Derek S. Hutcheson and Jean-Thomas Arrighi (2015), ‘“Keeping Pandora’s (ballot) box half-shut”: a comparative inquiry into the institutional limits of external voting in EU Member States’, Democratization 22(5): 884-905 [ISSN: 1351-0347] This is a pre-proof manuscript copy and it may contain minor typographical differences from the final published version. Please refer to the final published version if citing [http://dx.doi.org/10.1080/13510347.2014.979161] Contact details: Dr Derek Hutcheson, Associate Professor, Department of Global Political Studies, Faculty of Culture and Society, Malmö University, SE-20506 Malmö, SWEDEN e-mail: [email protected]. Tel: +46 40 6657379 Dr Jean-Thomas Arrighi, European University Institute, Robert Schuman Centre for Advanced Studies Via delle Fontanelle 19, 50014 San Domenico di Fiesole (FI), ITALY. E-mail: [email protected]. Abstract: The article compares the institutional constraints that limit the potential electoral impact of external voting in national legislative elections in the 28 Member States of the European Union (EU). It shows that the discrepancy between policy aims and outcomes can be mainly attributed to a variety of institutional constraints restricting the scope of the policy (through residence and professional qualifications); limiting eligible voters’ access to the ballot (through cumbersome registration procedures and voting methods); and reducing the electoral weight attributed to their votes (through distinct modes of representation). It argues that the discrepancy is at least partly the result of a combination of electoral and normative concerns about the influence that external voters could and should have in elections. Institutional restrictions on the franchise of external citizens may be interpreted as a way to keep the ‘Pandora’s Box’ of unexpected electoral consequences half-shut, by extending the suffrage to a traditionally excluded electorate while at the same time moderating the implications. Keywords: external voting; EU; swamping; tipping; access to the ballot HUTCHESON AND ARRIGHI 2015/cont. Introduction In 2006 Mexico joined the growing group of states that have extended the right to vote in national elections to their citizens abroad. With an estimated 12 million Mexican nationals residing in the United States alone, the introduction of external voting rights had significant potential to alter the electoral balance in a country that was about to experience one of the most highly disputed electoral competitions in its history. Yet in the presidential elections which took place a few months later, only 32,000 voters cast a ballot from abroad – less than 1% of the estimated expatriate population of voting age. Accordingly, the electoral impact of the reform was negligible. This remarkably low turnout can be mainly attributed to an extremely cumbersome registration procedure, which was purposively introduced in order to mitigate the uncertain yet potentially determinant impact of external voting on electoral results.1 This example is illustrative of a trend that can be observed on a global scale. On the one hand, the last three decades have seen a sharp increase in the number of states granting the right to vote in national elections to their citizens who reside temporarily or permanently abroad. Yet in the overwhelming majority of cases, the extension of the franchise has failed to alter expected electoral results significantly. In other words, there has been a noteworthy gap between the policy aims of extending the suffrage to all citizens irrespective of their place of residence, and the policy outcomes, characterised by low turnout and marginal electoral impact. 2 HUTCHESON AND ARRIGHI 2015/cont. In this article we compare the institutional constraints that limit the potential electoral impact of external voting in national legislative elections in the 28 Member States of the European Union (EU). We do not focus on the question of whether external citizens, normatively, should be able to vote,2 or the reasons why states may enfranchise their external citizens.3 Rather, we focus on how this proceeds. We show that the discrepancy between policy aims and outcomes can be mainly attributed to a variety of institutional constraints restricting the scope of the policy (through residence and professional qualifications); limiting eligible voters’ access to the ballot (through cumbersome registration procedures and voting methods); and reducing the electoral weight attributed to their votes (through distinct modes of representation). While we do not seek to establish a clear causality and systematically test a hypothesis, we nonetheless argue, based on a combination of existing scholarship and detailed empirical investigation of current electoral laws, that the discrepancy is at least partly the result of a combination of electoral and normative concerns. Specifically, we identify an electoral argument that external citizens may have undue influence over the outcome of national elections either by diluting the votes of existing voters or tipping the result of an election, and a normative argument that external citizens should have a lesser claim to influence in homeland politics because they are less affected or have a lesser stake in the polity. From this perspective, institutional restrictions on the franchise of external citizens may be interpreted as a way to keep the ‘Pandora’s Box’ of unexpected electoral consequences half-shut, by extending the suffrage to a traditionally excluded electorate while at the same time moderating the implications. 3 HUTCHESON AND ARRIGHI 2015/cont. The article is organised in two main parts. First, we discuss the normative and practical aspects of extending the franchise to expatriates. Second, we successively examine four aspects of the institutional set up in a comparative perspective across the 28 EU member states, using rich qualitative and quantitative data recently compiled by the European Union Democracy Observatory on Citizenship (EUDO- Citizenship). Expansive Aims, Restrictive Outcomes: the Institutional Limits of External Voting The idea that citizens living outside a country’s borders should continue to enjoy voting rights in their country of citizenship has gained increasing traction in recent years. Most states until the 1960s either formally limited the right to vote to residents alone, or neglected to set up institutional pathways to enable their citizens to cast a ballot from abroad. Their inclusion in the demos was seen as an anomaly that could hardly be reconciled with the requirements of representative democracy.4 By the early 21st century, the reverse was true (a recent survey indicated that only 45 states, out of 174 worldwide for which data are available, still completely disenfranchise their nationals living abroad),5 reflecting a wider shift in attitudes towards external citizens from benign neglect or outright hostility toward their gradual inclusion in the political community (see Lafleur in this volume).6 In most instances, however, the electoral impact of the ‘expatriate vote’ has been insignificant. Only occasionally has the mobilisation of external voters changed the electoral balance to the benefit of political parties or candidates that had failed to 4 HUTCHESON AND ARRIGHI 2015/cont. obtain a majority of domestic votes already. The 2006 general election in Italy provides one such exception. The scale of mobilisation and the clear partisan preference for left-wing parties among a newly-represented electorate of Italians abroad came as a surprise to most, and especially to Silvio Berlusconi’s right-wing coalition partners who had played an instrumental role in the introduction of external voting rights in 2001, under the assumption that they would be electorally rewarded for it.7 These seats were to prove crucial to the coalition. In a similar vein, the narrow re-election of Traian Basescu in the 2009 presidential elections in Romania has been widely attributed to the ballots cast by ‘Romanians abroad’ who, despite a relatively low turnout, voted en masse for the ruling party.8 Yet these much-publicised cases are the exception rather than the rule. The most important reason for this may be that claims about an ‘age of migration’ are over- stated, as the proportion of individuals living in a country other than the one in which they were born remains marginal as a percentage of the world population. For better or worse, most people still live sedentary lives and the membership boundaries of the demos and the territorial boundaries of the state remain by-and-large congruent.9 However, in the growing number of countries that do enfranchise non-resident citizens, turnout among external voters has been consistently lower than among domestic ones. Such differential turnout may derive from the fact that political parties have neglected their external constituencies, while external voters have neglected domestic politics. On the one hand, the organisational challenges of campaigning abroad require considerable resources that political parties may lack or be unwilling to mobilise. On 5 HUTCHESON AND ARRIGHI 2015/cont. the other, voters who left their country of citizenship long ago, or may never even have stepped foot in their putative ‘homeland’, may have limited interest in the politics of their country of origin. The interaction between an apathetic constituency and an indifferent political class goes far in explaining the low levels of political participation among an electorate that is generally less informed and more remotely affected by elections than their counterparts residing in the country. However, the expected costs and benefits for both parties and external voters are also determined by the institutional set up which regulates the potential electoral impact of non- residents. Hence, rules matter, and a thorough comparative examination of their scope and implementation is required in order to understand the distinctive opportunities and constraints that shape the behaviour of both voters and parties. Before turning to the aspects of the institutional set up limiting the electoral impact of external voting, we briefly discuss two reasons why states may be encouraged to introduce such restrictions in the first place. The introduction of external voting rights raises a combination of electoral and normative concerns. Electorally, external citizens may be perceived to have undue influence over the outcome of national elections, either by diluting the votes of existing voters or tipping the result of an election. Existing scholarship has pointed to the fact that the anticipated electoral impact of introducing external voting rights shapes to a considerable extent the scope and implementation of the policy.10 Whilst desirable from a Rawlsian perspective, the veil of ignorance has long been lifted on the preferences and positions of citizens residing in liberal democracies, but the socio-demographic characteristics and, most importantly, ideological inclinations of 6 HUTCHESON AND ARRIGHI 2015/cont. expatriates are largely unknown to political elites potentially competing for their votes. In political discourses, expatriate populations tend to be reified as ‘diasporas’, homogenous groups whose members share a clearly identifiable set of objective attributes and are collectively bounded by a common attachment to the homeland.11 However, their dispersion across vast geographical areas and the broad range of reasons that led to their absence from the homeland suggests that external citizens may in fact be a highly diverse and fragmented constituency, whose electoral behaviour is both volatile and uncertain. Bauböck has drawn an insightful distinction between the fears of ‘swamping’ and ‘tipping’.12 In the former, the pool of external voters is seen as disproportionately large compared to the domestic electorate and threatens to ‘swamp’ its influence. In the latter, the external vote may be expected to play a decisive role in determining the outcome of the election. This uncertainty is more pronounced before voting rights are introduced and tends to decline from one election to the next. If the tipping and swamping scenarios fail to materialise in the medium term (or prove greater than expected), we may expect modifications to the initial institutional constraints. While political parties may have contrary expectations – with some anticipating electoral rewards and others, liabilities – the necessity to reach a minimal consensus militates against the adoption of an ambitious reform, the consequences of which are unpredictable. Normatively, the argument focuses on whether external citizens should be represented on equal terms with resident citizens. It can be argued under different principles of enfranchisement that the degree of influence that should be given to 7 HUTCHESON AND ARRIGHI 2015/cont. different groups can be differentiated according to the extent to which they are affected by, subjected by or stakeholders in the state’s laws. 13 External citizens may not be immediately influenced by most day-to-day laws of a state in which they do not live. Thus the representation given to them need not be equal with that of resident citizens.14 Such arguments have been invoked in various countries against the extension of the franchise to external citizens. In Britain, for instance, initial parliamentary opposition to the principle derived from the fact that government formation in the UK is a by-product of an electoral system that explicitly elects individual local representatives without any proportional seat allocation at national level, and that a person’s link to a specific locality diminishes rapidly once he or she leaves.15 In the rest of this article we explore the ways in which various institutional frameworks interact with the enfranchisement of external voters to mitigate the potential impact of a population that is widely seen as having a lesser claim to influence over the polity than its fellow citizenry residing in the country. Our analysis complements earlier comparative studies which have found that that rights of enfranchisement beyond the realm of permanently resident national citizens vary widely by country, by electoral level and by category of exception.16 We place our focus on four factors and institutional constraints that contribute to determining the potential electoral impact of non-resident citizens on the electoral process: (1) Differences in the size of the external citizenry; (2) Restrictive rules of eligibility; (3) Restrictive rules of access to the ballot; 8 HUTCHESON AND ARRIGHI 2015/cont. (4) Different modes of representation. Electoral concerns of ‘swamping’ and ‘tipping’ are addressed by all four modes of restriction. ‘Swamping’ is most likely to occur where external groupings are large, in which case restricting the number of eligible citizens can be achieved either by using nationality laws to restrict the extra-territorial perpetuation of citizenship, or by limiting the eligibility of external citizens to vote in the first place, through a combination of residence-based and professional qualifications. Moreover, even if all or most external citizens are enfranchised in principle, the extent to which they can participate in practice depends on how easy it is to register and to cast a ballot. Once the external electorate has voted, the extent to which it threatens to ‘swamp’ or ‘tip’ the domestic electorate depends on how the non-resident voters are represented. Discrete or ‘special’ representation is often depicted as an enabling factor in guaranteeing that the distinct interests of the external community are represented in the life of the home state,17 though Bauböck has pointed out that non- resident enfranchisement should be based on these citizens’ continued stake inside the polity.18 Discrete representation can also be used to mitigate the potential swamping effect by reducing the external electorate’s relative representation in relation to native citizens’. Such separate representation leads, however, to a greater danger of ‘tipping’, which is most likely to occur where the external electorate is cohesive and collectively mobilised. This can be addressed by each of the four types of restriction but most effectively by dispersing the votes of external citizens and assimilating them locally across the country, thus dissipating their effect and reducing their ability to act as a cohesive bloc. We do not postulate clear causal mechanisms 9 HUTCHESON AND ARRIGHI 2015/cont. between these factors, nor do we suggest that these restrictions are always deliberate acts on the part of political elites with the explicit aim of alleviating fears of swamping or tipping. Moreover, we acknowledge that other factors may play a role and that the sample of 28 EU countries, whilst a substantial and coherent geographical set, may be not be large enough for individual national factors to be discounted from the equation. Nonetheless, we suggest that normative concerns may at least play a role in setting the agenda surrounding external enfranchisement, and that there is some degree of correlation between the openness with which states embrace their external citizens as part of the demos and the potential impact that that electorate may have. The Participation of External Citizens in Practice: A Comparative Overview of EU Member States To examine the qualitative differences between the electoral rights of external citizens we focus on four levels of restriction in one form of election – national parliamentary elections – across the 28 EU Member States. National parliamentary elections provide greatest comparability as they are held in every state, usually relate to the most powerful level of government within the country, and are fully under national jurisdiction. We utilise data recently compiled by the European Union Democracy Observatory on Citizenship (EUDO-Citizenship), in the form of electoral rights data drawn from a comprehensive database of current electoral laws across the 28 EU members.19 10