by Rainer Bauböck
Most academics who write about citizenship of the European Union tend to compare it with what they know best: nation state citizenship. It comes as no surprise when they conclude that the current construction of EU citizenship is internally incoherent, exter-nally not sufficiently inclusive and lacking in democratic legitimacy. I agree to a certain degree with these critiques, but I think that they apply a wrong standard of comparison and are therefore likely to promote false solutions. As the EU Treaties have spelled out clearly since the 1997 Treaty of Amsterdam, EU citizenship is complementary or addi-tional to member state nationality without replacing it. National citizenship is a constitu-tive element of EU citizenship and can therefore not serve as an external standard of comparison.
Academic scholars have for quite some time described the EU polity as a multi-layered system of governance and governments. It consists not just of the supranational institutions of the European Commission, the Council, the European Parliament and the Court of Justice of the European Union, but also of the national parliaments and governments of the member states. There is a corresponding system of multilevel citizenship in the Union that needs to be studied and evaluated as a constellation where individuals have plural memberships and where citizenship regimes are connected with each other across levels.
Local, national and supranational citizenship
Such a multilevel perspective avoids regarding EU citizenship either as a postnational alternative to member state citizenship or as a mere appendix filled with a few additional rights that does not deserve the label citizenship in the strong sense of a status of equal membership in a self-governing political community. In order to understand how a multilevel system of citizenship can work, we do not have to invent future worlds or travel far back in history. Every larger democratic state already contains within itself a multilevel citizenship regime. It is true that only a few federal states, such as Austria and Switzerland, formally acknowledge in their constitutions a citizenship of their provinces, but even highly centralized states, such as France, have elections for regional assemblies that enjoy a range of devolved decision-making powers. While unitary and federal constitutions differ strongly with regard to the political status and powers of substate territories, all independent democratic states, apart from micro states and city states, are subdivided into municipalities with democratically elected offices of local councillors or mayors. Local level citizenship is not only a common feature of contem-porary democracies but also a democratic requirement. If central state authorities were in charge of deciding all matters of local government, then the inhabitants of municipali-ties would be unjustly dominated by representatives of national majorities.
Conceiving of democratic states as polities with nested layers of local, regional and state level citizenship is not only a useful analogy for better understanding the EU citi-zenship constellation. Substate citizenships form an integral part of that constellation so that there are not two, but at least three distinct levels of individual membership in the Union that are universally present throughout the EU polity and include all its resident citizens: local, national and supranational citizenship.
Citizenship on a national level
If citizenship is at its core a membership status, then the first task in describing this triple level structure is to analyse the rules that determine who is a member at each level of the polity. For the national level, such rules are laid down in nationality laws. These laws differ enormously with regard to their specific legal provisions and conditions for acquisition and loss of nationality, not only globally, but also within the EU. However, once we compare them with the rules for determining citizenship in supranational and local polities, it becomes obvious that all nationality laws have a common basic struc-ture and purpose.
The fundamental principle of nationality law in modern states is automatic acquisition of citizenship status at birth, either by descent from citizen parents or by birth in the terri-tory of the state. Acquisition of citizenship by naturalisation and loss of citizenship through renunciation or withdrawal are merely corrective rules that serve to resolve marginal discrepancies between a citizenship population determined by birthright and a reference population that states want to exclude or include. The need for such correc-tive devices arises mainly because of migration that generates non-resident populations with, and resident populations without birthright citizenship. Correcting birthright allocation is, however, also exceptionally necessary when international borders change through state breakup and secession or through unification and territorial incorporation. Three different rules have been used for the initial determination of citizenship of popu-lations in newly independent states or incorporated territories:
- a zero option that includes all residents at the point of independence
- a restoration option that refers back to citizenship in an independent predecessor state or
- the transformation of a previous federal entity citizenship into that of an inde-pendent successor state.
The zero option has been chosen by the vast majority of post-Soviet states that had no prior history of independent statehood. Estonia and Latvia opted for a restoration model that excluded most of their large Russian minorities for access to citizenship at inde-pendence. In the violent breakup of Yugoslavia and the peaceful separation of Czech-oslovakia, the previously fairly insignificant citizenship of the various federal republics was upgraded into the new national citizenship of the successor states.
It is crucial to understand that only shifting international borders lead to automatic in-clusion or exclusion of entire territorial populations. Democratic states with stable bor-ders never include automatically first generation immigrants without asking for their consent. One might object that there is the exceptional case of co-ethnic immigrants in Germany and Israel who are automatically naturalised upon entry. However, these groups have been identified as members of the nation prior to immigration and accept-ing the invitation to “return” implies consent to acquire full citizenship status.
Correcting birthright allocation through naturalisation requires therefore an individual application and so does voluntary renunciation by non-resident citizens. Involuntary withdrawal of citizenship by the state is sometimes used as a sanction but may also affect persons who are seen to lack a genuine link to the state concerned – for exam-ple, if they have inherited their citizenship at birth abroad and never take up residence in their ancestors’ country of origin. In any case, in democratic states acquisition and loss that is not based on birthright is regulated by procedures that involve individual consent or qualifications for membership.
Birthright Citizenship and the nation state
What is the purpose of birthright citizenship and how could it be justified? All modern states are constructed as intergenerational political communities and birthright mem-bership is the crucial mechanism that supports their continuity. Citizenship is attributed at birth and normally retained for a full life and it is passed on to subsequent genera-tions. An emphasis on intergenerational continuity is often seen as a core feature of ethnic nationhood. But this is a short-sighted view since all states, including those which are plurinational in composition or embrace a civic conception of nationhood, determine their citizenship through birthright. Moreover, there are distinctly democratic reasons for birthright allocation. Governments of independent states wield comprehensive political powers over their subjects and take decisions that affect future generations in important ways. While this may also be true for some powerful non-state actors, such as big corporations, only political governments can be held accountable by and be made responsive to citizens. If all citizens regarded themselves as mere temporary residents living among other temporary residents, then they would have little reason to support long-term decisions for the sake of future generations. Instead of hoping to win a political argument or election next time round, exit would become the preferred response by defeated minorities who regard majority decisions as contrary to their fun-damental interests or convictions. For these and other reasons, birthright citizenship is essential for maintaining the democratic core idea of a self-governing people.
Yet in contemporary states citizenship at the local level is no longer determined through birthright. (1) Liberal democracies grant internal freedom of movement not only to their own citizens, but to all legal residents in their territory and local governments provide public services to all those residing within their jurisdiction. It is true that most democratic states still reserve the franchise in local elections to their national citizens. But these do not have to apply for local naturalisation after moving to a different munic-ipality; they are automatically included as local citizens with full participatory rights after some time of residence. Moreover, fourteen European states, twelve of which are member states of the European Union, have fully disconnected local from national citi-zenship by enfranchising also third country nationals.
At the local level we find thus a second type of citizenship regime based on jus domicilii, .e. automatic residential membership. Birthright citizenship at state level has a sticky quality due to its strong external dimension. It is not lost through emigration and can be passed on to at least the second generation born abroad. This is also a main reason why plural nationality is becoming more frequent. A growing number of children of migrant origin acquire several citizenships at birth and more and more states tolerate also dual nationality in case of naturalisation or voluntary acquisition of a foreign nationality. By contrast, local citizenship is fluid and generally singular at any point in time. Taking up residence in another municipality leads to automatic acquisition of a new citizenship and automatic loss of a previous one.
This arrangement can again be supported by democratic reason. Local governments are responsible for providing public services to local residents and ought to be ac-countable to these. Discrimination on grounds of nationality is arbitrary from the per-spective of local self-government. But why do arguments in favour of birthright citizen-ship not also apply to the local level? The answer is simply that local residential citizen-ship is not an independent regime, but is nested within intergenerational national citi-zenship, so that every local citizen is also a member of an intergenerational political community – either as an internal citizen of the encompassing state or as an external citizen of a foreign country.
By considering local and national citizenship as a combined multilevel structure, we can see how the two principles of residence and birthright supplement each other. The long-term perspective of democratic community that is supported through birthright at the national level provides a stable background for more fluid memberships at local level. Local citizenships are not for life and easily acquired and lost. Mobile individuals will therefore be multiple local citizens sequentially over the course of their lives, but not simultaneously, since local citizenship has only a very weak external dimension. In democratic states, there is an additional reason for keeping local citizenships singular at any point in time; as provinces and municipalities are integrated into a common structure of government and democratic representation, no citizen should have multiple votes across several substate polities.
EU citizenship as a derivative of the member state nationality
Intergenerational and residential citizenship are the two basic regimes that we find in contemporary democratic polities. EU citizenship represents a third and hybrid type. When asking who are the citizens of the Union, the answer is: the nationals of its member states. Individual membership in the EU polity is therefore neither determined by an EU birthright, nor by residence in the EU, but is derivative of member state na-tionality. Yet the control that member states retain over acquisition and loss of EU citi-zenship is exposed to a powerful force that operates at a transnational level: the right to free movement inside the territory of the Union. This residential aspect of EU citizenship is not only articulated in the narrowly conceived rights of territorial admission, settlement and access to employment but includes a general right of non-discrimination on grounds of nationality and applies also to political rights: EU citizens residing in member states other than their state of nationality can participate in local and European Parliament elections there.
The derivative nature of EU citizenship is not a historically unique construct. The same citizenship architecture was characteristic for early stages of federal statehood in Ger-many, Austria and the United States of America. Switzerland seems to be the only sur-viving case where federal citizenship is formally derived from cantonal citizenship. In Switzerland, as in the EU, the distinct polities of the union enjoy wide powers of self-determination with regard to naturalisation. The important difference is that birthright acquisition and loss of citizenship are regulated by federal law (as they have been in the US since the 14th amendment of 1867) rather than at the level of provincial citizenship. Member state self-determination in matters of citizenship is therefore stronger in the EU than in any of the historical or contemporary federal states. Even the much looser Nordic Passport Union envisaged a minimal harmonization of citizenship laws so that they would become compatible with free movement rights. No such coordination has been possible in the EU although member states can subvert each other’s immigration control by producing EU citizens with free access to the rest of the Union.
EU citizenship and the residence-based right to freedom of movement
The tension between the strictly derivative nature of EU citizenship and its residence-based free movement rights shows also in the differential treatment of EU citizens who reside in their country of nationality and those residing in another member state. Let us call the former ‘first country nationals’ (FCNs) and the latter ‘second country nationals’ (SCNs). The protection of EU citizenship applies in specific ways to those persons who have made use of their free movement rights or who are in other ways involved in ‘cross-border situations’. These persons enjoy, for example, extended rights to family migration that most member states deny to their own FCNs who want to invite “third country national” (TCN) family members to join them. Such instances of reverse dis-crimination have been a notorious sideeffect of a construction of EU citizenship that applies more directly to mobile populations than to sedentary ones. In a series of recent judgments, most prominent among which are the 2010 Rottmann and 2011 Zambrano cases, the Court of Justice of the European Union has expanded the meaning of cross-border situations to cover also many that were previously considered to be purely internal ones. In order to do so, the court must often apply a twisted logic that derives fundamental rights from a merely potential link with the exercise of free movement.
While a market citizenship logic of free movement generates substantial privileges for SCNs, their most important democratic citizenship right remains less secure than for FCNs. EU citizens residing in other member states enjoy voting rights in local and EP elections there but, with the exception of Irish citizens in the UK and British citizens in Ireland, they remain excluded from political representation in the national government of their host country. From the residential citizenship perspective, this is an oddity. One can hardly argue that the local franchise is necessary in order to prevent that SCNs suffer political disadvantage, but that being deprived of the much more important na-tional franchise is an acceptable restriction of their free movement rights.
Finally, EU citizenship generates another highly problematic distinction between mobile European SCNs and TCN migrants. The residential dimension of EU citizenship has imposed a special privilege of local voting rights for SCNs on often reluctant member states, such as Austria, France and Germany, all of which adhere to the constitutional idea of a unitary people consisting of identical members across all levels of the polity. This has led to a distinction between two classes of local citizens that is arbitrary from the perspective of local self-government. More generally, there are now two strongly contrasting approaches to the integration of migrants in the EU. For TCNs, member states and the EU itself promote active integration policies that combine sanctions and tests with affirmative measures, while for intra EU migrants a market citizenship logic dictates a laissez-faire approach, which assumes that unconstrained mobility and non-discrimination is all that is needed for social integration.
Possible scenarios for the rectification of the deficiencies of EU citizenship
Some of these problems could be addressed by weakening the derivative nature of EU citizenship and moving forward on the road towards a fully residential citizenship not only at the local, but also at the supranational level. Let me sketch briefly four possible steps on this road.
A first reform would introduce automatic acquisition of EU citizenship, but not member state nationality, by long-term resident TCNs. This proposal, which has been occasion-ally endorsed by migrant lobby organisations, MEPs and the Committee of the Regions as well as by some academic scholars, would create two classes of EU citizens: those for whom this status is derived from their nationality and those for whom it is instead derived from residence. While the reform would lead to more inclusion by providing long-term resident TCNs with local voting rights throughout the EU and all the other privileges of EU citizens, it could hardly overcome present concerns in member states about immigrant integration. Resolving these by removing them from the domestic agenda of member states can only breed further anti-EU resentment among the elec-torates there. Finally, the proposal would also remove the most powerful argument for opening access to national citizenship to all long-term resident immigrants. If these en-joy automatic access to EU citizenship, they will not only lack incentives for naturalisa-tion, but will also be perceived as having no substantive claim to full membership and political participation at the national level.
A second, and more radical proposal would address this latter problem for SCNs (and if it follows after the first step, also for TCNs) by abolishing any remaining distinctions between FCNs and SCNs and granting the latter a residence-based franchise in na-tional elections. This move would retain the exclusionary potential of nationality laws in regulating access to EU citizenship, but would effectively eliminate any traces of the derivative nature of EU citizenship with regard to its content of rights, leaving member state nationality behind as a hard but empty shell.
A third step would then respond to this outcome by abolishing birthright citizenship in member states and establishing it instead as the basic principle for determining EU citi-zenship. All those born in the territory of the EU (with possible conditions for prior pa-rental residence as in all current versions of national-level jus soli) and all those born to EU citizen parents outside the territory would automatically become citizens of the Un-ion and of all its member states. As a consequence, state level citizenship would have to be determined by residence. This move would effectively transform the EU into a federal state and downgrade the member states to provincial status.
Finally, we can imagine a utopian fourth step that would abolish birthright citizenship even at the level of the European supranational state and replace it with a uniform rule that in every polity all those and only those who are long-term residents will be counted as citizens. Political theorists have argued that birthright citizenship is a major source of violence between states or that it serves to maintain a globally unjust distribution of resources and opportunities). From this view, the three preceding proposals should be regarded as merely intermediary steps on the road to universal residence-based citizenship.
As my earlier discussion of the conditions for residential citizenship at the local level has made clear, I am not convinced by this project. Its third step, at which the current union would be replaced by a federal state, cannot be ruled out a priori. There may be future economic, political or military crises that are more dramatic than the current financial troubles and convince member states of the need for much deeper political integration. Yet such a possible response to a life-threatening challenge must not be confused with a hidden telos that supposedly pulls the EU towards becoming a federal state even in the absence of democratic support by its citizens.
The fourth scenario is, in my view, even more clearly a dystopian rather than a utopian one. It is hard to imagine how democratic political communities could be formed and maintained without assurances of intergenerational continuity provided by birthright membership. We can, however, not rule out this possibility on purely normative grounds. In a hypothetical world where most people are migrants living outside their countries of origin for most of their lives, maintaining birthright membership would amount to establishing a tyranny of sedentary minorities over mobile majorities. Current residence would then become the only justifiable basis for linking territorial jurisdictions to populations of citizens. I assume that in this scenario only minimal states could claim legitimate authority. Considerations of social justice that support public systems of edu-cation, health and welfare based on redistributive taxation would find little popular sup-port and democratic participation would be reduced to a small politically interested elite. I also assume that a need for belonging to associations with birthright membership will not vanish completely, but would then probably be articulated through the formation of non-territorial associations based on religion, class or ethnicity. What I cannot imagine is how democracy as we know it could survive such a radical disconnection between residence-based territorial jurisdictions and birthright-based non-territorial associations.
In our world, less than 4% of the global population are international migrants who reside for more than twelve months outside their country of birth; among the 500 million residents of the EU in 2010 just 4% are TCNs and 2.5% SCNs. In such a world, instead of dismantling territorial and intergenerational political communities with largely seden-tary populations for the sake of promoting geographic mobility, migrants must be ena-bled to integrate as equal citizens into these polities at all levels.
Modest reforms within the multi-layered system.
I conclude therefore that for the time being we should explore alternative ways of re-solving the deficiencies of EU citizenship. The starting point should be to accept it as a potentially coherent and normatively attractive constellation of three interconnected membership regimes: a birthright-based one at member state level, a residential one at local level, and a derivative regime with residence-based rights at the supranational level. This perspective suggests a couple of modest reforms.
The first among these would be to extend the local franchise to all residents in all member states. Instead of deriving local from national and European citizenship, the former would be finally based on its own distinct principle of inclusion, a principle that is already embraced by twelve member states and that is implicitly present in local de-mocracy in the other states as well. The main obstacle for this reform is the constitu-tional construction of a unitary demos across all levels within a state. The anachronistic character of this constitutional conception shows also in the fact that campaigns for a local franchise for third country nationals have been surprisingly resilient even in France, Germany and Austria, where reforms have been blocked by constitutional courts or councils.
The second reform would make sure that European citizens residing in other member states do not lose their representation at national level. This can be more easily achieved by introducing absentee ballots in those few member states that still have not done so (for example in Ireland and Greece) or by scrapping provisions in some other countries (such as the UK or Denmark) that withdraw voting rights after a certain period of residence abroad. Serious concerns in countries with large diasporas that a general right of external voting might impact too strongly on electoral results could be taken into account by limiting an absentee franchise to SCNs and excluding emigrants residing in third countries, or by reducing the weight of the external vote through counting it sepa-rately for specially reserved seats. There are reasons why external voting has recently become a global democratic standard and these reasons can be decisively reinforced through the imperative that free movement inside the EU must not lead to a loss of democratic representation at any level. A final argument for the external franchise solu-tion rather than the extension of national voting rights to SCNs in their country of resi-dence is that the former reform asserts the derivative nature of EU citizenship that the latter denies.
The third and most important reform would coordinate access to EU and national citizenship through some common basic standards for jus soli and jus sanguinis, for natu-ralisation, renunciation and withdrawal. Letting the ECJ expand the scope of EU citi-zenship rights while denying the EU any competence to harmonise member state poli-cies with regard to citizenship status will undermine the legitimacy of the Court, is likely to create conflicts between states that suspect each other of undermining their immi-gration control powers, and leaves radically incomplete the EU agendas of harmonising integration policies towards TCNs and promoting the political participation of SCNs in their host countries. None of these reforms would challenge the derivative nature of EU citizenship or the importance of birthright membership in the states that have after all created the Euro-pean Union. They would instead make explicit the as yet underdeveloped multilevel structure of citizenship in the European polity.
(1) This is a relatively recent development. In late 19th century Austria and Germany, birthright citizenship in municipalities (“Heimatrecht”) was used to restrict internal migration by denying poverty relief and access to local public services to citizens residing outside their municipality of birth. Switzerland’s “Bürgergemeinden”, in which membership is acquired at birth, are historical remnants of this system. Today’s “hukou system” in the Peoples’ Republic of China is an extreme case of local birthright citizenship as an instrument of exclusion from social welfare. It is based on jus sanguinis so that rural hukou status is even inherited by second generations of migrant origins born in the cities.
An earlier version of this essay was published in Richard Bellamy and Ulla Steiger (eds.) (2011): EU Citizenship and the Market. University College London, The European Institute: 6-1.
Rainer Bauböck ist Professor für soziale und politische Theorie am Europäischen Hochschulinstitut (EUI) in Florenz.