In the last few decades, many countries have enfranchised their emigrant populations. Voting from abroad is not new, but was historically reserved for diplomats or military personnel. However, starting in the 1990s, parallel to a growth in international migration, an increasing number of countries enacted laws allowing expatriates to participate in some fashion in the electoral process. Twelve Latin American nations have enacted these rights. Colombia (1962) was the first country in Latin America, and the second in the world, after Indonesia (1953), to extend suffrage to emigrants. Colombia was also a pioneer in Latin America in allowing expatriates to participate in congressional elections and elect their own representative to the Lower House (enacted in 1991 and exercised for the first time in 2002). Nonetheless, no studies of elections among Colombian expatriates had been done[i] until we decided to look at this case, not only because of its intrinsic interest but also to expand our general knowledge about expatriate voting by serving as an example for future comparisons.
Two main questions guided our study: First, who gets involved in Colombian electoral politics from abroad, and why? And second, what determines the choices of Colombian expatriates, and how does this “calculus of voting” differ in comparison to that of voters in Colombia proper? These questions about electoral choices are not new. On the contrary, they are classic questions that political scientists have been asking for a very long time. Our work differs in that we bring these questions to a relatively new and understudied type of constituency that has, nevertheless, grown significantly worldwide in recent years as expatriates have expanded their political rights. While the questions and the survey methodology are not new, the characteristics of expatriates as voters and their multiple locations demand particular research strategy designs. We take into consideration not only that voters are outside of the territory, which implies different access to information, different exposure to the media and political campaigns, and so on, but also that they are widely dispersed and located in very different contexts, which possibly influences their voting preferences.
In collaboration with a group of scholars and students from universities in France, Spain, the United Kingdom, and the United States, we carried out 1) a representative survey of the potential electorate, and 2) exit polls during the 2010 Colombian legislative and presidential elections in five cities in Europe and the United States. Simultaneously, comparable surveys were administered in five cities of high migration areas in Colombia. In total, 3,369 Colombians took part in the study, with sampling concentrated in Miami and New York (U.S.), Madrid (Spain), London (U.K.), Paris (France), and Barranquilla, Bogotá, Cali, Medellin and Pereira (Colombia). The inclusion of various cities abroad allows us to study the influence of the place of residence, while the inclusion of Colombian cities allows us to compare between migrants and non-migrants. Half of the survey interviews (N=1,567), corresponding to the potential electorate, were conducted before the Colombian congressional and presidential elections (March 14 and May 30, 2010). Expatriate Colombians seeking services at one of the consulates (e.g., renewal of passports, legalization of documents, etc.) in New York, Miami, Paris, London, or Madrid were randomly selected for participation in our study. The objective of this pre-election survey was to gauge political attitudes within the general Colombian emigrant population, including both absentee voters and non-voters. On the days of the elections, an additional 1.802 interviews were conducted with randomly selected voters as they exited polling places.
In order to vote, Colombian expatriates are required to register in the consulates and embassies and have to cast ballots in these locations or other voting stations created by the government in areas with large numbers of expatriates (New York and Miami, for example). Abstention is high in Colombia (55% and 51% for the legislative and presidential elections of March 14 and May 30, 2010, respectively) but much higher abroad. Colombians living abroad were estimated at 3.3 million in 2005 (8% of the total population), but only 105,312 expatriate voters participated in the presidential and 48,937 in the Congressional elections in 2010. We estimate abstention to be 74.6%, based on registered voters and 95% based on potential voters.
Who Votes Abroad?
In our first study,[ii] we addressed questions concerning expatriate electoral participation by analyzing the interviews (N=1,567) carried out before the elections. Specifically, we explored the influence of place of residence on voting intentions (expected participation) in the upcoming congressional and presidential elections. In order to estimate the extent to which participation was the result of residence in particular localities or of other factors, we first regressed the dependent variable on regional dummies. We then estimated multivariate regression models to control for the effect of several other possible factors influencing involvement. We found that neither socioeconomic characteristics and resources (education, income, and civil status) nor social capital variables (church attendance, living in a Colombian or Latin American neighborhood, being part of a community organization), all of which are critical in explaining electoral participation within national boundaries, had significant effects on the intentions of expatriates to participate. Direct contact with political campaigns, which is relatively low abroad, did not have much influence either. Motivational variables, on the contrary, proved to be very influential, particularly interest in Colombian politics and pre-migration electoral experience.
These findings underscore the argument of some scholars[iii] that emigrants who were electorally active at “home” remain active abroad; relatedly, we find a positive relationship between age and electoral participation. Ideology also plays a role, since left-leaning individuals were less inclined to participate, particularly in presidential elections. Our analysis shows that institutional variables, such as registration, are important factors in shaping electoral participation, a result to be expected given not only the challenge of setting up the infrastructure to carry out elections abroad, but also communication and information barriers within the expatriate electorate. The 2014 elections were a good test for these findings because the registration period was expanded from four weeks to ten months, and elections were held abroad over a period of seven days rather than a single day, as in earlier elections. However, and in spite of an increase of 25.3% in registration rates, participation itself did not increase. On the contrary, involvement decreased by 7.3%. Given this pattern, future work should investigate in greater detail the relationship between the mechanisms and infrastructure for administering elections abroad and actual turnout.
Our findings regarding migration factors (years abroad, desire to return, naturalization status, the sending of remittances, and trips back home) might appear contradictory at first because we found a moderately significant effect for U.S. naturalization in both congressional and presidential elections. Yet the impact of migration factors depends on other aspects because once other variables were introduced into the model it disappeared. At the same time, we found that time abroad was not significant in either of the elections. That means that neither naturalization nor years abroad, by themselves, influence intention to participate in elections. However, and here is the apparent contradiction, we found that emigrants who remit less (a moderately negative effect) and those who expected to return were more interested in participating in presidential elections than their counterparts. This apparent contradiction can be interpreted in light of scholarship which has underscored that political transnationalism is more common not among the recently arrived, but among the more stable i.e., those who have secured their position in the host country.[iv] Given their interest in Colombian politics and, in some cases, combined with their concrete or hopeful expectation to return one day, the more stable migrants, who remit less and have perhaps naturalized, are the ones with the time and resources to get involved in Colombian politics.
Going back to our initial question concerning place of residence, we found that after controlling for the effect of all the factors previously discussed, location does matter. The probability of expecting to vote was much higher in New York (0.34% and 0.47% for congressional and presidential elections, respectively) than in any other of the other cities included in the study, especially Miami (0.11 % and 0.20%).
How do People Vote Abroad?
In a second piece,[v] we examine how candidate favorability is shaped among Colombian expatriates, and how it differs from preferences among Colombians back home. We based this analysis on the two exit polls carried out on the day of the congressional and presidential elections (March 14 and May 30, 2010) in the five cities abroad and in the five Colombian cities. Because Colombian law prohibits specific questions about voting choices on the day of an election, we modeled respondent preferences, particularly relative preferences between the two main contenders: Juan Manuel Santos, the candidate selected by the incumbent president at the time, Alvaro Uribe, and Antanas Mockus, the candidate of the newly created opposition Green Party. Juan Manuel Santos won in Colombia and abroad, but he won abroad by a greater margin.
Our key dependent variable in this paper is an evaluative scale of the two main contenders. We subtracted the rating of Mockus from that of Santos, so that higher scores indicate a more positive attitude towards the ultimate winner, Santos. We estimated parallel ordinary least square regression models to examine the factors that account for variations in relative favorability ratings between Colombians abroad and Colombians in Colombia. We included local context, social and economic factors, political factors, and media factors.
Some democratic theorists and policymakers have expressed concerns about extending the franchise to voters who do not reside within a country, since emigrants may express preferences that are at odds with citizens who reside “back home.” Reassuringly, we found that Colombians abroad and back home do not differ substantially in the criteria upon which evaluations of the major candidates are based. Socio-economic factors and exposure to the mass media, for example, play similar moderate roles, whereas political factors are more important. Evaluations of President Uribe were equally critical in Colombia and abroad in shaping attitudes towards the two main contenders in the election. Similarly, the Internet was a major player in these elections both at home and abroad. Migration factors, which we expected to shape favorability differently among expatriates, did not have a significant influence. We also found that settlement context (whether one lives in Paris as opposed to Miami, for example) is not relevant in shaping opinions of candidates. In spite of these commonalities, we also observed some differences. Education greatly influences candidate favorability in Colombia, whereas among expatriates, income seems to be more relevant. We also found that religious practice shapes political judgments among expatriates to a greater degree than within Columbia proper.
Our aim in fielding this multisite-transnational study of Colombian expatriates is to contribute fresh insights to a relatively new and important area of research. Such broad-based comparative studies allow investigators to assess the role that emigrants play in elections, representation, and ongoing processes of democratization “back home.”
[i] See Jean-Michele Lafleur (2013) Transnational Politics and the State. New York: Routledge, and Cristina Escobar (2015) “Immigrant Enfranchisement in Latin America: From Strong Men to Universal Citizenship.” Democratization, Special Volume Electoral Rights in the Age of Globalization. (Forthcoming).
[ii] Cristina Escobar, Renelinda Arana and James McCann (2014)“Expatriate Voting and Migrants’ Place of Residence: Explaining Transnational Participation in Colombian Elections.” Migration Studies doi: 10.1093/migration/mnt030.
[iii] See Roger D. Waldinger, Thomas Soehl, and Nelson Lim (2012) “Emigrants and the Body Politic Left Behind: Results from the Latino National Survey.” Journal of Ethnic and Migration Studies 38(5): 711-736.
[iv] See Luis Eduardo Guarnizo, Alejandro Portes, and William Haller (2003) “Assimilation and Transnationalism: Determinants of Transnational Political Action among Contemporary Migrants.” American Journal of Sociology, 108/6:1211– 48; David Leal,, Byung-Jae Lee and James McCann (2012) ‘Transnational Absentee Voting in the 2006 Mexican Presidential Election: The Roots of Participation.” Electoral Studies, 31/3:540:549; and Alejandro Portes, Cristina Escobar, Alexandra Walton Radford (2007) “Immigrant Transnational Organizations and Development: A Comparative Study.” International Migration Review 41/1: 242-81.
[v] Cristina Escobar, Renelinda Arana and James McCann. “External Vote and Political Preferences in the Colombian 2010 Presidential Elections.” Latin America Politics and Society. 56(2): 115-140.