1) Denationalization protects democracy

One defense of the right to revoke hooks it to the importance of physical safety in democratic states. A defining feature of a legitimate and sovereign state is its willingness and ability to protect the safety of all citizens; this understanding of a sovereign state's key obligation goes back at least to Thomas Hobbes’ explanation for why individuals cede authority to a Leviathan, that is, why they trade physical insecurity for the protection he offers. If this is correct, then an individual who by her actions aims to threaten this physical safety, by engaging in crimes that threaten a state's security must be stopped and also prevented from doing so again in the future; one way to ensure this is to denationalize her.

This is more generally the claim made in contemporary discourse on revocation: the state has one central responsibility, to protect its national security, one aspect of which is to protect its citizens from grievous harms, including those caused by terrorism. Many political actors propose that terrorism poses unique threats to the physical safety of citizens. They tell us, we are in a state of “exception” or “emergency,” where civil rights can be sacrificed in exchange for ensuring security (for a discussion, see the contributions to Ramraj et al. Reference Ramraj, Hor, Roach and Williams2005). Or, put differently, under conditions of exceptional risks posed to citizens, states can reasonably claim that their sovereignty rights should trump the rights of citizens (Davis and Silver Reference Davis and Silver2004; Joyner Reference Joyner2004; Golder and Williams Reference Golder and Williams2006). This is precisely what UK Prime Minister Theresa May said in the wake of recent terrorist attacks, when she declared that she would not permit human rights legislation to restrain her terrorism-fighting strategies: “if human rights laws stop us from doing it, we will change those laws so we can do it” (Mason and Dodd 2017). Correspondingly, under these exceptional conditions, it is reasonable to accept that normally unjust punishments can be justifiably deployed.

Even granting the possibility that there is something uniquely dangerous about the terrorist threats many democratic states face, in the case of punishments as severe as denationalization, however, it seems reasonable to demand evidence of its contribution to protecting safety as a condition of its use. Yet, there is no clear evidence that this particular power serves to make democratic states safer (Lenard Reference Lenard2016a). Citizens and outsiders perpetuate acts of terrorism in states that have long had revocation laws on the books, in states that have recently passed revocation laws, and in states that have rejected revocation laws. More importantly, every democratic state possesses a robust criminal justice system through which crimes against the state can be effectively prosecuted. No defender of the right to revoke has provided reasons for believing that the existing suite of criminal powers in democratic states is inadequate to deal with the kinds of crimes that might, otherwise, trigger the revocation of citizenship. In particular, democratic states regularly incarcerate dangerous offenders, and while there are certainly many difficult moral questions to be asked about the appropriate use of jails and prisons, and the conditions under which inmates live, there is no clear reason to believe that the physical safety of citizens is protected better by expelling wrongdoers than it is by incarcerating them.

A variation on this protective argument draws heavily on the Ancient Athenian practice of ostracism I described earlier (Forsdyke Reference Forsdyke2009). This practice, which appears to have much in common with contemporary revocation laws, permitted Athenians to vote to expel citizens who had gained power—by amassing wealth for example or popularity—of a kind that threatened to undermine democratic practice in Athens. The exclusion from Athens was temporary—for ten years only—and was not accompanied by a loss of property or citizenship; after ten years, ostracized individuals were permitted to return to their untouched property, and the rights that citizenship in Athens entailed, but (ideally) with much reduced influence over democratic political life.

The thought that individuals threaten to undermine the practice of democracy in contemporary democracies is uncompelling, however. Whereas the foundations of democracy in Ancient Athens may have been unstable, the same is not the case in contemporary democratic states. The practice of ostracism may have served to protect Athenian democratic institutions, as David Miller notes in a recent article (Miller Reference Miller2016); although it is surely right, as Miller claims, that the stability of democratic states should not be taken for granted, it is hard to believe that the power to denationalize individuals is in any significant way related to the ability of contemporary democracies to protect their ability to function. The dis-analogy continues: those who were targeted by ostracism in Ancient Athens were the most powerful, whereas those who are at risk of being targeted by revocation laws are presently among the least powerful. It should not escape notice that those most likely to be targeted by revocation proceedings are disproportionately Muslim citizens; rather than offer protection to citizens in democratic states, there is considerable risk that such laws will instead be used to further victimize already vulnerable citizens.Footnote ¹⁴ The claim that the practice of democracy requires expelling individuals who will otherwise undermine its foundations is implausible.

A slightly different way of defending revocation suggests its purpose lies in its expressive role (Hampton Reference Hampton and Cragg1992). A democratic state typically relies on the fact that individuals have internalized its norms, and thus do not require costly and physical coercion in order to abide by them. One way that a democratic state communicates these norms is by delineating a range of unacceptable behaviors and their associated punishments. From the point of view of an expressivist theory of punishment, revocation may appear to be, logically and symbolically, the right punishment for certain categories of crimes, in particular, crimes which states understand as an attack on their foundations. This kind of defense does not risk being rejected for inadequate evidence that it is in fact the case that the crime undermines democratic stability. Rather, it is simply that any individual who has by her actions attempted to rob her fellow citizens of the benefits of living under democratic rule, by attempting to undermine the foundations for the effective function of democratic institutions, deserves to have their status rescinded, that is, to be denied the very rights and benefits she has attempted to deny to others. Ultimately, denationalization is a way of reasserting the importance of democratic values, and is thereby a fitting response to certain crimes.

But we should resist this sort of fittingness argument for several reasons. First, the practice of incarcerating those who harm, or aim to undermine, democratic institutions, appears to be fitting in much the same way as denationalization. In both cases, the purpose is to separate the criminal from society and thereby to reduce her access to a large list of basic freedoms that law-abiding individuals can take for granted, and in so doing, to signal that she is to be treated as a kind of partial or conditional citizen (Vaughan Reference Vaughan2000). To the extent that it makes sense to treat criminals as individuals who have violated the terms of the social contract that ought to bind citizens in a law-abiding state, the criminal is forced by both incarceration and denationalization to confront her failure to have lived up to the terms of the agreement.

Yet, there are reasons to believe that incarceration rather than denationalization is a more appropriate punishment, in the sense of more consistent with a range of democratic values. First, incarceration permits the harmed democratic community the knowledge that it is issuing punishment and ensuring that it is exacted. The state that has been harmed returns the harm, as it were, by constraining a criminal's freedom, perhaps forever. The criminal's ability to inflict additional harm is removed (they are prevented from perpetrating future harm, at least during the period of incarceration). The epistemic advantage of incarceration over denationalization is considerable; while democratic states can incarcerate wrongdoers, and thereby know with certainty that the appropriate punishment has been carried out and that the wrongdoer is prevented from carrying out further harm, denationalization can carry with it a range of uncertainties, including whether the state to which the criminal is subsequently deported will prove willing and able to ensure that appropriate punishment (or surveillance) is carried out, so that she cannot commit further harm.

Second, it is worth noting that among theorists of punishment, there is considerable disagreement about the purpose of punishing wrongdoers, and to the extent that finding the fitting punishment as I described above is an objective, it is only one of many. Another is articulated in a recent article by Elizabeth Cohen, in which she argues that part of what makes a punishment appropriate for democracies is that those who are punished are treated as capable of moral reform, and therefore as people who can return to the ranks of full and participating citizens in their own state (Cohen Reference Cohen2016). Denationalization denies opportunities for reform and is therefore an unacceptable punishment in a democratic state.

Third, there is a widespread belief that criminals should face trials, and be punished, in the communities where they have committed the crime. The general reasoning here is that a political community has a right to exact punishment on individuals who have done it harm; the crime is considered to be against a particular community (Duff Reference Duff2001). This thought runs through discussions of the appropriate locations for trials, domestically and internationally. Domestically, it is common practice to hold the trial in the jurisdiction in which a crime was committed; where there are deviations from this practice, and they are rare, it is normally because there is some belief that a local trial will not be fair. The argument for holding the trial—and subsequent punishment—in the jurisdiction in which the crime was committed is that the harm was done to a particular community, which is entitled to witness the pursuit of justice on its own behalf. This in part also explains the general view that when an individual engages in a criminal act outside of her territory of citizenship, she should face punishment in that jurisdiction—she should face those she has harmed—rather than her own jurisdiction.Footnote ¹⁵ The underlying logic is that the relations requiring reparation are between a community and a criminal who harmed a particular community. Note that the language here is of reparation, that is, it is language that assumes, as above, that rehabilitation of the criminal into the status of a law-respecting citizen is the goal, or at least a genuine possibility. Denationalization denies the criminal and the community she has harmed the opportunity for all stages of reparation.

One possible response to this set of objections to denationalization is this: the account just above assumed that one or both parties wants the relation to be repaired. But that may be true, in fact, of neither party. What if the criminal prefers denationalization over incarceration, or possibly even intends to sever connections with her state of citizenship by committing grievous crimes? Would that preference for denationalization justify it? One might think that the preferences of the criminal ought not to be respected since, after all, she is a criminal. But this is mistaken. For example, it is permissible, and perhaps even morally required, to consider the preference a criminal has for the location of her incarceration; for example, she may prefer to be incarcerated as close to her family as possible, and this can be justly considered a relevant preference. And yet even if one believes that criminals’ preferences for the conditions of her incarceration should be respected in some cases, it seems reasonable to deprioritize (at least some of) them where the crimes are grievous and aimed at undermining the foundations of democracy. In particular, it seems morally relevant that the reason for the preference is, at least in part, to trade a more burdensome punishment (in this case, incarceration) in favor of one that is less burdensome (possible freedom, elsewhere). Any criminal can presumably be expected to prefer a lesser sentence. But that preference appears morally unrelated to the question of what is a fair and just sentence; it is therefore, when offered as part of an apparent reduction in sentence, to be understood as coerced denationalization (as in the case of Hamdi, which I noted earlier), and thereby unjustifiable because the conditions of consent are not present.

Perhaps though the wrongdoer aims to signal—exactly as the defenders of revocation posit—that she is no longer loyal to her state of citizenship, that she no longer respects the norms and values that underpin the institutions that govern the relevant political community and indeed, that she means by her actions to sever the relations between herself and her co-citizens. Citizens are already permitted to sever citizenship bonds on a voluntary basis, if they so choose.Footnote ¹⁶ Wouldn't it then be consistent to argue that a wrongdoer's preference for denationalization is adequate to justify it, since it should count as an instance of the right that citizens possess, as a matter of international law, to renounce citizenship?

Although there is considerable plausibility to this view, I do not believe so. Here, the danger of a slippery slope is too high: accepting that some actions count as evidence of voluntary renunciation puts judges (or other similar figures) in the position of interpreting behaviors as evidence of attitudes in ways that lend themselves towards dangerous overreach. This view—that there is potential for dangerous overreach—finds expression in a key American Supreme Court ruling (Afroyim v Rusk, 387 U.S. 253 (1967)). In this decision, which itself overruled decades of jurisprudence that had treated certain actions by citizens as equivalent to their renouncing citizenship, the Court argued that the state could not reliably interpret certain behaviors, by citizens, as evidence of a desire to renounce citizenship. As a result, the revocation of citizenship in the United States was rendered unconstitutional (except in cases of fraud) (Weil Reference Weil2012, 175; Spiro Reference Spiro2014). The only act that can and should be interpreted as a desire to renounce citizenship is a formal renunciation, made under conditions of consent.

What if the denationalizing community would prefer not to repair relations with the wrongdoer? Aren't these preferences morally relevant, given that they are those expressed by the democratic majority, which is the aggrieved party? Again, in my view, they are not. It is worth repeating the democratic political community's obligation to treat a wrongdoer as though she can be rehabilitated. The political community is bound by a public culture that is or should be defined by a set of shared norms and values. The deliberations that determine the content of these norms and values must be open to all citizens if, indeed, the culture is to be understood as adequately public. The ways in which these norms and values are imparted to citizens are multiple; some formal and others less so. But it is fair to say, in broad terms, that one responsibility a political community—that is, a political community that is bound by a shared public culture—possesses is the responsibility to ensure that the shared norms and values are imparted to all members of the community, coupled with the requirement that the venue in which its content is deliberated is adequately open and inclusive.Footnote ¹⁷ Where it fails to impart these values and norms adequately, it must take responsibility for this failure. When citizens act in ways that are taken to be demonstrating disloyalty or as failure to have internalized the shared norms and values of the community, and who thereby act to undermine the foundations on which the political community is based, the political community may properly be understood as having failed its citizens, both those who are harmed as well as the criminal herself.Footnote ¹⁸ This responsibility requires taking charge of punishing the wrongdoer, within the boundaries in which the crime was committed; denationalization amounts to an unwillingness to take responsibility for its failure to successfully educate its citizens and for punishing them appropriately.

One might propose that a state does not possess this responsibility in the case of immigrants, in which their formative years were spent elsewhere (Miller Reference Miller2016). But, that proposal is mistaken—as in the case of more generally imparting norms and values to citizens born on a territory bound by a particular public culture, the process of immigration and naturalization to citizenship places a responsibility on the state for educating newcomers to citizenship, that is, into this public culture. To the extent that the relation between immigrants and the state they join is properly understood as a contract, the state takes on responsibility for providing the conditions under which the successful integration of immigrants—to the social, economic, and political communities they have joined—is possible and, indeed, likely. Where it fails to impart these norms and values, it again must take responsibility for its failures,Footnote ¹⁹ in the form of punishing citizens in such a way that they retain—perhaps in the far distant future—the opportunity to repair the relations they have damaged, and to return to the political community as full and equal members of society.Footnote ²⁰

2) Dual nationality is a privilege

In the discussion above about whether denationalization is compatible with democratic principle, I made only brief mention of a key element of many contemporary discussions of denationalization, which I will now highlight. This is that, in many states that permit revocation, it is permitted only if individuals would not thereby become stateless; as a result, only dual nationals are affected by revocation of citizenship laws. So long as the revocation law is written and implemented so as to make it impossible to render an individual stateless, say some of its defenders, revocation can permissibly be imposed on dual nationals who have committed or intend to commit grievous crimes (de Groot and Vink Reference de Groot and Vink2010). This argument is pressed in a variety of ways, and I shall refute them all.

To begin, it is worth noting that all of the following justifications for denationalization in cases of dual citizenship rely explicitly or implicitly on the notion that there is an unfairness of some kind generated by the fact of dual nationality (Herzog Reference Herzog and Wejner2010). One might simply deny that there is something unfair about dual nationality in the first place. Nationality may simply belong on the list of so-called morally arbitrary factors about an individual, including race, sex, gender, religion and so on, that should not have any impact on how one is treated by the social and political institutions that govern a state. This would be consistent with global trends towards recognizing the legitimacy of dual citizenship, a shift which generally reflects an abandonment of the view that loyalty to a state is indivisible and that, correspondingly, dual citizens cannot be trusted to demonstrate adequate loyalty (Bauböck Reference Bauböck2007; Bloemraad Reference Bloemraad2004). An accompanying thought is that, even if there are cases where dual nationals are significantly more loyal to one state than another, there are enough dual nationals for whom their status in both states is meaningful; correspondingly, even in cases where dual nationals generally reside in one state (as inevitably they will), it is a mistake to infer that their connection to a second state is insignificant. If these underpinning thoughts are genuine, they rule out treating the citizenship of dual citizens differently, by permitting its revocation as punishment for, or to deter, crimes. Defending revocation policies requires asserting that the citizenship of dual citizens is less robust than that of single-nationality citizens. One might therefore reject any claim that dual nationality is unfair, by responding that any differential treatment of the citizenship status of some individuals is a violation of the equality that citizenship is intended to protect in democratic states.

But, since some believe that dual nationality is a source of inequality, it is worth fleshing out the ways in which the objection—that it is acceptable to subject only dual nationals to the threat of denationalization—is presented, so that I can be more specific about why these attempted defenses of revocation ultimately fail.

One thought is that the possession of two nationalities is a kind of privilege that confers a significant advantage, which violates the foundational equality on which democratic states rely. It is not that dual nationality is impossible to come by—there are many cases where individuals choose to migrate and then to naturalize into a second nationality, which though time-consuming is within the realm of possibility. But, doing so is cumbersome enough that those who already possess these two nationalities can legitimately be understood to be advantaged. These advantages mainly consist in a larger pool of convenient travel, work, educational and residence opportunities; as a result, dual nationality violates equality and making dual nationals liable to denationalization is a permissible way in which to reestablish equality between single and dual nationality citizens.Footnote ²¹

It is, of course, a basic principle of egalitarian political theory that, where differences among citizens generate inequalities, they should be remedied as far as possible. Much hinges on what the relevant metric of inequality is thought to be—if the focus is on material inequalities, then it is typical to permit and encourage the redistribution of wealth. If the focus is on welfare inequality, then emphasis is given to protecting the equal access to welfare for all citizens. If the focus is on political equality, then emphasis is given to eliminating inequalities that threaten to disrupt citizens' ideally equal capacity to influence political decision making. The point is that where we have identified a preferred metric (or metrics) of equality in a political state, benefits (and burdens) are unjust where they disrupt that form of equality, and compensation can legitimately be demanded of those who benefit by those who are burdened. So, one question we might reasonably ask is whether the possession, by some citizens, of dual nationality is such that it disrupts a relevant form of equality.

One preliminary challenge that must be faced in responding to this question is how to measure the advantages of possessing multiple citizenships. If one believed that dual nationality was an inequality that demanded compensation because of the unearned benefits it offers to those who hold it, one would also have to acknowledge that some citizenships, as a matter of practice, offer greater benefits than do others. At present, US citizens can travel to 174 countries without procuring additional visas; Somali citizens can travel to only 31. Similarly, the number of opportunities that accompany living in the United States, and the quality of rights protection, are significantly greater than those that accompany living in Somalia. Thus, a state considering the value of the benefits offered by a particular second citizenship will surely have to consider that the benefits of some second citizenships are insignificant compared to the benefits of others. Correspondingly, revoking some citizenships is costlier than revoking others.

Complicating this accounting further is the question of whether a person should be penalized for having a second citizenship when the second country does not permit renunciation of citizenship. Many defenders of denationalization propose that such individuals should be exempt from denationalization policies. This willingness to exempt stems from the thought that those who would genuinely like to protect themselves from the risk of expulsion can do so “easily,” by renouncing the second citizenship. If this severance is impossible, then we cannot hold dual citizens responsible for their failure to hold only a single citizenship status. Yet, if one truly believes that benefits—even those that are unasked for, and which one cannot refuse—require compensation, one must also believe that the benefits associated with an unrenounceable second citizenship require compensation from its holder. Even if it cannot be abandoned, a second citizenship will still provide at least some greater opportunities to travel and reside abroad, and these require compensation in some form. Finally, any state engaging in the process of developing a schema to articulate the precise compensation for holding a second citizenship will have to confront that some states also impose costs on their citizens, regardless of whether they are residents. These costs include, mainly, national service requirements and tax burdens, and surely would have to be balanced against the residence and travel benefits provided by the second citizenship.

Let's say, though, that all of these measurement and balancing challenges were successfully meetable and met. What, then, is reasonable compensation for the possession of dual citizenship and its attendant benefits? Is liability to denationalization a reasonable compensation? To answer this question, it is worth noticing two things. First, to the extent that dual nationality does confer genuine advantages, these are advantages that do not in most cases bear directly on the internal material and political relations of a state.Footnote ²² As citizens interact with each other in political and social environments, whether some possess an additional citizenship is generally irrelevant.

Second and more importantly, recall that the essential protection offered by citizenship status is residential security. Generally, we can see that even where we believe that benefits for some small proportion of citizens generate an inequality, and where we believe that compensation is thereby due to fellow members of the state, this compensation must be reasonable in the sense that it returns the internal state of affairs to equality as much as possible. However, since as I said above, the benefits (though real) are minimal from the perspective of internal, political relations, it is not legitimate to demand that dual nationals compensate in the form of accepting weakened protection for a basic right. Notice, furthermore, that one possible response—dual-nationals can escape denationalization simply by refraining from engaging in heinous crimes—is unacceptable; it would be similarly unacceptable to adopt a law imposing greater criminal penalties on residents of the east side of a city than those on the west side, with the justification that east-siders can avoid these penalties by refraining from crime. The mere fact of having one subgroup liable to denationalization undermines residential security.

A defender of revocation might insist, nevertheless, that the small chance that dual nationals have of being subject to denationalization is justifiable for this reason: even if we make their status very slightly less secure in one state, they retain citizenship status elsewhere. Dual citizens are not at risk of statelessness, which is the grievous harm that must be avoided at all costs. Thus, we might believe, the reduction in security offered to dual citizens in state X is more than compensated for by the additional security provided by their citizenship in state Y. Although this argument satisfies the requirements imposed by the international dimension of citizenship, it falls afoul of the domestic dimension of citizenship, namely, its grounding in the strong interests citizens have in residential security here.

To remind, residential security permits the confident expectation that one can continue to live here, and make decisions about how one's life should go. The ways in which we interact—with political institutions, with a cultural environment, with particular people including family members and friends—are themselves shaped by and dependent on being present, here, in a particular political and cultural environment. Autonomous lives are lived by individuals in particular cultures (Kymlicka Reference Kymlicka1995). Some political theorists deny the claim that any particular individual has a right to live their lives in a particular and preferred cultural environment. To the extent that individuals require a cultural environment in which to live autonomously, they are entitled access to a cultural environment that can offer the prerequisites for autonomous living. But individuals may not demand support to sustain a particular and preferred culture, nor may they demand the right to join a culture—citing a cultural preference—whose members do not wish to include them. The right to cultural or national self-determination, or the slightly weaker right to access one's own political or public culture, is thereby denied on the grounds that while we all have a right to self-determination, which is often expressed in political forums, and while we all have a right to live in the context of a flourishing culture, we do not necessarily possess the right to do so in our preferred political culture.

The same objection could well be raised here, that is, that whereas individuals have a right to citizenship they do not have a particular right to citizenship in the state of their preference. This argument is what those who would permit the expulsion of dual citizens are, in effect, suggesting; no one can be denationalized if they would thereby become stateless, but dual citizens are not faced with statelessness, simply with being required to reside in, and accept state protection of their rights from, their state of second citizenship, even while preferring to reside in their state of first citizenship. The objection might continue, their (domestic) right to citizenship is not violated, because although they are not necessarily entitled residential security in the state of their preference, they are nevertheless guaranteed residential security in a (human rights respecting) state.

Expressed in one way, this position is defensible. A citizen of Norway who simply desires to migrate to Canada and adopt Canadian citizenship, because she prefers Tim Horton's coffee and hockey (the central aspects of Canadian culture), is not (necessarily) entitled to do so simply because that is what she prefers. Typically, although there are certainly important exceptions among open border advocates (Carens Reference Carens1987; Cole Reference Cole2000), many scholars agree that states may, under some conditions, legitimately exclude individuals who are not yet on their territory, and who do not yet hold citizenship in that state. In the case at issue here, however, an individual already possesses the relevant citizenship (i.e., they are not requesting access to residence, they are already entitled to it), and those who advocate the possible denationalization of citizens deny that such an individual—even though she possesses citizenship in the state in question—is entitled to keep that citizenship, and correspondingly to have her residential security protected. That dual citizens have an in-principle access residential security elsewhere is inadequate—dual citizens are entitled to residential security in the state in which they reside, where they have family and friends, and to whom they owe redress in cases of violations (including grievous ones) of shared norms and laws.

What about the state in which they do not normally reside? Can that second citizenship be justifiably revoked in these cases? No. Here let me first remind readers of my earlier observation about the nature of dual citizenship and its growing acceptance. This acceptance depends on committing to the view that a dual citizen can plausibly be loyal to two states at the same time.

One might then press the point, further, with this scenario: imagine a dual citizen, who was born in state X, who has resided all her life in state X, whose second citizenship (in Y) is an inheritance from a parent. She has not visited state Y; is not familiar with its language or culture; and has no friends or family there. Now, imagine that this individual either commits an (allegedly) denationalization-worthy crime in X, or that she crosses very briefly into Y and commits an (allegedly) denationalization-worthy crime there, before returning to X. Surely, a critic might propose, in this case, there is nothing problematic about state Y revoking her citizenship; the relevant dual citizen, by any measure, is simply not connected to state Y, and so denationalization from this second citizenship would constitute neither a great harm nor a significant erosion of the personal relations that matter to them, in particular, it does not appear to undermine her security of residence in a problematic way. The critic might even concede that it would be better if state Y chose against denationalization (perhaps simply on the grounds that it is better if states stay out of the business of assessing the inner attitudes of citizens), but propose that it is nevertheless permissible for a state to denationalize in these admittedly limited and unlikely circumstances.

I understand why one might accept the permissibility of revoking citizenship in these limited kinds of cases, in which security of residence appears not to be threatened by revocation. It should be clear, however, that I aim to resist this conclusion. Rather, as I have argued, citizenship status should be treated as permanent, that is, irrevocable, even in the kinds of egregious cases that have motivated recent attempts to reinvigorate or adopt denationalization policies. This commitment does not translate, however, into believing that all the rights of citizenship are protected for nonresident dual citizens.Footnote ²³ In other words, it is reasonable for states in which citizens are nonresident to treat some subset of rights as protected only for residents of that state. Protecting a subset of rights for resident citizens is a practice already in place in many states. Here is one familiar example: citizens who reside outside of a country for extended periods of time are normally not permitted to access the full complement of social services generally available to residents in the state in which they do not reside. Similarly, in many states, only resident citizens are permitted to vote. Citizens who reside outside of the country for extended periods of time can be denied the right to vote, while they are residing abroad. In both cases—of voting and social service provision—citizens retain the right to return to reside in their country of citizenship, and then to requalify for the associated rights and privileges once they have remet the residence requirements. For these policies to be justified, the residency period must be short, but it is reasonable to impose one nevertheless.

The point of this discussion is simply to observe that my defense of citizenship as permanently irrevocable does not demand that states retain the obligation to protect the full set of rights and privileges of citizenship for nonresident citizens. It is reasonable to require nonresident citizens, who return from abroad to reside, to fulfill some obligations before their full set of citizenship rights are returned to them. By their nature, dual-nationals are nonresident citizens of at least one state. Citizenship for nonresident citizens protects their right to return to reside, in the confident expectation that once they have met certain residency requirements, they can expect the full complement of citizenship rights to be protected. This view follows from how I have understood the right to citizenship in the first place, as protecting the strong interests that individuals have in residential security, even (as in this case) for citizens who are not presently residing on the territory in question.

What does this general view mean for the difficult case, in which a dual citizen with significant ties to country X but not Y commits a grievous crime? It means that normal criminal procedures should be followed in whichever jurisdiction the crime was committed. If the states agree, a prison sentence can be served in X even if the crime was in Y, as occurs in some criminal cases. But there is no implication for the citizenship rights of the perpetrator. If the normal procedures of Y are to deny nonresident citizens rights such as voting, they may do so in this case, but the crime is not relevant to that decision. When the criminal justice procedure has run its course, the dual-citizen should retain her rights to residence in both states; no state has the right to rescind citizenship.