Asymmetrical Voting Behavior. Janus-Voting Among Migrants

Abstract

There are cases in which the same electorate casts votes for ideologically detached candidates and parties in a short time-span. Migrants with dual voting rights often fall in this category. Examining the Turkish diaspora in Germany, I hypothesise that social identity and cross-generational interaction terms have influence on attitudes towards parties. I explore their effects on the probability of voting for a certain party through logistic regression using maximum like-lihood estimation. Introducing a new theoretical model addressing value shifts, this study also yields policy recommendations for social and integration issues.

1 Introduction

Elections and voting are integral components of democracies as they constitute mandate-awarding processes for political actors. Subsequently, social scientists allocate a great degree of their research to this field of inquiry. voters can express their preferences on policies and elect legislators whom implement those favoured by them. various factors guide vote choice, depending on the sender - the voter. political, economic and social judgments retain a high level of significance. According to rational choice theory, individuals act purposefully as for what they consider best for themselves (Scott, 2000). Interests are not only related to objective material benefits but can also encompass various notions, e.g. ideological beliefs etc. correspondingly, while it is plausible that vote choice can diverge in sufficiently distant elections, e.g. due to change of ideas or socioeconomic status¹, it is expected to be anchored more or less in the same ideological spectrum at one point of time.

However, there are cases in which the same electorate casts votes for ideologically detached can- didates/parties in a short time-span. Specifically, the case of migrants constitutes a puzzle: What are the main determinants of deviant migrant dual voting? Particularly in times of the revitalisation of the nation-state and continuous migration waves, this question becomes urgent to examine (Si­mon and Ruhs, 2008). Furthermore, the political and social emancipation of migrant communities, together with an increasing trend of autocratisation and polarisation in the origin countries, raises the necessity to scrutinise migrants and their bidimensional voting behaviour. vote choice provides a useful understanding of their identity and values. Considering that a certain part of migrants often have the right to vote in both the country of immigration and emigration, it offers an opportunity to explore the effects of social identity and socioeconomic variables on vote choice, especially if the response variable is tolerably different in both countries.

I introduce a novel theoretical approach to this research that bridges the gap in the literature between if and how migrants participate and addresses why they behave in this manner. After reviewing extant literature, I present a new framework interlinked with the value shift model. I then draft a research design and discuss my expectations and future research opportunities. This research thus, also yields insights for future policies vis-a-vis immigration and social integration.

2 Literature

Literature about the nexus between internal and external voting behaviour is very limited (collyer and Vathi, 2007; Chaudhary, 2018; Mügge et al., 2019; Grad, Gherghina and Ivan, 2020). The main focus on migrant participation lies on the questions whether they participate and if so, under which conditions (Scuzzarello, 2015; Boccagni, Lafleur and Levitt, 2016)? Additionally, inquiries about voting behaviour in both the origin and host country² and the interaction between them re­main insufficient while external voting rights have been extended in over 115 countries (Ellis et al., 2007; Mügge et al., 2019)³. Bauböck (2003) emphasises the duality of migrants as they are both, immigrants and emigrants and regards dual citizenship as the formal acknowledgement of transna­tional belonging. This is interesting as citizenship was always imagined as the connection between individuals and their physical residence. However, with increasing globalisation and migrant waves in the last 50 to 60 years, states as well as citzens attempt to continue this relationship even if it is not congruent with the traditional view of nation and nationhood (Mügge, 2012). Extra-territorial citizenship, as it is formally called, is viewed as an effort to build a purposeful bond to the diaspora abroad and also to be able to mobilise this group of people for long-term advantages (de Haas, 2006; collyer, 2014; Grad, Gherghina and Ivan, 2020). However, allowing migrants to participate in the electoral process in their origin country is a very controversial and highly tense topic.

Dual citizenship and voting “at home” is considered to impede the political, social, and cul­tural integration in the host country as it establishes cross-border affiliations (Boccagni, Lafleur and Levitt, 2016; Vermeulen, 2019; Mügge et al., 2019). Yet, the direction of this effect is not clear as policies of exclusion in the host country could lead to the strong attachment to the origin country, which would result in the claim of failed integration (Mügge, 2012; Scuzzarello, 2015). Criticism on immigration and integration policies presents itself especially in Europe. Too much tolerance would isolate migrant groups to a distinct sphere, a comfort zone, in which their old and incom­patible values remain not only the same but also are transmitted to the next generation (Kundnani, 2012; Vermeulen, 2019). Demands - explicit or implicit - for the conclusive integration into the ma­joritarian liberal society is a common denominator in the politics concerning immigrants (Joppke, 2004; Scuzzarello, 2015).

Migrants are situated in an environment consisting of two political and social communities, transcending borders and places of residence (Finn, 2020). This phenomenon prompts questions of democratic theory: Should participation be available only for those who are affected by the vote (Collyer and Vathi, 2007; Bauböck, 2005)? Collyer and Vathi point out the two rival concep­tions, following Weber⁴: the ethics of (republican) responsibility, which posits that only physically present citizens who have to bear the consequences of the vote may partake in the electoral pro- cess⁵, and the notion of the ethics of conviction, that would maintain the position that migrants are nevertheless citizens and thus, entitled to vote in the origin country. Finn (2020) takes the broader approach of examining migrant participation as a general phenomenon and differentiates between immigrant, emigrant, and dual transnational voting. political opportunity structures and the iden­tification with the host country are both important for analysing migrant participation. In contrast to the notion that dual citizenship is an obstacle for integration, other scholars underline that dual identity is possible and that the identification with the host country alongside the sense of belong­ing to a minority group has positive implications for political participation and hence, integration of migrants (Klandermans, Van der Toorn and Van Stekelenburg, 2008; Hopkins, 2011; Scuzzarello, 2015).

Chaudhary (2018, 438) differentiates two manifestations of migrant participation: (1) reso­cialization, i.e. the course of adjustment to the host country, “which ultimately weakens transna­tional political practices over time and in later generations”, and (2) complementarity, that expresses “an immigrant’s proclivity to engage in transnational politics with their countries of origin.” This study only provides unclear results: with more time spent in the host country, the odds of voting in the origin country decrease. However, voting in the most recent election in the host country increases the probability of also voting in the origin country.

Turning to the puzzle of deviant migrant voting, social scientists have contended that support­ive and inclusive structures and institutions can result in the empowerment of migrants’ democratic participation in the host country (Mügge, 2012; Scuzzarello, 2015). Yet, a discrepancy is dis­cernible: research about deviant migrant voting behaviour demonstrate that ideologically detached voting in origin and host countries occurs (Lafleur and Sanchez-Dominguez, 2015; Ahmadov and Sasse, 2016). Mügge et al. (2019) e.g. argue that Turkish immigrants tend to vote left in the major­ity of Western European countries while the Islamist and lately, right-wing government in Turkey also enjoys their support. This is supported by data collected by SVR (2017) in Germany: The majority of Turkish immigrants tend to back primarily Social Democrats and also Greens. Simul­taneously, Sevi et al. (2020) show that their vote overwhelmingly benefits the Turkish incumbent party, AKP.

3 Theory

How can this phenomenon be explained? Looking at the new V-Party data provided by Lührmann et al. (2020) in Figure 1, huge discrepancies between both vote choices become observable for the case of Turks in Germany⁶. This study does not analyse if and how migrants become involved but instead why they are participating in both countries the way they are, i.e. Finn’s (2020) dual transnational outline.

Migrants with dual citizenship have the right to vote in both countries. The Federal Statistical Office of Germany formally defines a person as migrant if she or at least one of her parents does not possess citizenship by birth (see also Statistisches Bundesamt (2020)). Yet, it is common to go back to the grandparents in order to look for a migratory background.

The support for a party is among other things determined by economic and political variables. Pocketbook and sociotropic voting are widely researched by economic voting scholars (Nannestad and Paldam, 1994; Dassonneville and Hooghe, 2017; Tilley, Neundorf and Hobolt, 2018). Mi­grants are not part of the majority group in society. Their migration story do, however, vary: For instance, Turkish emigrants in the United States are on average more educated and wealthier than the standard US-American while in Germany, they originate from lower class conservative workers. While the former constantly supports opposition candidates at home,the latter is inclined towards the incumbent party (Koinova and Tsourapas, 2018; Sevi et al., 2020). This particular category con­stitutes a socioeconomically weak minority and thus, strives for protection due to own endurance concerns. In order to ensure survival in the host country, they vote for minority-friendly parties, which mostly are leftist and diametrically opposed towards their conservative values (Mügge et al., 2019).

Abbildung in dieser Leseprobe nicht enthalten

Figure 1: Differences between the German centre-left parties (Social Democrats and Greens) and the conservative Turkish government parties (AKP+MHP alliance) in 2017 and 2018, respectively . Source: Lührmann et al. (2020).

Social identity theory (Turner, Brown and Tajfel, 1979; Tajfel, 1981, 1982) might explain this phenomenon: Turks in Germany are recognised as well as perceive themselves as a minority in­group with own dynamics and values and as an outgroup in relation to general society (Simon and Ruhs, 2008). In the process, they base their voting decision on experience about personal gains, predominantly material but also in regards to more chances. Subsequently, I posit my first hypoth­esis:

Hypothesis H1a: Migrants vote egocentrically and retrospectively IN THE COUNTRY OF IMMIGRATION.

Whilst migrants’ citizenship is usually examined from the view of the host country, they are also emigrants. Through the appearance of deterritorialisation, citizenship has become a “way of re-imagining the [...] state” (Collyer, 2014, 60). If the migrants belong or at least recognise themselves to belong to the majoritarian (in-)group in their country of origin, the voting behaviour undergoes a change as they switch identities. At the same time social remittances are also of great importance (Levitt, 1998; Boccagni, Lafleur and Levitt, 2016): They provide a connection with family and friends in the origin country, especially political cues and opinions. Moreover, the consumption of foreign origin media possesses direct information and bond-building purposes (Newland, 2010). However, because they do not have direct material gains through policies there, they tend to consider the whole - “what is good for the nation” - and not particular political stances. Hence,

Hypothesis H1b: Migrants vote sociotropically and prospectively in THE COUNTRY OF EMIGRATION. it is important to embed this phenomenon into a historical context - a path dependence. socioe­conomic attributes of the individual’s social environment have also an important effect on the issue determinant due to the aspect of socialisation. Factors like education, income etc. by all means have a positive effect on integration (see also Heckmann (2008)). Still, they should be put into a socialisation perspective: Even if attitudes like religion, education, and wealth have changed for the descendants of first immigrants, I hypothesise that cross-generation interaction terms regarding these variables influence the vote. While a secular and less religious person would be expected to vote more liberal, the attributes of her parents could influence the weight or direction.

[...]

¹ These are variables that principally experience long-term change instead of a sudden shift, see also campbell et al. (1960)

² “Country of settlement” could also be used in this context. However, as I intend to analyse dual citizen behaviour and address the puzzle of ideological divergent voting, I regard “host country” - for now - as more appropriate.

³ Since the end of the study of Ellis et al. (2007), more countries have granted their citizens the right to vote from abroad, which makes the argument even more compelling.

⁴ See also Weber (1919).

⁵ This mostly resembles Dahl’s (1971) stance as he considers taxation as prerequisite for participation and represen­tation.

⁶ The dataset provides information for the alliance per se. However, this is only an electoral alliance, the AKP constitutes the government (Seljuk and Hekimci, 2020; Esen and Yardimci-Geyik^i, 2020)