What Does It Take for Immigrants to Join Political Parties?

Political parties are crucial agents in democratic representation and political integration of persons of immigrant origin, a growing category of citizens in the European Union. Research demonstrates that citizens of immigrant origin are less likely to join political parties than persons without a migratory background. Nevertheless, party membership varies across countries and between immigrants. Accounting for such inter‐individual and cross‐national variations, this article uses secondary data from the European Social Survey, the Migrant Integration Policy Index, and the Varieties of Democracy (V‐Dem) project for 25 European democracies to uncover mechanisms that explain the party membership of immigrants. In our multilevel analysis, we test interactions between country‐specific variations in legislation on migration policies on the one hand and individual differences in political socialisation and political efficacy on the other. Our models suggest significant positive effects of exposure to a democratic regime in the country of origin and of internal efficacy on party membership of citizens of immigrant origin. Additionally, our findings highlight the significance of an inclusive national framework for immigrant integration, serving as a moderator to diminish the impact of political socialisation in less democratic countries on the decision of citizens with immigrant backgrounds to participate in political parties within their country of residence.


Introduction
Political parties are crucial agents of democratic representation and governance (Müller, 2000;Powell, 2004) and important gatekeepers in the process of elite recruitment.Party membership has been argued to be a central mechanism linking citizens to the state and a source of the democratic legitimacy political parties have in representative democracies (Gauja & van Haute, 2015;Scarrow, 2014;van Biezen et al., 2012).Despite the decreasing number of party members and the challenges traditional party membership faces (van Biezen et al., 2012;van Biezen & Poguntke, 2014), party membership continues to play an essential role in political representation and democratic legitimacy.
Research has demonstrated that party members compose a rather homogenous group of citizens that does not mirror the diversity in the population (Achury et al., 2020;Angenendt, 2023).Focusing on citizens of immigrant origin (CIO), characterized by their diversity in terms of ethnicity, religion, race, and language, studies have identified barriers immigrants face while incorporating themselves into political parties and becoming elected representatives (Dancygier et al., 2015;Zapata-Barrero, 2017).Though party membership is only one of the many forms of political participation, it is one of the most important routes to high political office and direct policy influence (Gauja & van Haute, 2015;Scarrow, 2014).Thus, immigrants' shortfall of party membership in their countries of residence may hinder their substantive and descriptive representation (da Fonseca, 2011).
Moreover, gaps in party activism between immigrants and natives indicate political inequalities between these groups and challenges for the political integration of immigrants.
Seminal studies of party membership have examined the general motivations for party activism (Seyd & Whiteley, 1992, 2002;Whiteley & Seyd, 2002).While these analyses have successfully uncovered motivations for party membership, they have generally not focused on socio-demographic sub-groups beyond gender and age.Although there is no reason to doubt that party members of immigrant origin are motivated by many of the same general incentives found for members from the autochthonous population, research suggests that the experience of belonging to an ethnic minority or immigrant group may affect a person's level and style of party activism (e.g., Arslan, 2011;Cyrus, 2008).Knowing that, in general, migration background reduces the likelihood of political action (Morales & Giugni, 2011;Ramakrishnan & Espenshade, 2001;Soininen & Bäck, 1993), party membership that is already rare for the autochthonous population can be particularly discouraging for immigrants.Even though existing studies of immigrant experiences add considerable information on individual perceptions, experiences, and motivations of party activists with immigrant origin, they are usually based on limited observations specific to particular individuals or groups in time.Hence, neither the larger-scale general membership studies nor the small-n studies of immigrant-origin members provide any information on the factors affecting the transition from non-member to membership status among immigrants and their descendants.Not least, they tell us very little about the effect macro-level covariates have on immigrants' decision to join political parties.
Given the small number of party members of immigrant origin and the lack of documentation in membership registers, modelling the decision of CIO to become formal party members has been difficult.This article seeks to shed light on this question.Relying on secondary data from the European Social Survey (ESS), the Migrant Integration Policy Index (MIPEX), and the Varieties of Democracy (V-Dem) project for 25 European democracies, we break new ground in modelling the choices to join a political party in the country of residence or not, made by CIO.Employing multi-level regression models, we combine (a) macro-level structural information on the countries of ancestral origin as well as the countries CIO currently live in with (b) individual-level data capturing their political socialisation and psychological factors recognised in the literature on political participation.Hence, building on established models of political participation, we develop an innovative approach to understanding the mechanisms behind the party membership of immigrants.This article complements Finn and Ramaciotti's (2024) article, which unpacks the factors shaping the legal frameworks defining the political opportunity structures for immigrants and adds a layer to how institutional frameworks can shape the political participation of CIO.Furthermore, it aligns with the work of Gherghina and Basarabă (2024) underlying the importance of the experience immigrants have in both their country of origin and their country of residence and further demonstrates how such experiences can shape the party membership of immigrants.

Party Membership of CIO: Interactions Between Macro and Micro Factors
Studies on ethnic minorities in political parties have provided great knowledge on the importance of incorporation of minority groups in political institutions, the complex multilevel factors associated with ethnic representation in electoral bodies (Saggar, 2000), as well as ethnic mobilisation in mainstream and ethnic political parties (Buta & Gherghina, 2023;Chandra, 2011).Research has demonstrated that descriptive representation of ethnic minorities mobilises citizens with similar demographic characteristics (Birnir, 2007).Despite the differences between CIO and ethnic minority citizens, these studies provide insight into the challenges minority groups face in political parties in ensuring political representation.
Political parties vary in the extent to which their values and policies support or oppose immigration and the interests of immigrants and their descendants (van der Brug et al., 2015), as well as the extent to which they encourage and foster the involvement of immigrants within their organisations (Alba & Foner, 2009, 2015).
An essential factor in explaining the immigrant representation gap is political parties and party gatekeepers in particular (Dancygier et al., 2021;Höhne et al., 2023).The supply-demand factors used to explain the underrepresentation of minority groups underline the role demands of selectors and the characteristics of those aspiring to be representatives (Norris & Lovenduski, 1993).Extensive research has demonstrated the undermining effect individual factors have on immigrants' decision to take political action.The limited access to resources such as time, money, education, civic skills, or social capital for the majority of immigrants has been used to explain some of the inequalities in political participation between immigrants and native citizens (Leighley & Vedlitz, 1999;Ramakrishnan, 2005;Verba et al., 1993).There is little reason to assume that involvement in a political party, a particularly resource-intensive form of political participation, would not follow patterns of inequality.In addition to resources, experiences such as discrimination (e.g., Schildkraut, 2005), social networks (e.g., Jacobs & Tillie, 2004), national attachments, and ethnic identity (e.g., Rapp, 2020;Schildkraut, 2005) can affect immigrants' decision to become politically active.
Furthermore, immigrants may have retained strong political ties with their ancestral countries of origin and can be mobilised by the political parties in the country of origin (Kernalegenn & van Haute, 2020).Hence, immigrants' decision to join a political party in their country of residence can be influenced by a complex set of micro-level factors.
The rights of immigrants according to a country's laws and other relevant elements of what might be called the 'structure of political opportunities' for political engagement are particularly important for immigrants (Kastoryano & Schader, 2014).There has been a long tradition of research examining the role of variations in political institutions and opportunity structures in explaining cross-national differences in aggregate patterns of immigrants' political participation (Bird et al., 2011;Freeman, 2004;Vogel, 2008).However, the micro-foundations in the political decisions immigrants make have generally remained understudied in research on political opportunity structures and political participation.

Politics and Governance
In this research, we test the influence opportunity structures may have on micro-foundations in explaining party membership of CIO.Though data demonstrate that CIO in Europe are less likely to join political parties in the countries they live in than persons without a migratory background (see Supplementary File, Appendix A2), some still decide to join political parties, whereas others do not.Moreover, party membership among CIO is far more widespread in some European democracies than in others (Appendix A2).These findings underscore the need for explanations to understand the micro-foundations of responses to contextual variables at the macro-level.Recognising the impact political parties as organisations at the meso level may have in explaining party activism among immigrants (Dancygier et al., 2021;Höhne et al., 2023).In this article, we disentangle the mechanisms explaining the decision of immigrants to join a political party, focusing on variation streaming from institutional frameworks and individual differences among immigrants.First, we estimate the effect that the level of democracy in the country of origin and political efficacy have on party membership while controlling for recognised predictors of political participation of immigrants, such as resources, citizenship, duration of residence, and the person's attachment to the country of residence, as well as a predictor to party membership such as age and gender.Secondly, we embed individual-level analysis in a multi-level design estimating the impact of the institutional context in the country of residence on the individual-level factors explaining the decision to join a political party.The most fundamental institutional variables relevant to our research are laws, policies, and practices that shape the rights and opportunities of immigrants in their countries of residence, including policies that deal with conditions of settlement (integration policies) and policies that regulate access to full democratic membership (citizenship policies; see Helbling et al., 2017).That being said, by applying multilevel analysis, we go beyond standard models of party membership and political participation to explain immigrants' party membership.

Institutional Context and the Effects of Political Socialisation on Party Membership of Immigrants
The process of political socialisation for first-generation immigrants starts in their countries of origin and continues in their countries of residence (Neundorf & Smets, 2017).Through intergenerational transmission, first-generation immigrants transfer their experiences to their descendants (Aggeborn & Nyman, 2021;Jennings et al., 2009).Thus, for both first-and second-generation immigrants, the political socialisation-or the process by which citizens develop their political information, values, and behaviour (Neundorf & Smets, 2017)-is associated with structures and agents of socialisation in the countries of residence and of ancestral origin.
From the perspective of resistance theory, the political socialisation of immigrants is conditioned by political learning in their formative life (White et al., 2008).Following this theory, "people tend to avoid or reject environmental messages that are inconsistent with orientations accumulated during the formative years" (White et al., 2008, pp. 269-270).Applying this theory to the case of party membership, immigrants who experienced political parties in their country of origin in a way that is inconsistent with the role of political parties in pluralist liberal democracies may reject joining political parties.Citizens from countries with less developed democracies may lack the civic skills and knowledge required for political involvement in liberal democracies (Aleksynska, 2011;Just & Anderson, 2012).Studies demonstrated that immigrants' experience with autocracy reduces their support for democracy (Bilodeau et al., 2010), lowers immigrants' trust in political institutions (Voicu & Tufiş, 2017), and can hamper their social trust (Xu & Jin, 2018).While the implications of the democratisation in the country of origin on the electoral behaviour in the country of residence are less conclusive (Bilodeau & Dumouchel, 2023;Bueker, 2005;Okundaye et al., 2022;Xu & Jin, 2018), immigrants' experience with autocratic regimes may hinder other forms of political participation in the country of residence; (Bilodeau, 2008;Bilodeau & Dumouchel, 2023).That said, party activism of immigrants, as one of the costliest forms of political participation, might be particularly affected by immigrants' experience with an autocratic regime in their country of origin.
Advocates of the theory of exposure argue that the more exposure immigrants have to new social influences, the more they will adapt their thinking and behaviour to these newly learned patterns (White et al., 2008).Some scholars used the duration of residence as a proxy for socialisation and exposure (e.g., Sumino, 2023).Following this argument, research has found duration of residence to be positively associated with political participation and voting (Leighley & Vedlitz, 1999;Ramakrishnan & Espenshade, 2001) and can contribute to stronger partisan identities of immigrants (Cain et al., 1991;Wong, 2000).
Nevertheless, cross-sectional research, which demonstrates differences across countries in immigrants' political behaviour (e.g., Bird et al., 2011) indicates that the duration of residence alone does not explain exposure to political learning sufficiently.
The effects of exposure and resistance on the willingness of CIO to join a political party in the countries of residence can be affected by factors in the wider institutional environment of these countries (Freeman, 2004;Wright & Bloemraad, 2012).The political rights immigrants enjoy in their countries of residence may facilitate political integration (Martiniello, 2006).For example, Pilati and Herman (2020) find that more inclusive citizenship and residence regimes tend to increase immigrants' political participation in their countries of residence.Despite the inconsistent results on the implication integration and citizenship policies have on integration outcomes (Ersanilli & Koopmans, 2010, 2011;Neureiter, 2019), the policies for integration and citizenship define the rights and opportunities of immigrants in the country of residence.
For example, these policies define the conditions required for immigrants to enter the educational system and the labour market, the opportunities for political participation, opportunities for cultural and social integration, laws for protection against discrimination, and requirements for membership.Thus, more open policies provide more opportunities for immigrants to engage in diverse dimensions of society and enable immigrants to have stronger exposure and socialisation with agencies from their country of residence.From here, we hypothesise: H1: The higher the level of democracy in the country of origin, the more likely a CIO is to join a political party in the country of residence.H2: The more inclusive the institutional framework for integration in the country of residence, the less pronounced the effect of the level of democracy on the likelihood of party membership.

Institutional Context and the Effects of Efficacy on the Party Membership of Immigrants
Scholarship has provided robust evidence on the predictive value both internal and external efficacy have on the decision to take political action, including the decision to join a political party (Chang, 2023;Craig et al., 1990).Internal efficacy refers to "beliefs about one's competence to understand and to participate effectively in politics," while external political efficacy is capturing "beliefs about the responsiveness of governmental authorities and institutions to citizen demands" (Craig et al., 1990, pp. 290).In addition to the direct effect internal efficacy has on the intention to participate in conventional action, it also mediates the predictive power political knowledge and personality traits have on citizens' decision to take political action (Reichert, 2016;Vecchione & Caprara, 2009).Furthermore, external efficacy is related to political trust (Bienstman et al., 2023;Craig et al., 1990) and perceptions of institutions' responsiveness (Abramson & Aldrich, 1982).CIO have often been described as less informed about the politics of the country of residence, with lower levels of civic skills or language skills and lower levels of political and social trust (Quintelier, 2009;Xu & Jin, 2018).
A lack of skills and knowledge relevant to political participation can reduce immigrants' self-confidence in their ability to understand and participate in the politics of political parties in their country of residence, while a lack of political trust can hurt their external efficacy.With that, party activism, which is already rare for native citizens, can be particularly discouraging for immigrants.
Political efficacy is developed under the influence of macro-level factors.For example, direct democracy has been shown to have a positive effect on efficacy (Chang, 2023) and the predictive power efficacy has on political participation can be affected by the distinctiveness of the political party system (Ikeda et al., 2008).
Moreover, individual perception of connectedness within the community increases both internal and external efficacy (Anderson, 2010).Hence, the institutional frameworks and social experience in the country of residence could shape immigrants' efficacious feelings.
Focusing on internal efficacy, research has demonstrated that it develops through political learning (Beaumont, 2011), civic education (Pasek et al., 2008), and gathering political information (Nie et al., 1969).Goodman and Wright (2015) reveal that immigrants are more confident in their political decisions in countries that implement civic integration policies with higher requirements of civic and language skills during the process of naturalisation.As Vecchione and Caprara (2009) suggest, political actions are embedded in broader social agencies that shape people's confidence in their abilities to contribute to the functioning of democracy.Building on that, policies that define access to diverse social agencies may have significant power over immigrants' sense of internal efficacy and, thus, their readiness to participate in democratic processes.
Studies on external efficacy have demonstrated that proportional electoral systems, by increasing external efficacy, have a positive effect on electoral turnout (Karp & Banducci, 2008).In general, external efficacy is directly affected by the political system and the performance of governments (Coleman & Davis, 1976).
Following this argument, immigrants' external efficacy will depend on how satisfied and represented they feel with the government.That being said, we argue that more open integration and citizenship policies represent immigrants' interests better than rigorous integration and citizenship policies.Thus, we hypothesise that: H3: The higher the level of (internal and external) efficacy of CIO, the higher their probability of joining a political party in the country of residence.H4: Where integration policies are more inclusive, the effect of efficacy (internal and external) on party membership is weaker than in countries where integration policies are more exclusive.

Data and Methods
To identify the extent to which the party membership of immigrants is influenced by institutional factors and to understand whether, and if so, how CIO differ from citizens without immigrant origin when deciding whether they wish to work in a political party, we need a sample of the total population.Despite some limitations (Aleksynska, 2008), data from large-scale comparative population surveys have been found to show general promise in the study of party members (Ponce & Scarrow, 2016).We base our empirical analysis on the second version of the ninth round of the ESS.Our sample is limited to member states of the European Union that took part in the ESS in 2018 (in 2018, the UK was still an EU member state).Also in the sample, we included Switzerland and Norway as countries with similar opportunity structures as EU states and a significant population of immigrant origin.
We operationalise a respondent's immigration status as a dichotomous variable registering a value of zero if a respondent did not indicate any migratory background themselves or in their parent's generation (reference category) and a value of 1 for persons who were born outside the country of residence ("first-generation CIO") and respondents who said they had at least one parent who was born abroad as a foreign national ("second-generation CIO").We opted to amalgamate first-generation and second-generation immigrants in our estimations, largely because of the small number of first-generation immigrants amongst party members in some countries.This strategy makes better use of the heterogeneity among CIOs in terms of relevant personal characteristics and provides robust statistical results.
For our dependent variable, worked in political party, we use an item in the ESS asking respondents whether they had worked in a political party or similar organisation in the past twelve months before the interview.
For the independent variable on the macro-level, level of democracy in the country of origin, we rely on the V-Dem project.From this data set, we use the liberal democracy index ranging from low (0) to high (1) to characterise democratisation in each country of origin the respondents in the sample reported.For the independent variable at the individual level-political efficacy-we constructed scales for internal and external efficacy.These scales were produced through a factor analysis of five ESS items, which refer to the individuals' assessments of the responsiveness of the political system and the government and their own expected ability and confidence to participate politically (see Supplementary File, Appendix A5).For the macro-level variable integration policies, we employed factor analysis to produce scales from the MIPEX designed to capture legislation on education, anti-discrimination, family reunion, political participation, permanent residence, and access to nationality.Thereby we generated an index on integration, which is strongly associated with the MIPEX variables education, political participation, anti-discrimination, and access to citizenship (Appendix A5).We used the factor scores for integration to estimate cross-level interactions in the model.More detailed information on our individual and contextual covariates can be found in the Supplementary File, Appendix A3.
Acknowledging the comprehensive range of influential factors affecting political engagement among CIO (Just & Anderson, 2012;Ramakrishnan & Espenshade, 2001;Rapp, 2020;Verba et al., 1993), our model controls for individual covariates.These include citizenship, resources (education and occupation), duration of residence, and attachment to the country of residence.The variable occupational status is measured with the ESS item "occupation," using the 2012 International Standard Classification of Occupations (ISCO08).This covariate was recoded to create an ordinal variable with four ISCO-based categories.The variable duration of residence, capturing political socialisation in the country of residence, was created by calculating the difference between the year a respondent first came to live in the country of residence and the year the interview was conducted.The variable national attachment is measured in the ESS with a ten-point scale ranging from not at all emotionally attached (0) to very emotionally attached (10).Additionally, following the findings on the relevance gender and age have on the likelihood of party membership (Achury et al., 2020;Angenendt, 2023), our model controls for gender and age.
Hierarchical models are a useful estimation method because they adjust parameter estimates in relation to the clustered nature of the data (Gelman & Hill, 2006;Snijders & Bosker, 2011).Because our dependent variables are dichotomous, we fit multi-level logistic regression models.To assess whether the application of multi-level logistic regressions is justified, we first estimate the proportion of the second-level variance compared to the total variance by calculating the intra-class correlation coefficient (ICC).The ICC always ranges between 0 and 1, where ICC = 1 indicates that the entire variance is due to differences between the groups.In contrast, ICC = 0 would indicate that the total variance is due solely to differences within the groups.If the latter were true, there would be no variation between upper-level contexts and, therefore, no need for a multi-level analysis.Although there is no predetermined reference value, a rule of thumb frequently found in the literature is that a multi-level analysis is statistically necessary from a value of ICC = 0.05 (Hox et al., 2018;Snijders & Bosker, 2011).With an ICC of 0.089, the null hypothesis stating that there is no variance between contexts can be rejected.Thus, the application of multi-level analyses is statistically recommended.Our multi-level logistic regressions account for the random effects of the dependent variables concerning contextual differences.

Results
In this section, we focus on the results of our multi-level logistic regression model of party membership.
A descriptive overview of the used variables and more information on the descriptive statistics can be found in the Supplementary File, Appendixes A3 and A4.
Since the mean values of our factor scores (integration policy, internal and external efficacy) are zero and the mean value of the index of democracy in the country of origin is close to zero, coefficients in Table 1 can have a meaningful interpretation.The results in Table 1 demonstrate, for an integration factor of zero, that the higher the index of democracy in the country of origin the higher the chances that CIO will join a political party in the country of residence.Hence, this finding supports our first hypothesis.This result follows the line of research that argues that immigrant citizens with origin from an autocratic country face more challenges for political integration in the countries of residence than immigrants coming from established democracies (Bilodeau & Dumouchel, 2023;Bilodeau et al., 2010).Moreover, this result demonstrates that in addition to electoral participation and protests (Bilodeau, 2008;Bueker, 2005), the effect streaming from the political system in the country of origin can be extended to party membership in the country of residence.That said, this result provides empirical support for the "transferability argument" (Bilodeau et al., 2010;Voicu & Tufiş, 2017;Xu & Jin, 2018) and demonstrates that both political learning in the country of origin (first-generation) and political learning in the family (second-generation) can influence immigrants' party activism in the country of residence.
Furthermore, results in Table 1 show that, if the moderator variables (internal, external efficacy, and index of democracy in the country of origin) are zero, the inclusiveness of integration policies has a statistically significant positive effect on party membership for CIO, meaning that CIO are more likely to join political parties in countries with more open integration policies than in countries with more closed integration policies.
For example, Sweden and Portugal, countries that report the highest scores in openness of integration policies, have among the highest rates of party members with immigrant origin.On the contrary, Bulgaria and Hungary, countries that rank in the lowest ranks for integration policies, report the lowest rates for party members with immigrant origin (see Supplementary File, Appendixes A2 and A5).This result contributes to the literature on immigrant political participation and demonstrates that integration policies matter for the political integration of immigrants.For a more vivid interpretation of the effect the institutional frameworks have on the outcomes of party membership across immigrants, we visualise our statistically significant multilevel analysis testing the moderation effect the integration policies in the country of residence have on the predictive value the "level of democracy in the country of origin" has on the party membership of immigrants.Results in Table 1 demonstrate that integration policies moderate the effect of "democratisation in the country of origin."This result supports the argument for political socialisation as lifelong learning (Neundorf & Smets, 2017), as well as theories on exposure underlying the importance of political learning and political resocialisation in the receiving country (White et al., 2008).
The plot of predictive margins in Figure 1 shows that supportive integration policies play a significant role in reducing the gap in party membership found for respondents from countries with relatively low values on the V-Dem index of liberal democracy.This result supports our second hypothesis.Moreover, this finding adds to the existing knowledge on political resocialisation and demonstrates, in addition to time spent in the country (Sumino, 2023), that policies governments implement influence the "exposure" immigrants experience in the country of residence and can shape the outcomes in the political participation of immigrants.Further, in our model, we test the predictive value internal and external efficacy have on party membership.
Results, again under the respective condition that the integration factor is zero, demonstrate that internal efficacy has a strong predictive power.The effect of internal efficacy holds for CIO and the autochthonous population and confirms earlier findings in the literature on political participation (Chang, 2023;Craig et al., 1990).External efficacy, by contrast, is significantly associated with party membership for the autochthonous population, but not for CIO.These results demonstrate that although for autochthonous populations both internal and external efficacy are equally relevant in the decision to work in a political party, for immigrants, only their self-confidence in political actions matters, while their perception of institutions' responsiveness does not play a significant role.Hence, this finding partially supports our third hypothesis.Similarly to our findings, Michelson (2000) identified that external efficacy does not have the same power to vote for Latinos as it does for Anglos.Nevertheless, this finding leaves an open question of why external efficacy is not a relevant predictor for party membership of CIO as it is for native citizens.
Finally, our model tests whether integration policies will moderate the significant effect efficacy has over party membership.Our results demonstrate that the inclusiveness of integration policy does not moderate the effect of internal efficacy.Hence, our fourth hypothesis is not supported.This result follows findings identifying that immigrants' internal efficacy is not associated with the enforcement of immigration policies (Rocha et al., 2015) and contradicts studies arguing for significant effects of civic education and political learning on immigrants' internal efficacy (Beaumont, 2011;Pasek et al., 2008).Thus, though specific immigration policies may influence the internal efficacy of immigrants, the integration policies (operationalised as an overall score including various dimensions and segments of the integration process) do not have a significant impact on the internal efficacy of immigrants.
Reporting for our control variables (see Supplementary File, Appendix A1), results demonstrate that occupational status is a statistically insignificant factor for party membership for both CIO and autochthon citizens.Education, on the contrary, is a statistically significant factor for party membership of autochthon citizens while it is not for CIO.In line with existing research (Martiniello, 2006), naturalised immigrants are more likely to join political parties than non-naturalised immigrants.Age is a statistically significant predictor for both CIO and autochthon citizens, while gender is statistically significant only for autochthon citizens.
Finally, duration of residence and attachment to the country of residence are not significant predictors for party membership of immigrants.

Conclusion
This article provides a unique knowledge of factors explaining the decision of immigrants to join a political party in their country of residence, which is the initial step towards the incorporation of immigrants in political parties.
Presenting comparable models on the factors associated with the choice of party membership in the autochthonous population and immigrant population allowed us to see how factors shaping the decision to join a political party differ between the two groups.In our model, we find a significant positive effect of the level of democracy in the country of origin on the likelihood of immigrants joining a political party in the country of residence.This finding supports the literature arguing that political socialisation in the country of origin (Bilodeau & Dumouchel, 2023;Bilodeau et al., 2010) and political learning in the family (Jennings et al., 2009) matter for the political behaviour of immigrants.In addition to the effect integration policies have on the party membership of immigrants, we find that in multilevel interaction the inclusiveness of the integration policies reduces the negative effect of an autocratic country of origin on the party membership of immigrants.With that, this research supports the theories arguing that institutional frameworks influence political learning (Soss, 1999) and opposes the "resistance" theory of political learning (White et al., 2008).
More importantly, this finding contributes with a new perspective on how macro factors can shape the exposure to political learning immigrants experience in their country of residence and influence the outcomes of immigrants' party activism.
Furthermore, this research shows that internal efficacy is a powerful psychological driver of party membership irrespective of the migration status.The effect of external efficacy is positive for both citizens with and without immigrant origin, but it is statistically significant for the latter only.Though this finding supports our assumption that factors shaping the likelihood of party membership might not be equal for immigrants and natives, it leaves us with the question of why external efficacy does not matter for party membership of immigrants as it does for natives.The insignificant moderation effect between integration policies and internal efficacy indicates that specific programmes for political learning (Beaumont, 2011;Pasek et al., 2008) and electoral contexts (Chang, 2023;Ikeda et al., 2008) might have a stronger impact on internal efficacy than the broader institutional framework for integration.This article makes a unique contribution in explaining the party membership of immigrants and uncovers the impact the institutional frameworks, both in the country of origin and country of residence, can have on the party activism of immigrants in the country of residence.Our findings imply that institutional frameworksin our case, inclusive integration policies in the country of residence-matter as moderators reducing some disadvantages of party activism for immigrants.Bearing in mind the high number of international migrants in Europe originating from underdeveloped or developing democracies, our findings increase the importance of the implemented policies for integration countries.
Though this research provides a good base for immigrant party membership, further research can extend the knowledge of how additional factors can shape the party activism of immigrants.Following the rising number of war refugees in Europe, as well as recent findings on the importance experience with political violence and political conflicts in the country of origin can have on the political orientation of immigrants (Okundaye et al., 2022;Soehl et al., 2023), further research will benefit from a more extended contextual framework in the country of origin and the implication it may have on the political behaviour of immigrants in their country of residence.Though the cross-sectional analysis has enabled us to provide generalisable knowledge on migrants' patterns of party membership in different countries in Europe, the low number of cases did not allow us to test further interactions with specific immigrant groups or ethnic groups within the countries.

Table 1 .
Multi-level logistic regression model of party membership.