Keywords

The vast extent of migration following the accession of countries of Central and Eastern Europe (CEE) to the European Union after 2004 has potentially significant political implications. On the one hand, with electoral turnout in countries of the region already significantly lower than elsewhere (International Institute for Democracy and Electoral Assistance (International IDEA), 2022), the outcomes of domestic elections may be distorted by the abstention—whether enforced by law or by inconvenience—of a large cohort of voters some of whom might otherwise have cast a vote. On the other hand, those who do cast a vote overseas may cast that vote differently, with exposure to alternative sets of values and political contexts in their country of residence influencing their political choices back home. Their choices may also be influenced by the different contexts within which their votes are collected, counted, and translated into seats.

While some migrating EU citizens may choose to naturalize in their country of residence, EU freedom of movement means that there is little incentive for them to do so. With votes in elections to the European Parliament cast in the country of residence, external voting in countries of origin is in many cases the only means these migrants have to express political preferences on a national level. Yet migrant voting behavior remains understudied. Despite much recent interest in democratic backsliding, populism, and illiberalism and the consequences of these phenomena for party competition in the region and beyond (see Vachudova, 2021, for a comprehensive overview, and Buzogány, 2017, and Pirro & Stanley, 2022, for in-depth studies of the key cases of Hungary and Poland), there has been a little empirical study of the political impact of outward migration (although see Dancygier et al., 2022) and transnational political remittances (although see Erdal et al., 2022).

This partly stems from a lack of data. While qualitative studies of the motivations and preferences of voters are feasible, large-n surveys of the kind that would yield data that can be directly compared with national-level election studies are logistically challenging and expensive to conduct among diaspora communities, particularly if probability samples are required.

However, the data we do have at our disposal—the results of voting in electoral commissions set up outside country-of-origin borders—remain underutilized. Most comparative research on external voting to date has focused on turnout (Burgess & Tyburski, 2020; Ciornei & Østergaard-Nielsen, 2020), with studies of the actual results of external voting focusing primarily on non-European democracies (Lafleur & Sánchez-Domínguez, 2015).

External voting results are neither a perfect nor a complete indicator of diaspora political preferences, and there are obvious limits on what can be achieved with aggregate-level data, but comparing the results of voting among diaspora communities with the choices made by their origin-country counterparts can reveal patterns that are substantively informative in their own right, and which generate further research questions and hypotheses that may be tested at the individual level.

In this chapter, we compare the post-enlargement external voting of CEE diasporas in Western Europe with voting patterns observed at the domestic level, focusing on turnout, different levels of support for parties in general, and different levels of support for parties with respect to key ideological dimensions and issues. To carry out these analyses, we use external voting data gathered by the authors on all parliamentary and—where applicable—presidential elections held in Bulgaria, Czechia, Latvia, Lithuania, Poland, and Romania from the election prior to EU accession in each country until April 2021. These six countries were selected on the basis of their high levels of outward migration and because external voting is potentially politically significant due to the number of votes cast or the symbolic importance of outward migration in domestic political discourse. External voting data for these countries’ elections was collected from 15 Western European countries: 12 EU member states (including the United Kingdom), plus Norway, Iceland, and Switzerland.

Following a brief description of the political contexts of the six origin countries, we describe the data and methods of analysis and then present the results of analyses regarding turnout, overall divergence, or convergence in voting behavior, and ideology-specific divergence or convergence.

The Political Contexts of the Countries of Origin

Diaspora and origin-country voters alike choose from the same menu. The six countries in this analysis have different types of party system and varying levels of party system “closure” (Casal Bértoa & Enyedi, 2021, pp. 34–35), but all offer a reasonably legible set of choices to their electorates. The five countries which hold direct presidential elections also offer broad slates of candidates. While the main purpose of this chapter is to analyze the differences between the choices of diaspora and origin-country voters, a brief contextualization of the circumstances in which those choices are made will help elucidate what follows.

Bulgaria

For the first decade of post-communist transition, Bulgarian politics was structured in accordance with a divide between an ex-communist left and anti-communist right. After 2001 this gave way to a period of political instability, with nationalist and populist movements and parties coming to the fore, and the role of ideological differences downplayed amid a focus on the issues of corruption and crime. Karasimeonov (2019, p. 13) characterizes Bulgarian parties as “predominantly parties of power” which exploit their access to power for clientelistic purposes. Between 2009 and 2021 the political scene was dominated by the center-right Citizens for European Development of Bulgaria (GERB), but in 2021 Bulgarian politics entered a period of significant ferment, with three parliamentary elections during that year amid the refusal of increasingly popular challenger parties to cooperate with parties of the establishment. Amid these changes, the office of president has remained relatively stable, with successive incumbents emerging from major political parties until 2016, when current incumbent Rumen Radev, an independent, was elected to his first term.

Outward migration is a phenomenon of increasing significance, with the proportion of Bulgarian citizens of working age who are residents in another EU country increasing from 6.0% of their home-country resident population in 2010 to 10.3% in 2020.

Czechia

Following the dissolution of Czechoslovakia in 1992, the Czech party system offered a broad spectrum of ideological options from extreme left to extreme right. The party system was initially characterized by divides over economic issues, with the right-wing Civic Democratic Party (ODS) and the Czech Social Democratic Party (ČSSD) as the main points of orientation. While the party system remained open to the emergence of new formations, this dynamic remained dominant until 2010, after which the erosion of bipolarity between 2010 and 2013 (Balík & Hloušek, 2016) saw the emergence of a variety of populist forces from across the political spectrum, in particular billionaire Andrej Babiš’s Action of Dissatisfied Citizens (ANO). This reshaped the ideological complexion of the party system, with corruption in particular becoming an issue of key importance. Prior to 2013, the president was elected by parliament, with former dissident Václav Havel and then ODS leader and former prime minister Václav Klaus holding the post for two five-year terms apiece. Former ČSSD prime minister Miloš Zeman then won two successive elections in 2013 and 2018.

Czechia has the lowest rate of outward EU migration among the CEE countries that joined after 2004. In 2010, the proportion of Czech citizens resident in another EU country was only 1% of the home-country resident population, and there was barely any increase by 2020, when the figure stood at 1.1%.

Latvia

Latvian party politics has been marked by extreme multipartyism, which has led to a plethora of short-lived governments. Nevertheless, legible political cleavages can be discerned beneath the unstable surface. Auers (2015, pp. 110–111) identifies a socio-economic dimension relating to the nature and pace of post-1989 market reforms which has declined in relevance, while an ethnic cleavage rooted in post-Soviet policies over language and citizenship has persisted, and a political value divide has emerged between corruption-fighters and oligarchs.

In 2010, the proportion of Latvian citizens resident in other EU countries was 2.6% of the home-country resident population. This rose to 5.9% by 2020.

Lithuania

Lithuanian party politics has been marked by a predominant left-right divide that shifted from a “regime-oriented conflict” over the Soviet past in the 1990s to a classic socio-economic divide in more recent years (Duvold & Jurkynas, 2013, pp. 137–138). While populist parties made significant gains after 2000, these parties did not fundamentally disrupt the underlying stability of the political system, with almost all ruling coalitions led by representatives of one side or the other of the left-right divide. This pattern was disrupted in 2016, when the agrarian Lithuanian Farmers and Greens Union (LVŽS), which convincingly won the parliamentary elections in 2016, capitalized on voters’ disenchantment with established parties. Four years later, the conservative-liberal Homeland Union—Lithuanian Christian Democrats (TS–LKD)—returned to power, but the poor showing of the Social Democratic Party of Lithuania (LSDP) raised some uncertainty as to whether the left-right divide would persist as the most significant point of ideological orientation in Lithuanian politics. With the exception of Rolandas Paksas, impeached and removed from office in 2004, the office of president has been marked by relative stability, with two incumbents, Valdas Adamkus and Dalia Grybauskaitė, serving two five-year terms each.

In 2010, the proportion of Lithuanian citizens resident in other EU countries was 4.0% of the home-country resident population. By 2020, this had risen to 6.5%.

Poland

The first decade of post-transition party politics in Poland revolved around the post-communist divide between the Democratic Left Alliance (SLD), the successor party to the communists, and an ideologically eclectic variety of parties drawn largely from the ranks of the former opposition. After the watershed election of 2001, in which three parties contesting the legitimacy of Poland’s transition settlement entered parliament, the party system transformed into one defined by a divide between nativism, populism, and “solidaristic” economic policy on the one hand and a relatively socially progressive and pro-free market orientation on the other (Stanley, 2018, pp. 20–22). While the party system remained open to newcomers, particularly in ideological niches insufficiently catered for by existing parties, the main line of competition ran between the conservative-liberal Civic Platform (PO), the main party of government from 2007 to 2015, and the populist-nationalist Law and Justice party, which following its victory in 2015 set Poland on a course of democratic backsliding that has taken it from one of the front-runners of post-communist democratization to one of the progenitors of a new form of electoral autocracy. Since the victory of SLD candidate Aleksander Kwaśniewski in 1995, successive presidents have emerged from one of the major parties of the day, with the incumbent usually ideologically congruent with the current government.

While Poland has seen the largest wave of outward EU migration in absolute terms, a relatively small proportion of the working-age population has emigrated. In 2010 this stood at 2.7% of the home-country resident population, rising to 4.1% in 2020.

Romania

The violent end to communism in 1989 led to a gradual transition over the next decade that was dominated by the competition between forces associated with the old regime. The urban/rural divide, which contained a significant socio-economic component, was a key determinant of voting behavior as transition reforms generated significant cohorts of economic losers (Crowther & Suciu, 2013, p. 376). By the second decade of transition, the influence of former communists had waned to the extent that a new divide emerged that focused on issues of clientelism and corruption, the battle against which would play a defining role in Romanian politics over the subsequent two decades, informing the creation of formal coalitions and informal alliances. During this period, the presidency was also the locus of related disputes, with two-term president Traian Băsescu twice suspended and subjected to impeachment proceedings over alleged abuses of power.

In relative terms, Romania has seen by far the largest migrant exodus among the countries analyzed here. Only three years after accession, the proportion of EU mobile Romanian citizens was 11.5% of the number of home-country resident population, and this rose dramatically to 18.6% in 2020.

Data and Methods

The analyses in this chapter use two sources of data. The first is the aforementioned dataset on diaspora voting, which supplies information on electoral turnout and aggregate votes cast for political parties among both diaspora and origin-country communities. This is supplemented with data on the ideological placements of parties from the V-Party expert survey (Lindberg et al., 2022), as described below.

For the analysis of turnout, we simply calculate the proportion of voters participating at a given election as a percentage of all voters participating in that election. The remaining analyses require more detailed explanation. The basic unit of observation in our analyses is the percentage share of the vote cast for individual parties in each of the country of residence, per country of origin. For the analysis of how diaspora voting converges with or diverges from origin-country voting, we use the Pedersen index (Pedersen, 1979) to derive a measure of disparity between diaspora and origin-country electorates. This index was originally developed to measure electoral volatility in terms of the net change in party electorates between consecutive elections. We use it instead to compare the differences between votes cast for parties by origin-country voters and votes cast by diaspora voters. The index consists of values between 0, where there is no difference between origin-country and diaspora voters, and a theoretical limit of 100 (in practice, the variable rarely exceeds 60).

For the analysis of how diaspora voting converges with or diverges from origin-country voting in terms of ideology, we derive an ideological “center of gravity” for both diaspora and origin-country voters. This measure is calculated as a weighted mean of party positions on a given ideological dimension, where the weight in question is the party’s level of support. This measure thus combines information about the ideological positions of parties (which are the same in each origin-country/diaspora community dyad) with measures of support for parties (which differ in each origin-country/diaspora community dyad). This gives us a composite measure of the extent to which origin-country and diaspora electorates favor parties with particular profiles on each of the ideological dimensions.

The V-Party data provide a wide variety of ideological dimensions and specific issues. The 12 we selected for our analyses are intended to cover both broad and specific questions about the nature of contemporary democratic systems, socio-cultural values, and economic distribution. They are as follows:

  • PluralismFootnote 1 (the extent to which a party shows a commitment to democratic norms prior to elections)

  • Populism (the extent to which party representatives use populist rhetoric)

  • Economic left-right (the extent to which a party supports or rejects an active and redistributive role for government in the economy)

  • LGBT rights (the extent to which a party supports or opposes social equality for the LGBT community)

  • Gender equality (the extent to which a party has women in national-level leadership positions)

  • Equal participation of women (the extent to which a party supports or opposes equal participation of women in the labor market)

  • Reference to religious principles (the extent to which a party invokes God, religion, or sacred texts to justify its positions)

  • Immigration (the extent to which a party supports or opposes immigration into the country)

  • Cultural superiority (the extent to which a party leadership promotes the cultural superiority of a specific social group or the nation as a whole)

  • Welfare (the extent to which a party prefers means-tested welfare policies, which are based on the application of resource-based criteria for recipient eligibility, over universalistic welfare policies, where no such criteria are applied)

  • Clientelism (the extent to which a party provides targeted goods and benefits to keep and gain votes)

  • GAL/TANFootnote 2 (the extent to which a party’s ideological profile is green, alternative, and libertarian compared with traditionalist, authoritarian, and nationalist)

Those variables not already bounded by 0 and 1 were standardized to that range for ease of interpretation.

These aggregate disparity and ideological position variables were used as dependent variables in a series of Bayesian ordinary least squares regression models. There are three independent variables in these models. The first is a measure of time, operationalized as an index variable with a lower bound of zero, in which the unit is year, and the starting point is the election preceding a given country’s accession to the European Union. The squared term of this variable was also included to capture nonlinear change over time. The second independent variable is a set of dummy variables for countries hosting diaspora communities, and the third is a set of dummy variables for the country of origin. As the aim is to ascertain how trends over time vary across origin-country and diaspora contexts, the independent variables interacted with each other.

The use of models with dummy variables and interactions means that an interpretation of the results of our analysis cannot be undertaken on the basis of regression tables, as their coefficients relate to an arbitrary reference category. Furthermore, the length of these tables precludes their publication. They are available from the authors on request. Instead, we rely on marginal effects plots of our quantities of interest, which allow us to identify the extent to which general and ideological disparities vary over time, and whether there are statistically significant differences between diaspora electorates and their origin-country counterparts.

Results of the Analysis

Turnout Patterns and Change Over Time

Patterns of turnout at parliamentary and the first round of presidential elections are presented in Fig. 3.1, which plots turnout in each diaspora as a percentage of all votes cast in the given election. Each trend line starts at the election prior to EU accession and ends at the most recent election for which data were collected.

Fig. 3.1
A graph compares the parliamentary and presidential diaspora turnout between 2000 and 2020. Each plot has 6 increasing curves of Bulgaria, Czechia, Latvia, Lithuania, Poland, and Romania.

Diaspora turnout as percentage of overall turnout

Full size image

A general trend of rising turnout is evident, although not in all origin-country cases and not at the same rate. In some cases, such as parliamentary elections in Latvia and Poland, the trend rises relatively smoothly over time, while in others, for example, Bulgaria, it fluctuates. However, with the exception of parliamentary elections in Czechia, there is at least some discernible increase from the initial position.

The rise in general trends can be explained in part by the increasing size of CEE diaspora communities in the countries of Western Europe. Before accession to the EU, migrant communities had a much smaller presence, and in some cases only a negligible one. Following accession, CEE migrants to Western Europe constituted a significant proportion of the citizenry of their respective countries of origin. However, the fluctuation in these trends suggests that rising turnout is not necessarily attributable simply to rising numbers of migrants. If it were, we would expect a rise in diaspora turnout proportional to the increase in migration and the increasing enfranchisement of those who migrated as children or were born abroad.

The Romanian presidential elections of 2019, where the proportion of diaspora voters participating in the first round abruptly jumped from around 1% of all voters in 2014 to over 5%, is an exemplary case of how factors other than the growth of migrant communities may influence turnout. Following criticism of the hitherto inadequate infrastructure for administering elections abroad, the Romanian authorities introduced reforms that substantially facilitated participation by diaspora voters, including a three-day voting period and longer opening hours for polling stations (Comai, 2019). With diaspora communities also galvanized by political factors (the vast majority of votes abroad were cast for liberal incumbent Klaus Iohannis), turnout in the second round was even higher, at over 10% of all votes cast. On the other hand, while turnout in the 2020 parliamentary elections also increased, it did so by a smaller margin. One possible explanation for this is that while diaspora voters participate in presidential elections on the same basis as origin-country voters, translating diaspora votes into parliamentary seats is a more complex process that often renders the link between vote and outcome less legible to the voter.

On the other hand, it should be noted that across the five origin countries which hold direct presidential elections (Latvia’s president is elected by parliament), the diaspora turnout is not consistently higher than in parliamentary elections, suggesting that the more “direct” nature of the election is not a decisive factor. While turnout is generally rising, it is impossible to identify a single factor driving this. Rather, it appears to be a function of the number of migrants, the socio-demographic profile of the diaspora communities, the extent and character of the polling infrastructure, and the nature of the political contest.

Overall Disparity

Figure 3.2 shows the marginal effect of time on disparity, conditional on origin country and diaspora community. Each of the black points corresponds to a single origin-country/diaspora-community dyad. The trend line shows the median disparity, and the dark-to-light shaded areas correspond to credible intervals of 0.95, 0.80, and 0.50, respectively. Each of the subplots shows three essential pieces of information. The first is the extent to which diaspora communities as a whole diverge from origin-country voters, indicated by the level of disparity (as presented on the y-axis), and the credible intervals for the median measure. The second is the extent to which diaspora communities vary, as shown by the dispersal of the black points. The third is the extent to which disparity changes over time.

Fig. 3.2
A graph compares the parliamentary and presidential elections. It plots diaspora voters between Bulgaria, Czechia, Latvia, Lithuania, Poland, and Romania.

Disparities between country of origin and diasporas

Full size image

Parliamentary Elections

As shown in the topmost line of subplots in Fig. 3.2, there is clear variance in the extent to which diaspora communities differ from non-diaspora ones. The plotted points are often widely dispersed, indicating that there are significant differences between diaspora communities with respect to the extent of disparity. Yet some clear trends can be identified.

In the cases of Bulgaria, Latvia, and Poland, an initial post-accession convergence was followed by divergence. Disparity values among Bulgarian diaspora communities decreased on average between the last pre-accession election in 2005 to the election of 2014, meaning that the votes of diaspora Bulgarians began to resemble more clearly those of origin-country voters. However, by the first of the three parliamentary elections Bulgaria held in 2021, the level of disparity had clearly increased again. There is a similar pattern in the case of Latvia, where the average disparity fell between 2002 and 2010, only to rise again by 2018. In Poland, this trend was more pronounced: in 2001, the disparity level stood at 35, falling to just over 20 in 2011, only to rise again to around 30 by 2019.

Czech and Romanian diasporas show a pattern of increase over time. In the case of Czechia, disparity was on average around 30 in the last pre-accession election in 2002. With each successive election, disparity increased so that by 2017 average disparity was around 45. In Romania, this figure rose from around 25 in 2004 to over 40 in 2020, although there was considerably more variation than in the Czech case.

Lithuania is the only case in which there has been no increase over time, but the level of disparity is high. Between the last pre-accession election of 2000 and the 2016 election, average disparity figures remained stable at around 40. Only in the most recent election, held in 2020, did that average change, falling slightly to just under 40.

Presidential Elections

There is also considerable variation in disparity when it comes to presidential elections.Footnote 3 In the case of Bulgaria, average disparity fell from 20 in 2006 to around 16 in 2016. In Lithuania, Poland, and Romania, there was a pattern of initial decrease in disparity followed by an increase. In 2004, the disparity between diaspora voters and origin-country voters in Lithuania was over 30. This fell to around 26 in 2009 and 2014, only to rise again to over 30 in 2019. In Poland, there have been particularly large variations in levels of disparity between specific diaspora communities. On average, disparity fell from just under 30 in 2000 to 20 in 2010, mirroring the convergence observed with respect to parliamentary elections over the same period, only to rise again to 30 by 2020. Meanwhile, in Romania, disparity fell from over 40 in 2000 to less than 30 from 2009 onward.

There are only two points of comparison in the case of Czechia, which started electing its president in 2013. In both the 2013 and 2018 elections, there was a high level of disparity, at around 40. No countries have shown consistent election-on-election increases in disparity.

Ideological Disparity

Figures 3.3, 3.4, and 3.5 present measures of ideological divergence and convergence. These plots show the marginal effect of time on ideological position, conditional on origin country and diaspora community. In this case, the black points show the weighted mean for each diaspora community per origin country on the relevant issue or dimension, with the black line showing the median and the shaded areas credible intervals as detailed above. The red points are the corresponding weighted means for origin-country voters. The red line does not have any substantive meaning but connects the red points for clarity. If the red points fall outside the credible intervals, then it can be concluded that there is at least a 95% probability of ideological disparity between the aggregate vote choices of a country’s diaspora and its domestic voters on a given ideological issue or dimension.

Fig. 3.3
A graph compares pluralism, populism, economic left right, and G A L over T A N between 2000 and 2020. It plots diaspora voters and origin country voters between different countries.

Ideological disparities between diaspora and origin-country voters: pluralism, populism, economic left-right, and GAL/TAN

Full size image
Fig. 3.4
A graph compares L G B T rights, gender equality, equal participation of women, and reference to religious principles. It plots diaspora voters and origin country voters between different countries.

Ideological disparities between diaspora and origin-country voters: LGBT rights, gender equality, equal participation of women, and reference to religious principles

Full size image
Fig. 3.5
A graph compares immigration, cultural superiority, welfare, and clientelism. It plots diaspora voters and origin-country voters between different countries.

Ideological disparities between diaspora and origin-country voters: immigration, cultural superiority, welfare, and clientelism

Full size image

Pluralism

Overall, diaspora voters are more inclined than their origin-country counterparts to vote for parties that support the principle of pluralism. However, this varies by country. While in Latvia and Lithuania there is no statistically significant difference, in the remaining countries the ideological center of gravity for origin-country electorates is at least in some years significantly lower than the median of diaspora voters. The clearest difference can be seen in Czechia, where diaspora voters have been consistently more likely to vote for pro-pluralism parties, and in which the gap between the two sets of voters has increased in recent years. In the remaining countries, diaspora voters mirror origin-country voters, albeit imperfectly. In Bulgaria and Romania, voters at home and abroad have been more likely to opt for pluralist parties in recent years, but diaspora voters have been somewhat more likely to support these parties. In the case of Poland, support for pluralist parties has fallen, but to a greater extent among origin-country voters. While both sets of voters were equally likely to vote for pluralists in 2011, by 2019 diaspora voters were more likely to support pluralist parties, even if—reflecting changes in ideological supply—the overall level of support for these parties had fallen.

Populism

Unsurprisingly, given the considerable conceptual overlap with pluralism, the opposite relationship can be seen in the case of populism, where most diaspora communities are less likely to support populist parties than their origin-country equivalents. Czechia again stands out for the most significant different on this issue dimension; while support for populists has grown in recent years among all voters, origin-country voters are significantly more likely to vote for such parties. In Bulgaria and Romania, support for populist parties initially increased, but in recent years has waned. Both diaspora and origin-country voters follow this trend, and—with the exception of the 2021 election in Bulgaria—there is no statistically significant difference between diaspora and origin-country voters. In Latvia, support for populist parties initially waned but has increased in recent years, with no significant differences between voters at home and abroad. In Lithuania, where levels of support for populists remained low, diaspora voters were less likely to support such parties. In Poland, where support for populists remained relatively high, most diaspora communities were less likely to support such parties.

Economic Left-Right

Diaspora voters from Czechia stand out in particular on this dimension, being significantly more likely to support parties with a more right-wing economic profile. This is also the case to a somewhat lesser extent in Lithuania and Romania. While origin-country voters in Bulgaria were initially more likely than many diaspora communities to support more economically left-wing parties, there has been convergence in recent years, and in 2021 there was no difference between origin-country and diaspora voters on this question. In Latvia and Poland, meanwhile, diaspora voters were more likely than many origin-country voters to vote for right-wing parties, but the difference was increasingly negligible.

GAL/TAN

There are no significant differences on the GAL/TAN dimension in the cases of Bulgaria, Latvia, and Lithuania, with diaspora and origin-country voters alike voting for successively more TAN parties in Bulgaria and successively more GAL parties in Latvia and Lithuania. In Czechia and Romania, there has been an ideological divergence on this dimension; while initially there was no difference among voters, in recent elections diaspora voters have been more likely to vote for GAL parties. In the case of Poland, there was no statistically significant difference from 2001 to 2011, but following the turn of the Law and Justice government toward more conservative socio-cultural positions from 2015 onward, a gap emerged on this dimension, with origin-country voters significantly more likely to opt for more TAN parties.

LGBT Rights

Diaspora communities are generally more likely to vote for parties with a more positive approach to LGBT rights. In Bulgaria, where support for such parties has slightly declined in recent years, the difference is not substantial, and in the most recent election, there was no difference between diaspora and origin-country voters. In Latvia and Lithuania, support for pro-LGBT-rights parties has increased in recent years, but over the last decade, origin-country parties have been significantly less likely than diaspora voters to vote for such parties. The remaining countries have seen a divergence on this question. Until 2010, diaspora voters from Czechia were as likely as their origin-country counterparts to vote for parties with a more pro-LGBT-rights orientation. Since then, the two groups of voters have diverged, with diaspora voters now significantly more likely to vote for such parties. In Romania, such a separation occurred in 2012, since when diaspora voters, while supporting pro-LGBT-rights parties to differing extents, have been more likely on average in recent years to support these parties. In Poland there has been relatively little change in support for pro-LGBT-rights parties, but since 2011 diaspora and origin-country voters have increasingly diverged on this question. In the 2019 election, diaspora communities were substantially divided in their level of support for these parties, but in all cases, they were more likely than their origin-country counterparts to vote for them.

Gender Equality

In Bulgaria and Czechia, there has been little change over time in the average position of parties on this issue, and no consistent difference in the voting behavior of diaspora and origin-country voters. Elsewhere, parties have become, on average, more likely to have women in national-level leadership positions. For the most part, diaspora and origin-country voters do not differ on average in their support for more gender-balanced parties. This is in large part attributable to the fact that there is considerable diversity in the positions of diaspora communities, particularly in the cases of Latvia and Poland.

Equal Participation of Women

There is also little difference between diaspora and origin-country voters on the question of whether women should be able to participate equally in the workforce. This is perhaps due to the relatively low salience of this issue. With the exception of Bulgaria, where parties have become less likely to mention this issue over time, there has been little change in the extent to which parties mention this issue. At times, origin-country voters in Bulgaria and Czechia have been slightly more likely than diaspora voters to vote for parties who mention this issue, while the reverse has been the case in Poland and Romania, but for the most part there are no significant differences.

Reference to Religious Principles

There are clear differences between origin countries both in the extent to which religious principles are important aspects of political discourse and the extent to which those principles matter to origin-country and diaspora voters alike. In Czechia, the importance of religious principles to parties has declined over the period of analysis, and while in previous years diaspora voters were slightly more likely than origin-country voters to vote for parties espousing such principles, in the most recent election there was no difference. Romania has also seen a decline in the importance of this issue, but in this case diaspora communities—although divided on the question—are significantly less likely than origin-country voters to support such parties. In the case of Poland, religious principles have remained an important element of party-political discourse throughout the period of analysis. Prior to 2015, while diaspora communities were significantly divided in their support for parties prioritizing this issue, origin-country voters did not differ from the median diaspora community. Since then, diaspora voters have remained divided on the question, but significantly less likely than origin-country voters to support parties that make greater reference to religious principles. In Latvia and Lithuania, meanwhile, diaspora and origin-country voters do not vote differently on this question, and there has been no discernible change in the ideological profiles of parties over time, suggesting it is of negligible importance.

Immigration

In most cases, diaspora communities—migrants themselves—are more likely to vote for pro-immigration parties than their origin-country counterparts. The clearest exception to this is Latvia, where origin-country voters have been slightly more likely to vote for pro-immigration parties. In Bulgaria, the difference is minimal, but in the remaining countries there is a clear gap between the diaspora median and the positions of origin-country voters. In Lithuania parties have become on average more pro-immigration over time, and the choices of both diaspora voters and origin-country voters have tracked that increase. The same is true of Romania since 2012. However, in Czechia and Poland parties have become less positive about immigration in recent years. In Czechia, the choices of origin-country voters with respect to immigration were initially close to those of diaspora voters, but since 2014 a substantial gap has opened up, with the former much less likely than the latter to support pro-immigration parties. There has been a similar change in the case of Poland after 2015, although in this case there is significantly greater variation among diaspora communities, with some much more likely than others to vote for pro-immigration parties.

Cultural Superiority

There is also significant variation in the case of cultural superiority. In Latvia and Lithuania, there has been very little change over time in the average position of parties on this issue, with parties more likely to oppose narratives of cultural superiority than support them. There is no significant difference between the choices of diaspora and origin-country voters in both cases. While parties have become more likely to support narratives of cultural superiority in Bulgaria and somewhat less likely in Romania, the choices of diaspora voters do not diverge substantially from those of origin-country voters. However, in Czechia and Poland there are clearer differences. In Czechia, support for parties espousing narratives of cultural superiority has declined over the period among the diaspora, but origin-country voters were as likely to support such parties in 2018 as in 2002. In Poland, support for such narratives initially declined, only to increase again after 2015. While there is significant variation in the responses of diaspora communities to this question, origin-country voters are more likely to support parties which advance narratives of cultural superiority.

Welfare

In keeping with the more general preference for right-wing economic positions among the diaspora, those voting abroad are more likely to support parties whose welfare policies are based on means testing. Poland is a clear exception here; while the positions of parties have tended toward more universalistic welfare provision in recent years, the Polish diaspora is no less likely than its origin-country counterpart to support parties who take this position on welfare. Elsewhere, party positions have largely been moving in the direction of means-testing, with diaspora voters more likely to support parties which offer such a policy.

Clientelism

For the most part, there is very little difference between diaspora and origin-country voters with respect to their support for parties which practice clientelism. In Bulgaria the average party is more clientelistic than in Czechia, Latvia, Lithuania, and Poland, but all voters, regardless of their migrant status, are equally likely to vote for such parties. Romania stands out for significant divergence on this issue. While in the last pre-accession election in 2004 and the first post-accession election in 2008 diaspora communities did not differ from origin-country voters, since then a large gap has opened up. While there is substantial variation in the degree to which diaspora voters are willing to support clientelistic parties, in all cases they are significantly less likely than origin-country voters to give such parties their support.

This chapter has analyzed the aggregate voting behavior of diaspora voters, comparing them with voters in their country of origin. After examining levels of turnout since the last pre-accession elections in each country of origin, we then analyzed the disparities in aggregate party choice, looking both at overall levels of disparity and then at disparities of an ideological nature. The evidence marshaled in this chapter has shown that while post-EU-accession diaspora voters—facing as they do more onerous barriers to participation—are less likely to turn out in elections than their origin-country counterparts, their participation has not diminished with time, and those who do vote make choices that are legible with respect to origin-country political dynamics and relatively consistent over time.

Those choices can also be understood with recourse to many of the same established theories we apply to the study of voting behavior in more institutionally predictable and conventional origin-country settings. A recent study of Finnish migrant voters by Peltoniemi (2018) not only emphasizes the particular importance in diaspora settings of easy access to infrastructure for voting but also points to predictors such as interest in politics, age, and education. Mügge et al. (2021, p. 416) show that socio-cultural attitudes and socio-economic status play a significant role in the vote choices of Turkish migrants in the Netherlands, albeit often in ways which are inflected by the host-country context. Such studies suggest that diaspora voting is driven by many of the same classic drivers of voting behavior that inform choices back home, but that idiosyncratic features of the diaspora experience must also be taken into consideration.

Aggregate-level data of the kind used in this chapter can identify persistent patterns of similarity and difference between diaspora and origin-country voters, and help generate research questions and hypotheses about the differences we observe. However, such data can only take us so far. Individual-level observational and experimental studies are needed if we are to delve beneath the broad tendencies identified here to understand more about what informs the distinct character of diaspora voting.

Where turnout is concerned, we observe that interventions to increase the density and accessibility of polling infrastructure have resulted in increases in turnout. This—allied with the general tendency for turnout to increase rather than decrease over time—suggests that there is a significant “latent” electorate among diaspora communities; that is, there are many who are willing to vote, but are dissuaded from doing so by the lack of adequate infrastructure for their participation. We therefore forward the hypothesis that decreasing the investment of time and effort required to participate in elections will significantly increase the probability that a diaspora voter turns out to vote.

The persistent—and in many cases relatively high—disparities between voting behavior among diaspora voters and their origin-country counterparts point to significant and persistent structural differences between these two groups that are relatively resistant to change. One likely source of these differences is socio-demographic composition. While the nature of socio-demographic differences is likely to vary from diaspora to diaspora, and from origin country to origin country, the persistence of the disparities we observe despite often quite substantial changes in the ideological menu of parties on the supply side leads us to hypothesize that the most significant predictor of diaspora/origin-country voting disparities is socio-demographic rather than ideological in character.

That is not to say that ideology does not matter. Clearly, as our analyses have shown, there are some significant differences between origin-country and diaspora electorates in terms of the ideological nature of the parties they are willing to support. The clearest and most consistent differences concern overall economic orientation and attitudes to welfare. However, it is not clear whether this outcome reflects socio-demographic differences of the kind identified above, or distinct diaspora experiences. One hypothesis is that the greater propensity for diaspora communities to support more economically right-wing parties can be explained by the overrepresentation in those communities of socio-demographic groups more likely to cast such votes. Alternatively, it may be hypothesized that this difference reflects a lower sense of communal obligation to society as a whole, stemming from a migration-induced sense of remoteness from the origin-country community.

Finally, we also advocate the over-time exploration of the impact of host-country society and politics on the behavior of diaspora voters. While this investigation has focused on aggregate patterns across Western European host countries as a whole, each data point is also a potentially different context in time and space, with the origin-country voting behavior of diaspora voters potentially influenced not only by origin-country considerations but also by the impact of their experiences as members of distinct diaspora communities in diverse contexts.