s xtlogit command; numbers in italics indicate odd ratios; standard errors are listed in parentheses; *p≤.05, **p≤.01, ***p≤.001(two-tailed). Those who arrived in their host country 6-10 years ago is the reference category for other periods of arrival, secular individuals - for religious groups, and ESS round 5 – for other survey rounds.
To examine this conditional relationship in more detail, I plot the marginal effects of my key variables of interest with 95% confidence intervals (Berry, Golder, and Milton 2012). Using the results of the interaction model (Model III in Table 1), Figure 2a presents the marginal effects of party polarization on partisanship acquisition for citizens and non-citizens among foreign-born individuals. Figure 2b subsequently provides citizenship’s marginal effects as we move from -1 to +1 standard deviation from the polarization’s mean. 8
The results in Figure 2a reveal that party polarization’s marginal effect on the probability of partisanship acquisition is positive and statistically distinguishable from zero for all first-generation immigrants. However, as expected, this effect is larger for non-citizens than for citizens: the respective values are .081 and .059. Figure 2b provides further evidence that party polarization on immigration control and citizenship interact in shaping partisanship acquisition among foreign-born respondents. It shows that citizenship’s marginal effect is positive and statistically significant at all levels of party polarization on immigration control. However, this effect drops from .082 to .036 as we move from -1 to +1 standard deviation from the polarization’s mean. Taken together, the results confirm that party polarization on immigration control is positively linked to partisanship acquisition among first-generation immigrants and that this relationship is more pronounced among foreign-born non-citizens than foreign-born citizens. 9

Marginal Effects of Party Polarization on Partisanship Acquisition among Foreign-born Citizens and Non-Citizens.

Marginal Effects of Citizenship on Partisanship Acquisition Among Foreign-born Individuals by Party Polarization on Immigration Control.
Polarization and Party Choice Among Immigrant Partisans
While the aforementioned results indicate that party polarization on immigration control contributes positively to immigrant partisanship acquisition, they do not tell us with which parties foreign-born individuals are more likely to identify. One possibility is that newcomers become attached to pro-immigration parties. As noted earlier, immigrants generally favor liberal immigration policies either because they personally gain from these policies or because they feel kinship and solidarity with other newcomers (Jiménez 2007; Just and Anderson 2015; Huddy, Mason, and Horwitz 2016; Yang and de la Garza 2017, 715). If this perspective is correct, we should find a positive relationship between party polarization on immigration control and identification with a pro-immigration party among foreign-born partisans.
However, it is also possible that party polarization on immigration control divides and polarizes immigrants, pushing them in different directions as they align with parties in their host country. Existing research shows that while foreign-born residents are on average more pro-immigration than their native-born counterparts, there is considerable heterogeneity in immigrant attitudes toward immigration (Branton 2007; Jiménez 2007; Rouse, Wilkinson, and Garand 2010; Knoll 2012; Just and Anderson 2015; Braakmann, Waqas, and Wildman 2017). If this heterogeneity is large and newcomers choose parties closer to their views, the impact of party polarization on identification with pro-immigration parties may be weak or even insignificant.
To test these expectations, Table 2 reports the results of my models focusing exclusively on partisans among foreign-born individuals. The dependent variable in these models is identification with a pro-immigration party. To construct this variable, I used the survey item indicating with which party a respondent identifies and matched it with information on party positions on immigration control from the CHES data. If an individual reported feeling close to a party that favored a more liberal position on immigration control than the party system mean (weighted by party vote shares) in his/her host country, this respondent was coded as identifying with a pro-immigration party. Alternatively, if a respondent indicated feeling close to a party that favored a more restrictive policy on immigration control than the party system mean, this respondent was coded as identifying with an anti-immigration party (for more detail, see Table A2 in the Supplemental Appendix). 10
Party System Polarization on Immigration Control and Identification with a Pro-Immigration Party among Foreign-Born Partisans in 17 West European Democracies, 2002–2017.
Note: These are multilevel (random-intercept) logistic regression estimates obtained using the Stata’s xtlogit command; numbers in italics indicate odd ratios; standard errors are listed in parentheses; *p ≤ .05, **p ≤ .01, ***p ≤ .001(two-tailed). Those who arrived in their host country 6–10 years ago is the reference category for other periods of arrival, secular individuals - for religious groups, and ESS round 5 – for other survey rounds.
My analyses reveal that first-generation immigrants were more likely to identify with a pro-immigrant party in party systems that were more polarized on immigration control. The polarization’s coefficient was positive and statistically significant for foreign-born residents with at least one foreign-born parent (Model I) and when focusing more narrowly on foreign-born individuals with two foreign-born parents (Model II) among partisans. Taken together, the results confirm that when parties take more distinct policy positions on immigration control, first-generation immigrants are more likely to not only develop party attachments in their host country but also identify with a pro-immigration party.
Robustness Checks
To test the robustness of my findings, I conducted several additional analyses. While my main models controlled for party polarization on the left-right continuum, some scholars suggest distinguishing between the economic left-right and social (gal-tan) dimensions (Marks et al. 2006). 11 I, therefore, re-estimated my models while controlling for party polarization on these alternative dimensions instead of using the overall left-right scale (see Tables B1a-b in the Supplemental Appendix). I then provide the results with party polarization on European integration as an additional dimension that shapes political contestation in contemporary Europe (Gabel 2000; Bakker, Jolly, and Polk 2012) (Tables B2a-b). The results of these models revealed no change in my key findings. Furthermore, I found that citizenship interacts with party polarization on immigration control, but not with party left-right polarization (Table B3), confirming immigration control’s relevance in structuring partisanship acquisition among first-generation immigrants.
There is also a possibility that the effects of party polarization on immigration control are not exclusive to foreign-born individuals but extend to native-born residents as well. I, therefore, re-estimated my models with an additional control for the percentage of partisans among natives in the models of partisanship acquisition (Table B4a) and the percentage of pro-immigration party identifiers among natives in the models of pro-immigration party partisans (Table B4b). I found that although the coefficients of party polarization on immigration control were somewhat reduced in substantive terms, they remained positive and statistically significant.
Since my empirical analyses rely on data from Western Europe, I additionally examined whether my results are sensitive to the inclusion of EU citizenship. I found that having citizenship of an EU member state had no impact on the patterns of partisanship among foreign-born individuals, while my key findings remained unchanged (Tables B5a-b). Another possible concern is that sample sizes of foreign-born respondents in some country-rounds are too small to generate reliable estimates. Dropping country-rounds with fewer than 100 observations from the sample, however, did not alter my main findings in any way (Tables B6a-b; for information on sample sizes by country-round, see Table A4).
While my main results were estimated using multi-level statistical techniques with individuals nested within host country-rounds, I considered an alternative multi-level model specification with cross-classified intercepts for both migrants’ host and origin country but found no change in my key findings (Tables B7a-b). Finally, I tested my findings’ validity by re-estimating the models with “jackknifed” standard errors. The results of this resampling technique that omits one macro-level unit at a time showed that my main results remained essentially the same (Tables B8a-b). Taken together, these additional analyses confirm that party system polarization on immigration control plays an important role in shaping the patterns of partisanship among foreign-born individuals and that these findings are robust to different model specifications and statistical techniques.
Conclusion
Party loyalties have long occupied the center stage in democracies. They motivate people to engage politically, guide their electoral choices, and act as an efficient heuristic in making sense of a country’s governance. Scholars have therefore been increasingly interested in how foreign-born residents develop party allegiances in their adopted homeland (Alvarez and García Bedolla 2003; Dancygier and Saunders 2006; Abrajano and Alvarez 2010; Hajnal and Lee 2011; Street, Zepeda-Millán, and Jones-Correa 2015; Huddy, Mason, and Horwitz 2016; McCann and Nishikawa Chávez 2016). While high levels of non-partisanship among new arrivals are usually attributed to immigrants’ limited familiarity with their host country’s politics (Hajnal and Lee 2011), existing research has paid little attention to what may mitigate this information barrier and thereby enhance partisanship acquisition among foreign-born residents. This article suggests that, by clarifying party stances on policies that directly target immigrants, party polarization on immigration control both enables and motivates partisanship acquisition among first-generation immigrants. Thus, party polarization on immigration control contributes positively to immigrant political integration in their adopted homeland.
My analyses confirm that foreign-born individuals are not insensitive to their host country’s political environment and become partisans in response to the distinctiveness of party policy positions. Moreover, the results show that the relevant policy dimension for measuring polarization is immigration control, rather than the left-right continuum. 12 These findings are in line with previous studies showing that immigration control is of central importance in shaping immigrants’ political orientations (Pantoja, Ramirez, and Segura 2001; Bowler, Nicholson, and Segura 2006; Street Abrajano and Alvarez 2010; Schildkraut 2013; Collingwood, Barreto, and Garcia-Rios 2014;, Zepeda-Millán, and Jones-Correa 2015) and that this policy dimension does not easily align with the left-right divide (Junn and Masuoka 2007; Hajnal and Lee 2011).
This article also expands our understandings of citizenship’s political consequences. Consistent with previous research (Wong 2000; Hajnal and Lee 2011; Sears, Danbold, and Zavala 2016), I find that citizenship is positively linked to partisanship acquisition among foreign-born individuals. However, there is also evidence that citizenship interacts with polarization, as foreign-born non-citizens are more likely to become partisans in response to party polarization on immigration control than foreign-born citizens.
While the available data indicate that these findings are robust, the empirical analyses have several limitations. Given the ESS data’s cross-sectional nature, I could not rule out the possibility of a reciprocal relationship between party polarization on immigration control and partisanship acquisition among first-generation immigrants. Future studies using panel data could shed more light on this issue. Another shortcoming is that the data do not allow us to distinguish between authorized and unauthorized immigrants among foreign-born non-citizens. Whether foreigners who are ineligible for naturalization due to their legal status respond to party polarization on immigration control in the same way as authorized immigrants is a question that deserves further research.
Paying closer attention to elite polarization on immigration control may also enable researchers to address some of the existing puzzles in the scholarship on immigrant political behavior. For example, whether political threat mobilizes immigrants or leads to their political alienation has so far received mixed empirical evidence (Jones-Correa, Al-Faham, and Cortez 2018). One explanation of this puzzle could be that immigrants’ response to threat depends on party system polarization on immigration. Lacking distinct policy alternatives, newcomers may see little opportunity to counteract the impending threat and remain politically inactive.
Besides polarization, other features of political context may facilitate immigrants’ partisanship acquisition as well. Some studies suggest, for example, that compulsory voting creates information-rich political environment that consequently boosts partisanship among ordinary individuals, including immigrants (Jiang 2016, 430). Stable party positions and low or moderate party system fragmentation may similarly help new arrivals identify parties closer to their policy preferences and to become partisans. These conjectures suggest the need for more research on how political context shapes immigrant political incorporation in contemporary democracies.
Supplemental Material
Supplemental Material, Online_Appendix - Party System Polarization, Citizenship, and Immigrant Party Allegiances in Western Europe
Supplemental Material, Online_Appendix for Party System Polarization, Citizenship, and Immigrant Party Allegiances in Western Europe by Aida Just in International Migration Review
Footnotes
Notes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
Supplemental Material
Supplemental material for this article is available online.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
