Migrants' Voter Turnout in the Home Country Elections: Non‐Integration or Political Anchor?

The transnational political participation of migrants has been extensively analyzed in the literature. Previous explanations focus on individual determinants ranging from political interest or efficacy to social ties or socio‐demographic characteristics. So far, little attention has been paid to the contrast between factors related to their lives in two different countries. The present article adds to this burgeoning literature by identifying and comparing the effects of several attitudes and behaviors of migrants in the host and home country on their voter turnout in home country elections. We use individual‐level data from a survey conducted in 2022 on 1,058 Romanian migrants living around the world. The results indicate that migrants who remain anchored in the politics of their home country—without necessarily striving to return—and those who are engaged in their host communities are more likely to vote. Migrant voter turnout is not determined by poor integration in the host society.


Introduction
Many countries around the world allow their citizens living abroad to vote in national elections.The procedures used to facilitate voting vary and include polls organized in the host country, postal ballots, or a requirement to return to the home country for election day.The basic principle behind the expansion of voting and representation rights beyond borders remains the same: to allow non-resident citizens to have a voice in the decision-making process of their home country, especially in the context of large contemporary flows of migrants (International Organization for Migration, 2020).The voting patterns of migrants have produced sizeable effects on home election results across several countries (Gamlen, 2015;Gherghina, 2015) and there is an increasing discrepancy between how the diaspora and the electorate in their home countries vote over time (Szulecki et al., 2023).Under these circumstances, the turnout of migrants in home countries' national elections has attracted extensive scholarly interest over the past three decades.
There is a consensus in the literature that migrants' political participation is shaped by the interaction between what they experience both in their home and host societies, reinforced by their political, economic, and cultural links (Chaudhary, 2018;Ciornei & Østergaard-Nielsen, 2020;Tsuda, 2012).However, we do not know which of these two types of determinants has a stronger influence.The present article seeks to address this gap in the literature by analyzing the effects of migrants' individual-level characteristics in the host and home country on their voter turnout in home country elections.It focuses on Romanian migrants abroad as the most likely case where voter turnout may occur.This case has four characteristics: (a) there are special seats for the diaspora in the Romanian Parliament; (b) there is relatively high access to voting through postal ballots (since 2016) and through the creation of polling stations in their host countries; (c) Romanian migrants have voted extensively in the past, especially in the presidential elections, and Romanian parliamentarians often directly address the diaspora's priority issues in their speeches; and (d) there is a history of transnational digital activism supporting protest causes in Romania.
We use individual data from a survey conducted in 2022 among 1,058 migrants living in 31 countries, with most respondents living in six European and North American countries that tend to be preferred destinations for Romanian migrants.Our analysis covers, in chronological order, the voting of Romanian migrants in home elections for the European elections (May 2019), the presidential elections (November 2019), and the national legislative elections (December 2020).The choice of these elections was for methodological reasons: Each election is different, thus we can gauge the general participatory behavior of migrants rather than their turnout in a specific type of election and as they were all organized relatively close to the timing of our survey, which increases the probability that respondents will accurately report their turnout.
This article brings three contributions to the existing literature.The first is that it moves beyond the discussion of the incentives generated by diaspora politicization and formal access to the political process in the home country (Burgess, 2014;Lafleur & Sánchez-Domínguez, 2015).By contrasting the motivations linked to the host and home country, our approach seeks to identify the degree to which migrants channel their voting behavior as a result of their transnational lives.Relatedly, the second contribution is the addition of individual characteristics to the existing knowledge that migrants' voting behavior in their home country is embedded in the political context of the two countries to which they are connected (Belchior et al., 2018;Ciornei & Østergaard-Nielsen, 2020).The analytical model suggested in this article complements earlier accounts, in that instead of looking at the host-home country linkages as favoring electoral turnout, we test the effects of several factors linked to each of the two countries.Without downplaying the importance of institutional or system-level variables, our study explores the relevance of attitudes and behaviors in both the host and home country.Finally, we advance the empirical knowledge about the individual motivations of voting behavior beyond the often-studied demographic and socio-economic characteristics (Bevelander & Pendakur, 2009;Burgess, 2014;Escobar et al., 2015).
Section 2 reviews the literature on the motivations for transnational electoral participation and provides the theoretical underpinnings for several testable hypotheses.Section 3 includes details of the study's data and methodology.In Section 4, we present the main findings and interpret the results, before Section 5, with the conclusions, where we discuss the broader implications of this study and provide some directions for further research.

Host vs. Home Country Determinants
The individual-level determinants for migrants' turnout in home country elections have been often investigated from two different angles, which are reflected in the exposure, transferability, and resistance theories that have been developed in relation to migrants' attitudes towards their host and home country, respectively.The exposure theory posits that migrants abandon the values they possess before they arrive in the host society and change their behaviors when they come into contact with the values and institutions of their new place of residence.The transferability theory argues that migrants use their pre-migration values to adapt to their host society, while the resistance theory explains that migrants hold onto the values acquired during their socialization in their home country prior to migration and their actions reflect these values (Tsuda, 2012;van Londen et al., 2007;Voicu & Comşa, 2014;Wals, 2011;White et al., 2008).
We build on these theories to test the explanatory power of several variables associated with the two main categories of explanations covered by these theories (host vs. home country).The analytical framework we deploy in our study is presented in Figure 1.The following lines formulate testable hypotheses for both main categories.The theoretical arguments behind the four hypotheses related to the host country set out the idea of migrants' limited exposure to new values and their poor integration in the new environment of residence.

Politics and Governance
Extensive time spent by migrants in the host society allows them to become accustomed to it, to understand and develop an interest in its functioning, and to learn about possibilities to engage (Togeby, 2004).
For example, the longer migrants live in host societies with higher political participation, the more they will vote (Mügge et al., 2021;Voicu & Comşa, 2014).A long period of stay is conducive to better adaptation and engagement in the host society and to alienation from their home society (Finn, 2020;Gherghina & Tseng, 2016).On the contrary, migrants who have only stayed for a short period in their host country maintain connections to their home country and are likely to express their voice in home elections.The same logic applies to the planned stay.The intention to spend more time in the host country enhances interest in the host society and boosts migrants' motivations to create connections because they will pay off in the long run (Umpierrez de Reguero & Finn, 2023).Migrants who plan for longer-term integration into the host society are expected to distance themselves from their home country, including its elections (Gherghina & Tseng, 2016).Short-planned stays often mean that the host society is seen as temporary, so the migrants may avoid investing resources in adapting to it and their main anchor may remain the home country.
Migrants' level of integration in the life of their host country is lower if they do not engage in their new local community.A high degree of civic engagement reflects both the efforts made by migrants and the opportunities that they have to learn about the social dynamics of the host country (Gherghina & Tseng, 2016;Lee et al., 2018).Low engagement can mean that migrants disregard the possibilities in the host country and continue to nurture ties with their home country.Finally, problems encountered in the host country with respect to their legal status, workplace arrangements, or language can lead to migrants feeling unwelcome and uncertain about their future.In these circumstances, they can maintain ties with the stable society at home and be more oriented towards expressing a voice in their home elections.Following all these arguments, we expect that relative to their experience in the host country: H1: Migrants with a shorter stay are likely to vote in home elections; H2: Migrants with a shorter planned stay are likely to vote in home elections; H3: Migrants with lower community engagement are likely to vote in home elections; H4: Migrants who have encountered problems are likely to vote in home elections.
The theoretical underpinnings of the four hypotheses presented relate to the home country ties combined with the society of origin and the migrants' desire to return to that society.Political interest is one of the most common drivers for voter turnout (Blais, 2006;Tsuda, 2012).The mechanism is straightforward: Individuals who are interested in the political process are more inclined to cast a vote because they believe there is something at stake for them and that the act of participation is important.This variable has also been considered relevant to the electoral behavior of migrants, with the general expectation that those with an interest in the political system of their home country may turn out to vote in elections there (Escobar et al., 2015;Lafleur & Chelius, 2011;Spies et al., 2020).Attachment to the home country is another variable that could influence voter turnout.This rests on the theory of social identity according to which people differentiate between in-groups and out-groups (Huddy, 2001;Tajfel & Turner, 1979).When the national identity (of their home country) is salient for migrants, they will associate themselves with that group.A high satisfaction in belonging to that group, i.e., a strong emotional attachment to the home country, can stimulate electoral participation in home country elections (Simon & Grabow, 2010;Spies et al., 2020).
Transnational home visits by migrants are another indicator of national identity practice.Earlier research shows that migrants often visit their home countries to visit family and, in some instances, these family-motivated visits are combined with business or educational journeys (Marschall, 2017).Other studies find that the main purposes of travel are the nurturing of kinship relations, cultural obligations, and maintaining family values (Feng & Page, 2000;Hung et al., 2013).In general, migrants' visits reflect the persistence of symbolic and social ties with the homeland (Waldinger, 2008), which can also push them to vote.Elections are means through which they can influence and to some extent control what happens in the society to which they continue to feel close.Migrants' return intentions, which are defined as their plans to move back one day to their home country, can also shape voter turnout.Migrants may decide to return if they encounter problems with work, family, and socio-cultural integration in the host country (Bettin et al., 2018;Bonifazi & Paparusso, 2019;Constant & Massey, 2002), or if they perceive or expect positive benefits in social, economic, or family life if they return to live again in their home country.Migrants planning to return in the near future may pay more attention to events in the home society and seek to influence its political development because that will affect their lives directly.Following these arguments, we expect that relative to the home country: H5: Migrants with a high interest in their home country's politics are likely to vote in home elections; H6: Migrants with high attachment to their home country are likely to vote in home elections; H7: Migrants who often visit their home country are likely to vote in home elections; H8: Migrants with short-term return intentions are likely to vote in home elections.
In addition to these main effects, we control for age, education, and gender since earlier research indicates that these can influence voter turnout among migrants (Bevelander & Pendakur, 2009;Leal et al., 2012;Mügge et al., 2021;Wass et al., 2015).Apart from these variables, we control for several immigrant-specific variables that may have had an effect on migrants' voter turnout, such as the existence of ethnic networks in the host country, the perception of discrimination in the host country, or the existence of homeland parties that actively campaign abroad (Burgess, 2014;Spies et al., 2020;Umpierrez de Reguero & Finn, 2023).In addition, we also control for general variables such as the perceived usefulness of political parties in the home country, trust in the home country's parliament, and social class (Mügge et al., 2021).However, none of these were observed to have either large or statistically significant effects on the level of knowledge so we do not report them in the article to keep the statistical models parsimonious.

Data and Methodology
We use individual-level data from a survey conducted in June-July 2022 among first-generation migrants from Romania.The Romanian migrants were selected as the subject of study for four reasons that make this the most likely case to identify voter turnout in home elections.First, the Romanian diaspora is one of the largest in the world relative to the country's population (International Organization for Migration, 2020) and, since 2008, has had dedicated seats in the Romanian Parliament.Romanians abroad thereby form a separate constituency that is represented by two senators and four deputies in the national legislative elections.Second, since 2016, there has been relatively high access to voting through postal ballots and polling stations in host countries.Third, Romanian migrants have voted extensively in the past, especially in the presidential elections, and Romanian parliamentarians directly address the diaspora's priority issues in their speeches (Gherghina, 2015;Gherghina et al., 2022).Fourth, the Romanian diaspora has previously engaged in transnational digital activism in support of protests in Romania (Mercea, 2018), which reflects an interest in the state of affairs in the home country.
The dataset includes 1,058 respondents who provided complete answers to the survey.The dropout rate was around 25% of the total number of people who started the survey (1,372 potential respondents).
No specific question triggered the abandonment of the survey.We used a purposive sampling technique in the form of maximum variation sampling to increase the variation in several key variables for research (Emmel, 2013).Only estimates are available on the number of Romanian migrants, as official statistics about their number or profile are lacking, thus the drawing of a probabilistic representative sample was not possible.Although the findings cannot be generalized to a broader population since the sample is not representative of a broader population (Schreier, 2018), they are nevertheless valuable for an internally diverse segment of the population that cannot otherwise be studied.The sample has high variation in terms of migrants' age, education, gender, the area in which they lived prior to migration, country of residence, and length of stay.The distribution of some variables closely matches the limited available information about the broader population of Romanian migrants.For example, the respondents to our survey are an average of 41 years old, close to the average age of Romanians gaining residence abroad, which is 38 years old ("Statistica românilor stabiliți," 2021).The respondents come from 31 countries and roughly 80% of those who filled in the online survey live in one of the six countries that have traditionally been the preferred destinations of Romanian migrants: France, Germany, Italy, Spain, the UK, and the US (Vintila & Soare, 2018).Most of the remaining 20% live in other European countries or Canada, but some reside in countries that are geographically far from Romania such as Australia, South Africa, Taiwan, or the United Arab Emirates.
The respondents were neither pre-selected nor part of a pool of available individuals.The survey was conducted online and was distributed through messages on Facebook groups, discussion forums for Romanians living abroad, and e-mails sent to representatives of Romanian associations and organizations.
The questionnaire covered several main themes including challenges in the host country, ties with the home country, political attitudes and behaviors, and post-materialist values.The use of social media to collect responses has advantages such as cost-effectiveness, broad audience reach, the absence of observer bias, and the potential to explore (Lehdonvirta et al., 2021).However, two main disadvantages are the coverage bias rooted in unequal internet access (van Dijk, 2005) and self-selection bias due to topic saliency (Bethlehem, 2010).The latter is reflected in the distribution of the dependent variable (see Table 1 in the Supplementary File)-we have a higher percentage of voters in the sample than in reality.Moreover, there may also be an issue of over-reporting in this survey related to the social desirability of turnout, as is common in voting behavior surveys.In brief, people say that they have voted even when they have not.
The over-reporting can happen when citizens tend to have a stronger sense of civic duty, are more educated, or have political knowledge (Górecki, 2011).In the case of migrants, this can be exacerbated by the fact that they face greater participation costs when living abroad than they would in their home country.
The dependent variable of this study is a cumulative index that measures the vote in the 2019 presidential elections, the 2019 European elections (for Romanian members of the European Parliament), and the 2020 legislative elections.While in other countries the turnout in European elections is substantially lower than in the national legislative elections, this was not the case in Romania for the most recent elections at these levels.The country-level turnout for the 2019 European elections was approximately 51%, while for the 2020 national legislative elections, it was approximately 33%.In the diaspora, the turnout for the 2019 European elections was higher (295,000 voters) than for the 2020 national legislative elections (265,000 voters; Rezultate Vot, 2022).The respondents were asked if they voted in each of these elections, with a positive answer coded 1 and a negative answer coded 0. The final index is measured on an ordinal scale that has values between 0 and 3.
The independent variables are measured in ordinal scales.The length of stay (H1) is measured with a question regarding the total period of time spent in the host country.The available answers are recorded on a five-point ordinal scale: less than six months (1), half a year to one year (2), one to three years (3), three to six years (4), and more than six years (5).The planned stay (H2) is also measured on a five-point scale through the answers to a question about the length of time for which the migrants planned to stay in the country.The values of the scale are the following, with the explanations in brackets made available to the respondents: short-term (undecided, wanted to check first how it is; 1), temporary for one year (2), medium-term (one to five years; 3), long-term (more than five years; 4), and permanent (5).The level of community engagement (H3) uses the self-reported degree of involvement in the problems of the locality in the host country.This degree ranges between very little and very large, coded ascendingly from 1 to 5. The problems in the host country (H4) are represented by a cumulative index that refers to eight problems personally encountered by the migrants relating to their legal status, access to the job market, education, housing, underemployment, environment adaptation, isolation, and language.Each identified problem is given a score of 1, otherwise a 0, and the final index ranges between 0 and 8.
Interest in the politics of the home country (H5) is measured through a standard survey question: "How interested are you in general in Romanian politics?"The available answers ranged from: not at all (1) to very much (6).Attachment to the home country (H6) is measured via a question about how emotionally attached the respondents feel to Romania.The available answers are coded ascendingly from 1 to 5, representing values between: very little (1) and very much (5).Home visits (H7) uses a four-point ordinal scale with never (1) and twice per year or more (4) as answers to the question about how often the respondents had visited Romania between 2017 and 2022.Return intention (H8) is measured through the answers to the question, "On a scale from 0 (not at all) and 10 (very high), how do you estimate your likelihood of returning to Romania and living there in the next two years?"The control variables are measured in a straightforward manner, similar to international surveys.Age was measured in the survey through the number of years at the time of the survey, which was then recorded in age categories ascendingly from 1 to 5 to create the following cohorts: 18-30 years old (1), 31-40 years old (2), 41-50 years old (3), 51-60 years old (4), and over 60 years old (5).Education is recorded as the highest degree completed, coded on an ordinal scale ranging from primary or secondary school (1) to post-graduate studies (5).Gender is a dichotomous variable: 1 for women and 2 for men.
The analysis uses ordered logistic regression.For all the variables, the "DK/NA" answers are treated as missing values and excluded from the analysis.We ran models with country-fixed effects to account for variations in the degree of democracy in the host country which could impact voter turnout (Ciornei & Østergaard-Nielsen, 2020;Wass et al., 2015).The results of these models resemble those without country-fixed effects.For reasons of parsimony and ease of interpretation, this article presents the version without fixed effects.We tested for multicollinearity and the results indicated no highly correlated predictors: the highest value is 0.42 and the variance inflation factor values are lower than 1.31 for every estimate.

Analysis and Results
Romania uses a closed-list proportional representation system for its national legislative and European elections with an electoral threshold of 5% for political parties and 8-10% for electoral alliances and coalitions.This system was reintroduced in 2016 for the national legislative elections after problems with the mixed majoritarian system that was experimented with in the 2008 and 2012 elections (Gherghina & Jiglau, 2012).National legislative elections are organized every four years-usually in the fall-for a bicameral parliament (senate and of deputies) in which the two chambers are elected in similar ways; the country has not had any early elections in its post-communist history.European elections are organized every five years, on a date decided at the European level.Romania has a multiparty system in which majority government has only been possible in isolated instances.Coalition governments usually include three or more parties, but there have been instances where only two parties governed together.The largest party in post-communist Romania is the Social Democratic Party (PSD), which receives an average of one-third of the votes, the party reached its electoral peak of 46% in the 2016 national legislative elections.The party or its predecessors (with different party names) has won all but one of the popular votes in national parliamentary elections since 1990 and has formed the government many times.The National Liberal Party (PNL) is the second-largest party in the country and receives an average of one-fifth of the votes.It has had a continuous presence in parliament since 1996 and has governed many times since then.The third-party with a continuous presence in the Romanian party system is the Democratic Alliance of Hungarians in Romania, which gets 6-7% in every national election, a percentage proportional to the share of Hungarians in the country.The party is included in many coalition governments due to its pivotal role in parliament.
The other two parties that gained seats in the 2020 national legislative elections were Save Romania Union (USR) and Alliance for the Unity of Romanians (AUR), both relatively new-USR was formed in 2015 and AUR in 2019.USR has a strong pro-European stance and joined the coalition government for one year after the 2020 elections.In contrast, AUR is a radical right party with nationalist and anti-EU discourse, which gained 9% in the 2020 elections.In the diaspora, AUR was the third most voted-for party after USR and PNL.The latter two secured most of the diasporan electoral support in 2016.PSD has had very poor electoral results in the diaspora, e.g., in the 2019 European elections it received slightly more than 3% of the votes and, in the 2020 national legislative elections, it got less than 4% of the votes (Rezultate Vot, 2022).Apart from these five parties, the People's Movement Party gained seats in the 2019 European elections.It emerged in 2013 and gained national parliamentary seats in 2016 but failed to gain representation in 2020.
The Romanian presidential elections have been organized every five years since 2004.Until then, the presidential term in office was four years and coincided with the parliamentary term and a two-round majority system is used in which the second round is a runoff between the top two candidates if no candidate gets an absolute majority of votes (relative to the size of the electorate, not to turnout) in the first round.The diaspora generally does not support PSD presidential candidates.For example, in 2019, the PSD candidate received less than 3% of the votes in the first round (out of roughly 675,000 votes) and 6% in the second round (out of almost 950,000 voters; Rezultate Vot, 2022).
Romanian migrants have been able to vote in home elections since 1992, the first year when elections were organized for complete terms in office of the parliament and country president after the regime change in 1989.The first term in office (1990)(1991)(1992) was considered temporary and was half of the usual term.Since 2008, Romanian migrants have been represented in parliament by two senators and four deputies, while since 2016 the diaspora has been able to use postal voting for legislative and presidential elections.Romania is a semi-presidential regime in which the country's president is elected directly by the population for a five-year term.The voter turnout of Romanian migrants increased from roughly 45,000 in 1992 to 295,000 in the 2019 European elections, falling slightly to 265,000 in the 2020 legislative elections.A record turnout of 950,000 migrant voters was recorded in the second round of the 2019 presidential elections (Rezultate Vot, 2022).This increase took place in the context of high emigration from Romania over the past three decades.Estimates position the total number of Romanian migrants at somewhere between six and eight million (Pavaluca, 2022).
Figure 2 presents the distribution of voter turnout among the survey respondents.Approximately half of the respondents had voted in one election or none, while roughly one-third declared that they had voted in all three elections.The sample is likely to be biased towards respondents who voted since this share can hardly match the numbers provided earlier in this paragraph.We ran three ordinal logistic regression models: one for the variables related to the host country, the second for the variables related to the home country, and the third a full model including all the main effects plus the control variables.The following lines interpret the results for the full model only since the effects are consistent in terms of size and statistical significance across the three models (Supplementary File, Table 2).Among the host country variables, we find empirical support for the first two hypothesized relationships.Romanians who have lived for a shorter period of time in the host country and who only plan to stay in the short-term vote more than those who have settled in a host country or plan a permanent stay (Figure 3).This finding is in line with the arguments from the literature (Gherghina & Tseng, 2016;Umpierrez de Reguero & Finn, 2023) and could mean that the surveyed migrants who see the host country as a temporary stop in their lives vote more in the home elections due to uncertainty about their future.They do not know where they may live in the future and their vote influences the politics in a potential destination.

No vote One elecঞon Two elecঞons All elecঞons
The effect observed for community engagement (H3) goes against the theoretical expectation and indicates that the migrants who are highly involved in the local community in their host country vote more in home elections compared to migrants who are not.One possible explanation for this finding is the existence of a positive relationship between the two variables, which has long been established in the literature.High involvement in civic associations provides people with the skills and experiences that prepare them for future political activity, resulting in higher political participation (Brady et al., 1995;Olsen, 1972;Putnam, 1993).
Romanian migrants appear to use the skills acquired through social engagement in their local communities in the host country to vote in their home country.This finding is consistent with earlier results on other groups of Romanians abroad (Gherghina & Tseng, 2016).We find no evidence for the effect of problems in the host country (H4), which means that the hurdles encountered by migrants in the host country do not make them vote in home elections as a way to safeguard their future or as a refuge from these problems.Interest in home politics (H5) has a strong effect on voting in home elections.Apart from confirming the arguments for the positive relationship between these two variables, this result also indicates that living abroad is not equivalent to a drop in political attentiveness among Romanian migrants.This is in line with previous observations, according to which life abroad does not hinder attention to politics from a distance (McCann et al., 2019).There is also a strong effect of home visits (H7) on voter turnout, which may indicate that electoral participation is driven by a sense of social belonging manifested through continued physical encounters with home society.However, the sense of belonging in the form of emotional attachment to the home country makes no difference in terms of voting.There is also no effect of the return intention, which implies that the surveyed migrants' electoral participation is not linked to a physical presence in the community in which they vote.Their turnout could be driven by considerations relating to their family or social network in the home country or by a broader sense of duty as citizens of the home polity, irrespective of their distance and living intentions in the future.

Length of stay
Among the controls, education is the only one that has strong and statistically significant results.Highly educated migrants are approximately 1.25 times more likely to vote than those with low levels of education.
The absence of an effect for age shows that younger and older surveyed migrants are equally likely to vote, which contrasts with the usual picture from elections in Romania in which the older population turnout is considerably higher.One reason for this observation is the low share of young people in the sample: only 13% were between 18 and 30 years old.

Conclusions
This article has sought to understand the drivers of voter turnout in home elections among Romanian migrants.It contrasted the variables related to their host and home country and produced three main findings.First, the home country variables seem to have a stronger effect on turnout than experiences in the host countries.The model has a better fit (pseudo  2 ) and the effect size for the variables is larger.This means that migrants engage in home elections because of the ties they maintain with their society of origin, rather than because they feel alienated from the host society.Second, domestic civic activity enhances political activity in a transnational context.Migrants who are more socially engaged in their host communities vote more in their home country compared to socially passive migrants.Third, turnout is driven by political interest and frequent contact with the home society rather than by distant assessments of the facts.This observation indicates that voting by diaspora members appears to be much more of an informed decision than an activity engaged in by migrants who are out of touch with the realities at home.
These results have implications for the broader study of migrants' transnational political participation.
Although the sample used in this study is non-representative, the findings indicate the existence of several trends that are informative for the scientific community.For example, we show that migrants' voting behavior is not exclusively linked to their (non-) integration in their host country or to their ongoing ties with their home country.Instead, it is the result of their transnational lives.We find that specific attitudes and behaviors beyond socio-demographic characteristics in both the host and the home country foster their electoral participation.The turnout of Romanian migrants is driven by political interest, civic values, and strong ties to the community where the vote is cast.Such characteristics correspond to informed and responsible behavior, which could bring benefits in the medium and long run to political representation in the country and beyond its borders.While these findings are hardly generalizable due to the sample characteristics and the features of the Romanian diaspora, they invite future research that tests their relevance in other political settings and migrant populations.They also advance the existing knowledge about the electoral participation of migrants and could represent a useful starting point for comparative analyses.Further work may refine these conclusions with a focus on the relevance of knowledge, civic duty, and partisan attachments, factors that were not explored in this study.

Figure 1 .
Figure 1.An overview of the analytical framework.

Figure 2 .
Figure 2. The voter turnout of respondents in the three elections organized in the home country.

Figure 3 .
Figure 3.The effects on voter turnout.