Migrants by plane and migrants by stork: can we refuse citizenship to one, but not the other?

ABSTRACT States combine the routine refusal of citizenship to migrants with policies that grant newborns of citizens (or residents) full membership of society without questions asked. This paper asks what, if anything, can justify this differential treatment of the two types of newcomers. It explores arguments for differential treatment based on the differential environmental impact, different impact on the (political) culture of the society in question and differences between the positions of the newcomers themselves. I conclude that, although some justification for differential treatment exists, the case for it is weaker than one may expect and the grounds on which it can be justified are surprising and problematic.


Introduction
Every year the Netherlands admits roughly 170,000 citizens without questions asked. France about 750,000 (Eurostat 2020) and Japan (with strict entry policies) roughly 840,000 (Ministry of Internal Affairs and Communications Japan 2021). These newcomers are not confronted with borders or immigration services, many are festively welcomed. Two of these countries explicitly prefer more such newcomers (UN 2013). These numbers are not evidence of a generous attitude to refugees, but the number of children born into citizenship. Far fewer outsiders received citizenship: roughly 35,000; 110,000 (Eurostat 2019) and 8,000 (OECD 2016) respectively. Policies are made to discourage them from applying for citizenship, fences and walls physically prevent them from entering. Paula Casal puts it nicely: I will take a statist and not a cosmopolitan theory of justice as a starting point, i.e. I assume that the scope of the full demands of justice is limited to the state. 2 Even if we take the statist view, differential treatment is hard to justify. Statists can appeal to indirect justifications 3 for the right to exclude (like cosmopolitans) and appeal to arguments that rely on special obligations to compatriots (unlike cosmopolitans). If statists cannot defend differential treatment, neither can cosmopolitans.
In which country one receives citizenship at birth has a profound effect on one's chances in life. The combination of birthright citizenship and restrictions on changing citizenship imposes more costs on some than on others. For a disadvantaged migrant requesting Canadian citizenship, a lot depends on the reply. If Canada refuses, this person is owed a justification. Unlike a tennis club with over-used tennis courts, Canada cannot say that it is not accepting new members. It is: newborns. Differential treatment also raises a theoretical problem. If statists and liberal nationalists cannot provide a justification for differential treatment, they face a dilemma. They either accept that states may exclude both newborns and migrants, or that states must be open to both. As Ferracioli points out, the former is highly counter-intuitive, and the second weakens statism to such a degree that it 'might no longer be worth having ' (20,182,867).
The aim of this paper is twofold. First, theoretically, the paper aims to find out how much room statists have for defending differential treatment. I argue that the defences that have emerged in the literature fall short, and look at whether other arguments for either the inclusion of newborns or the exclusion of migrants identify a morally relevant difference between them. Second, practically, I want to ask whether states can consistently refuse migrants access to citizenship while at the same time welcoming newcomers by stork. Do actual states -with these statist arguments in hand -have a consistent, good faith, justification to offer to those migrants they turn away?
A few terminological clarifications. 'Citizens' refers to full members of a society. A 'prospective citizen' is someone who does not have citizenship of a particular state, but wants to acquire it. This paper focuses on citizenship, so 'Migration' refers not to a person's physical move across borders, but to the act of acquiring citizenship. Newborns are all children whose parents are citizens of the society in question (not children born to foreigners) and who will acquire citizenship based on the nationality (or location) of their parents. 4 Although conceptually distinct, I do not distinguish between a right to permanent residency and citizenship. There seems to be wide consensus amongst opponents and defenders of the right to exclude that it is not permissible to have permanent non-citizen residents. 5 By granting this, the bar for statists to show that differential treatment is justified is lowered. After all, all they need to show is that newborns are owed permanent residency, and the right to citizenship will follow.
The ambition of this paper is limited. It does not argue for or against the right to exclude, or against granting children citizenship at birth, but just asks what -if anything -the morally relevant difference that can justify differential treatment between migrants and newborns is.
The paper starts at literal beginning: procreation. Migrants exist, children are brought into existence. And they are brought into existence by current citizens. Second, I move to parenthood: children are -usually born into parent-child relations. The small literature on newborns and migrants has so far focused on the nexus of rights and duties that people have as parents. These arguments, I argue, cannot be the full story. The rest of the paper focuses on arguments for the right to exclude migrants instead, primarily but not solely as defended by David Miller. I ask whether states lack the right to exclude on similar grounds when it comes to newborns. I conclude that the case for differential treatment is surprisingly weak and has some implications that I suspect statists will find hard to stomach, too. And the degree to which actual states can appeal to these arguments is limited. 6

From procreation to citizenship
There are two distinct moments at which the state can attempt to prevent migrants by stork entering society. First, it could design policies to prevent children from coming into existence. The second moment at which the state can prevent someone from acquiring citizenship is once she exists. Let me turn to potential and actual newborns in turn to explore whether the differences between them and migrants might ground a justification for differential treatment.

Migrants and possible people
Migrants exist, but the government can implement policies to discourage the creation of newborns. What does this striking difference do for differential treatment? One question is what are the stakes in coming into existence. Although some argue that there is an existential benefit (Holtug 2001;Arrhenius and Rabinowitz 2010), let us assume that it is not bad for newborns not to come into existence. 7 Migrants, on the other hand, exist when they apply for citizenship of a society: they have an (often significant) interest in acquiring citizenship. A clear disanalogy presents itself. But it points away from differential treatment as defined above, which favours the inclusion of newborns. The child's interest in coming into existence is much weaker, and hence much less 5 E.g. because the inequality between citizens and non-citizen residents is unjust (Walzer 1983, 58-60;Miller 2008a). 6 Other possible arguments for differential treatment could be grounded in the freedom of association view, see Brezger and Cassee (2016) and Bou-Habib (2019). 7 Or at least not bad in a way that makes it wrong not to create them. Narveson famously, expresses the sentiment that it is not bad clearly: 'we are in favor of making people happy, but are neutral about making happy people ' (1976, 73).
likely to override the interest of the migrant to acquire citizenship. The merely potential person may lose a whole life, but it is not really a loss to them. One may interject that we do not wrong newborns but prospective procreators by preventing possible people to join our ranks. The means one would have to use to prevent children from coming into existence (also Laegaard 2013, 660-661) are arguably more problematic than the means used to prevent people from acquiring citizenship. First, because the former means constrain citizens (whose interests have greater weight on statist accounts) while the latter applies to foreigners. Second, the means to prevent procreation are arguably more intrusive. Michael Blake argues in favour of the right to exclude migrants based on the right not to acquire responsibility for the protection of others' rights. Tellingly, Blake gives procreation as an example of where we do acquire responsibility for others without consent: the rights of my friends and colleagues to control their own bodies is more central than my right to avoid unwanted obligations towards their children, and this means in practice that any attempt to prevent those children from coming into the world would be morally impermissible (2013, 119).
Brezger and Cassee offer a plausible reformulation of his claim applied to migration: 'Blake holds that the right to procreate, which is justified by reference to individual selfdetermination, is more important than the right to avoid imposed obligations ' (2016, 375). We could rephrase Blake's claim in terms of a disanalogy argument about citizenship: Permissibility of prevention disanalogy argument: the moral costs of preventing children from coming into existence are much higher than the moral costs of preventing migrants to enter.
There are several worries about this argument. The moral costs for preventing migration are higher under some (real-world) circumstances. Think, for example, of the thousands of people who die trying to cross the Mediterranean every year. Moreover, some ways of influencing how many people come into existence are unproblematic, for example improving the position of girls and women leads to lower fertility levels: limiting how many newcomers arrive by stork, like limiting those by plane, can come at higher or lower costs. 8 Blake agrees that sometimes the cost of exclusion from a territory is too high, and agrees that refugees for example cannot legitimately be forcefully excluded from the territory if their rights elsewhere are threatened. 9 However, even if exclusion from the territory is not permissible, this does not mean that granting citizenship is required. Granting citizenship (or a right to remain indefinitely) is not always a part of what states owe to those within their borders. Likewise for newborns. Even if limiting procreation is impermissible (e.g. Heyward 2012), it does not follow 8 As a referee points out, by influencing the determinants of fertility one does not really exclude anyone in a straightforward way from citizenship. In the migration case the equivalent would be to address the underlying causes of migration (e.g. tackle poverty). This limits the number of migrants without exclusion. My point here is that migration and procreation are not that dissimilar in that regard: both preventing newborns from coming into existence and migrants from entering our territory may come at low costs (by tackling determinants for example) or high costs (use of force, limiting (bodily) autonomy). 9 I thank a referee for insisting on this. See Blake 2013: 127. that the resulting children should receive citizenship or permanent residency. This requires an extra step. Possible arguments for granting citizenship to newborns created by citizens as a basis for differential treatment is what we will turn to next.

Causation and responsibility
Some argue that procreators are responsible for their offspring because they brought them their existence. This is an argument for procreative responsibility. An argument for birthright citizenship needs extra steps: P4.States have an obligation to enable people to act on obligation C1.
P5.Shared citizenship between parents and children is crucial for the fulfilment of obligation C1.
C2.States have an obligation to provide citizenship to citizen's offspring.
One could deny that an equivalent of P1 or P2 applies to migrants. I will focus on P2 here and turn to P1 in the next section.
Dependency based disanalogy argument: the dependent position of newborns is the consequence of decisions made by current citizens of the state, whereas the dependent position of migrants is not -therefore, newborns are owed citizenship and migrants are not.
In so far as the causes of poverty are domestic, destination societies are not causally responsible for the vulnerability of migrants. Hence, it has no corresponding obligation to alleviate the vulnerable position of migrants.
The claim that destination countries are not causally responsible at all for the dependent position of migrants is empirically false in many cases. Certainly the world's richest states (often destination countries) are often partially responsible for upholding the global institutional arrangements that greatly contribute to the bad life prospects of those in the world's poorest countries (Pogge 2004: ch. 4;also Wenar 2008). Rich destination countries cannot fully dodge responsibility for migrants' positions. Additionally, the effects of climate change on some of the world's poorest countries are undeniable. Destination countries in the EU and North America contributed most to climate change, so some destination countries have caused the dependent position of prospective citizens: climate refugees broadly understood. At best, then, the disanalogy argument holds for the subset of migrants whose dependent situation is in no way the consequence of the economic global system or climate change. 11 This disanalogy argument, if correct, supports a limited and conditional justification for differential treatment because there is only a severely limited right to exclude newcomers by plane: it can only justify differential treatment of prospective migrants who are not in a dependent position, or those prospective migrants for whose dependent position the destination state carries no responsibility whatsoever.
However, the argument fails on other grounds, because P3 is not plausible. It assumes that one has obligations to the dependent if and only if one renders them dependent. This is an assumption that is hard to defend. To say that obligations to children are exclusively grounded in causal responsibility would imply, for instance, that there is no general obligation to care for neglected children. P3 seems false. Causing people's dependency is sufficient for acquiring obligations to act on this dependency, but not necessary. An intuitively much more plausible argument is that children have a right to a good enough life, and if their creators do not fulfil their duty, others need to 'pick up the slack'. Dependency itself generates a sufficient claim for support, not being rendered dependent.
What does this imply for the comparison at hand? The difference between all migrants below the relevant threshold (i.e. a good enough life) and newborns disappears. A migrant who does not have the prospect of leading a good enough life, but would acquire such prospects by becoming a citizen of a state, has the same kind of claim as a newborn. Both are owed this equally, unless another relevant difference can be identified.

Vulnerability and responsibility
Perhaps the relevant difference between migrants and newborns lies in what is at stake for them. For many migrants leading a good, or good enough, life is at stake whereas for newborns life itself is at stake. They would not survive their first hours in this world without support from others. Children are not only dependent, but vulnerable. Could we justify special obligations to children by appealing to their vulnerability? 12 Let's say someone is vulnerable if and only if one sees her basic rights violated or runs the risk of seeing them violated, avoidably so. An argument for inclusion of newborns may run as follows: P1:the state has a general duty to protect the vulnerable; P2:all newborns are vulnerable; C1:therefore, the state has a duty to protect children from vulnerability.
In order for a right to citizenship to follow from this argument, we need: P3:citizenship is the best way to protect newborns against vulnerability; 11 There is a non-trivial sense -although anti-cosmopolitans deny it is a relevant sense -in which destination countries cause the dependent position of all migrants: the mere fact that states impose restrictions on migration places people in this dependent position. Wherever a prospective migrant could get access to a good (enough) life as a citizen, the border restriction is a necessary condition for someone's dependent position (e.g. Van Parijs 2007, 649-650). 12 Goodin (1984, 81) makes a similar move from causation to vulnerability as a ground of parental obligation to children.
C2:the state owes children citizenship.
A disanalogy argument aimed at P2 would have to show that newborns are vulnerable in ways migrants are not, so that the equivalent of C1 does not apply to migrants. If children are abandoned at birth they will die, but migrants are not necessarily vulnerable in this sense: Vulnerability-based disanalogy argument: children are vulnerable and migrants are not. Therefore, children have a right to citizenship while migrants lack this right.
There are three ways to disarm this disanalogy argument. First, one could deny that this difference always holds. Some migrants are vulnerable. Miller for example understands refugees to be those migrants who are at risk of seeing their basic rights violated (Miller 2015, 395), hence they are vulnerable. 13 Although statists diverge on this issue, manylike Miller -do think quite substantive and demanding obligations are owed to refugees. Some prospective migrants, then, escape the disanalogy.
Second, one could deny that vulnerability generates a claim to citizenship. Many statists do so. Miller does: on his view states have an obligation to do their fair share globally in meeting people's basic needs, not on the needs of each and every vulnerable individual. The only strict duty states have is a negative duty not to violate people's basic rights. Vulnerability, in other words, is not a sufficient condition to generate a claim to citizenship for all refugees. Hence, it cannot be sufficient to generate a duty to include newborns either.
Third, it is not clear that P3 is true. Miller, for one, denies it in the case of refugees. On his view, what refugees -i.e. vulnerable people (trying) to cross the border -are owed is limited. First, as we have seen, a particular state is required to take care of its fair share in protecting vulnerable foreigners. But what a refugee can claim is severely limited even if a state gains responsibility for her: The scope [of the refugee's claim] is limited to whatever is necessary to protect her human rights. It does not extend in any immediate way to the full set of rights and opportunities that a state may make available to its own citizens (Miller 2015, 395).
Refugees are entitled to citizenship if one of two conditions is met: either citizenship is the only way to protect the refugees' rights, or if the protection of their rights requires entry for many years. 14 Furthermore, a particular refugee does not have a claim to access to a particular state because she is vulnerable. The state may 'provide asylum outside of its borders by agreement with another state on condition that human rights requirements are fulfilled there' (Miller 2015: 396). In other words, there are other ways to discharge the duty to protect basic rights. Setting up refugee-camps, providing food in case of famine, water in case of drought, administrative support in case of a failed regime and so on. Only if this is not possible should states host their fair share of refugees. Granting citizenship, on such a view, is far from the default response to vulnerability.
Hence, the fact that newborns are vulnerable does not automatically lead to an obligation on behalf of the state to grant them citizenship. Unless one shows that newborns' rights can only be protected through citizenship of their parents' country, newborns are owed the same as other vulnerable non-citizens: that their basic rights and needs are protected. It is at least imaginable that one could meet these needs in different ways. For example, set up facilities in low-wage countries where professional nurses would provide for the children up to the basic rights threshold, or even above. 15 One could argue that this fails to do justice to babies' real basic needs: to have a loving family and to be parented. Even if this is a right, and children have such rights, differential treatment is not salvaged. A state could in principle subsidize families abroad to adopt newborns (although it is not even clear why states would have to do this if they already take care of their fair share of vulnerable non-citizens). Will these families provide loving homes? 16 Some will and others will not, but this is true for citizen-parents as well. As long as the adoptive parents are good enough parents (e.g. Shields 2016), this is sufficient.
Finally, one may argue that the relevant difference between newborns and migrants is that almost all migrants already have citizenship somewhere, whereas newborns do not. If one thinks people have the right (perhaps generated by vulnerability) to be a citizen of a country, this is a difference between migrants and newborns. However, even if there is such a right, it follows that newborns have a claim to citizenship of a country, not to a particular country. Even if one thinks that the duty to provide newborns with citizenship falls on the state of their parents (the reasons for this would need to be spelled out), the responsible state could arrange for citizenship to be granted by other states in order to fulfil this right.
To be sure, I do not recommend sending children abroad for care or citizenship. I aim to illustrate what seems to be compatible with the argument under consideration. The argument from vulnerability does not lead to citizenship rights for migrants. Barring additional reasons, this is true for newborns as well. The fact that some migrants are vulnerable and all newborns are vulnerable does not justify differential treatment.
Even if one denies this conclusion, and claims that children's vulnerability does ground a right to citizenship, differential treatment is not saved. Children's vulnerability as a ground for citizenship casts the net too wide to justify differential treatment. If there is a general obligation to protect vulnerable children, it is owed to all childrennot merely to the children born to particular people (citizens). Why protect 'our' children from vulnerable circumstances and not other equally vulnerable children? Without further argument we have not yet established a special duty to children of current citizens.

From parenthood to citizenship
Let us now turn to reasons to think about duties to particular children. Newborns are born into a family, into the nexus of rights and obligations between parents and children. One may take the interest parents have in sharing citizenship with their children as a starting point. This is a clear morally relevant difference between migrants and newborns. The argument for the inclusion of newborns seems straightforward. P1.Citizens have a (conditional) right to parent. P2.Sharing citizenship is necessary for parenting. C1.Parents have a right in their child being granted citizenship. The disanalogy justifying differential treatment runs something like this: Parenthood based disanalogy argument: procreator-citizens have a right to parent their children. Citizens do not have a similarly weighty interest in having migrants admitted. Therefore, the state has a duty to grant citizenship to newborns but no equivalent duty to prospective migrants. This is the line that Thomas Carnes pursues. He argues that: 'a state's right to exclude non-members is constrained when that right conflicts with its members' more central right to procreate and raise families ' (2018, 38). Let us grant that having the opportunity to procreate and parent is indeed important (e.g. Brighouse and Swift 2014: ch. 4;Gheaus 2011;Meijers 2020). The argument for birthright citizenship takes, under such an argument, the shape of arguments for chain migration -and indeed it provides grounds for denying the right to exclude if migrants are like newborns in this particular respect. 17 As Joseph Carens points out, it follows from this type of argument that states lack the right to obstruct family reunification: 'nobody should be forced to choose between home and family' (2013, 187). Miller concurs when he says that family reunification claims are valid because they arise 'from the right of existing citizens to have the opportunity for family life, and therefore to be able to bring their spouses and other immediate family into the country' (Miller 2015, 401).
One could argue that having access to parenthood is compatible with not granting citizenship to newborns and deny P2. The interest in parenthood could be realized by allowing children to stay in the country, and have an admission procedure at a later point (e.g. Shachar 2009). That way, parents could enjoy the goods of parenthood without granting children citizenship at birth. Such a proposal would be based on a severe misunderstanding on the goods of parent-child relationship. What matters for parent (and child) are continued relations between parents and children later in life. In addition, having long-term residents without citizenship effectively creates second-class citizens, and presumably living in a state for all of one's non-adult life is sufficiently long-term to lead to a claim to citizenship on these grounds (Carens 2013, 23-24).
It is important to point out whose interests is doing the work on this argument. Suppose that Jack and Marie are madly in love and want to get married and share their lives. Jack is American and Marie is Guyanese. The US owes Marie citizenship not because Marie is entitled to US citizenship, but because Jack -a citizen -has a rightsgenerating interest in sharing citizenship with her. The argument for inclusion of newborns is parent-focused: if Jack and Marie have baby John, the US grants John citizenship not because he has an interest in citizenship, but because Jack and Marie have an interest in sharing citizenship with John. John, too, has an interest in being with his parents. However, because John is not yet a citizen of the US, his interests are treated as those of an outsider. Because baby John is not a citizen, his interest in becoming a citizen is as relevant as the interest of any other non-member who wishes to acquire citizenship.
This view seems under-inclusive. Suppose that Marie accidentally gets pregnant with baby John, despite the fact that Jack and Marie took all reasonable precautions against pregnancy. They do not want a child. In fact, having any child would make the pursuit of their plans in life very difficult, perhaps impossible: both are committed to their careers, like parties and travelling, and highly value the ability to spontaneously radically change their life, such as moving abroad. Suppose that their plans are genuinely incompatible with parenting a child. John, in other words, would be a huge nuisance to his biological parents. Given that his claim to citizenship is completely conditional on current citizens having an interest in sharing nationality with him, John is now without a claim to citizenship to the country he is born into (although he may have other claims). To push discomfort with the implications of this view a bit further, we could imagine a case in which two children are born to parents in completely equivalent circumstances, both having the same means to provide their children with a very good life. But the one is wanted and therefore has the right to citizenship, whereas the other is unwanted and is, therefore, without a claim to citizenship. 18 Carnes is aware of this implication (2018: section IV) and thinks it is not a hard bullet to bite, because this is what happens to children put up for international adoption, too. They are given citizenship of the country of their adoptive parents. This is not convincing. First, states tend to set very high standards for international adoption. States act as if they have special duties to a child born on their territory that go well beyond the protection of basic rights, even if put up for adoption internationally. Part of what makes this implication intuitively acceptable, I submit, is that we assume the children put up for adoption will be taken care of at high standards. 19 But this high standard is only required if the state has special duties to this particular child, and hence a special duty owed to newborns and not to prospective migrants is presupposed in international adoption practices. One may think states lack such responsibility to set high standards for adoption, but this seems rather counterintuitive: without further argument unwanted newborns have no special claim on the state of their parents.
Second, there may be circumstances in which there is a 'surplus' of children, and some children -older, or with a handicap -may not find a home elsewhere. If their right to citizenship is conditional in the way just described, these children seem to be without a claim to the state they are born into altogether. Although they may have a general claim to citizenship somewhere, they seem to be without a claim to any particular state. This seems like a very difficult bullet to bite.
So perhaps the interest that newborns have in acquiring citizenship should be our starting point. Joseph Carens (2013: ch. 2) takes this line to defend birth-right citizenship. He argues that if children stay in the state in which they are born, they will form all sorts of relationships with the people in that state. Being refused citizenship of a state with which one is so closely connected, and which constitutes one's identity, would be unjust. Children of citizens are entitled to citizenship (and migrants are not) because of their ties to the community. This seems correct: granting children permanent residency but not citizenship would be unjust. However, this does not get us very far in justifying differential treatment. Permanent residents coming from abroad would have an appeal to citizenship on similar grounds. This argument works for Carens, because he argues for open borders: non-citizens have a general right to enter and stay (and in due time apply for citizenship). Carens does not defend differential treatment. Statists generally do not buy the arguments for open borders. So, outsiders are not allowed, unlike newborns, to enter and stay and, as a result connect and identify with the society they wish to enter. For Carens' argument for birthright citizenship to have traction in a statist story, we need to answer a prior question: why newborns have a right to enter and -more importantly -stay to begin with? We cannot defend differential treatment by appealing to his argument for birthright citizenship, without presupposing exactly the kind of duties to newborns we are trying to establish. Someone drawing on this argument would have to show why newborns, and not migrants, are entitled to permanent residency.
Luara Ferracioli offers a sophisticated account of a duty to include based on the interest of the children in acquiring citizenship. She argues that children, as a matter of justice, are entitled to be subject to effective paternalism -which parents and states can only provide working in tandem (Ferracioli 2018(Ferracioli , 2870. This, she claims, grounds citizenship rights. There is a clear morally relevant difference here between (adult) migrants and newborns: liberals believe it is permissible -and required -to act paternalistically over children, where it is not required -and often prohibited -when it comes to adults.
The argument, however, does not fully succeed in defending differential treatment. First, the view seems to rest on a circularity similar to the one observed in Carens' view. The right to citizenship is derived from a right to stay in the territory. Ferracioli (2018, 2873) argues that 'it is the fact of on-going residence which triggers a duty on the part of the state to exercise effective paternalism.' But this is what is at stake: one would need to show that that newborns have a right to ongoing residency, whereas those who want to take up residency can be excluded. A related worry is that Ferracioli argues that children are owed subjugation to effective paternalism as a matter of justice. But on a statist account, again, this assumes the newborn falls under the scope of justice Being owed full justice is conditional on being a full member of society. By arguing that children fall under the scope of justice, we presuppose what we're trying to prove. Ferracioli is aware of this worry, and points to the special duties states have for those who stay within their territory -but it is unclear why granting citizenship would flow from the duties to those present in the territory. Even if she is right that we owe children effective paternalism, this could be achieved in another state. 20 To conclude this section, the parental route grounds an disanalogy argument, but the disanalogy is either under-inclusive (excluding some who intuitively need to fall under the scope of special responsibilities) or presupposes the special duties the disanalogy argument is supposed to support. The next sections ask whether disanalogy arguments based on the right to exclude migrants do a better job.

Culture-based disanalogy arguments
One may argue for the right to exclude migrants on the basis that the right to exclude is necessary to protect the national culture or the political culture of a state. Let me discuss each in turn.

Preservation of national culture
David Miller offers two kinds of arguments for exclusion linked to the importance of cultural preservation. The first focuses on the effect of migration on population density, the second mostly on the qualitative change migrants bring.

Culture and population density
Miller argues that the interest nations have in controlling population size grounds a right to exclude, because some cultures value space: preserving natural beauty or certain landscapes. This is a concern about national population size. Miller argues that: . . . members of a territorial community have the right to decide whether to restrict their numbers, or to live in a more ecologically and humanly sound way, or to do neither and bear the costs of a high-consumption, high-mobility lifestyle in a crowded territory. If restricting numbers is part of the solution, then controlling immigration is a natural corollary (2005a, 202).
For this justification for exclusion to be compatible with differential treatment, one would have to defend: Population Density Disanalogy: the right to exclude is needed for the capacity to limit population size, whereas the right to exclude newborns is not.
For this disanalogy argument to work, one needs to show that migrants have a different impact on population size than newborns, but both newborns and migrants add to the total population of a country in similar ways. 21 As long as states incentivize natality or fail to curb fertility levels and grant citizenship to all newborns, this argument fails to justify exclusion of migrants. Only for those countries that actively try to limit population size using the effective and permissible means can justify a form of differential treatment. Most popular destination countries rate their fertility as too low, and many (but not all) claim to encourage higher fertility (UN 2013). A country that claims fertility levels are too low, and actively seeks to incentivize more births, cannot in good faith exclude migrants for reasons of population size. Insofar as states have an important interest in controlling the size of their population, there is a conditional case for differential treatment.

Preservation and control of national culture
One could defend a right to exclude by an appeal to the importance of preserving national culture: P1.States have a right to preserve (or control the development of) national culture; P2.The right to exclude is necessary for the preservation (or control over the development) of national culture; C1.Therefore, states have the right to exclude.
Miller thinks there are two distinct ways in which P1 is true (2005b). First, being part of a national group is intrinsically valuable (also Walzer 1983: ch. 2). Second, being part of a national group is instrumentally valuable, without it people could not be sufficiently motivated to act on demanding obligations of justice. 22 This brings us to two possible disanalogies.
Cultural Preservation Disanalogy: granting newborns citizenship is compatible with (or necessary for) the preservation of national culture, whereas the granting of citizenship to migrants is not.
An alternative formulation of P2 about migrants points to a possibly important difference, key to this disanalogy: P2*:inclusion of newborns is compatible with the preservation of national culture.

Or:
P2**:inclusion of newborns is necessary for the preservation of national culture.
Replacing P2 with P2* or P2**, does not lead to the conclusion that the state has a right to exclude newborns. The disanalogy disappears if one either shows that both migrants and newborns pose a threat to the preservation of national culture, or that neither poses such a threat.
When something is a threat to national culture is hard to say, because it is difficult to establish what a national culture is exactly -both in theory (a matter of definition) and in practice (people disagree on what key elements are). Claiming that both kinds of newcomers will not change national culture is implausible. Newcomers will make their mark. So, dissolving the disanalogy requires us to show that both kinds of newcomers similarly influence national culture, denying P2* and P2**. P2* is stronger than P2**: if newborns are necessary for the preservation of national culture they are compatible with it as well. If P2* is false, so is P2**.
When can we consider a culture preserved? Liberal nationalists like Miller do not think cultural preservation entails a static culture, but that people have an interest in trying 'to maintain culture continuity over time, so that they can see themselves as bearers of an identifiable cultural tradition that stretches backwards in history ' (2005a, 200).
Can we reject P2*? Imagine a country, Isolatia, which completely closes its borders. Would Isolatia be immune from cultural change, or in control of the development of its culture? No. First, in a globalized world, migration is not necessary for external influences to affect a culture. Even if we assume that Isolatia is truly isolated from all 22 But see Pevnick (2009) or Abizadeh (2002 external influences, its culture would not be immune to change. Young people will formulate their plans in life both in reference and opposition to the existing culture and, in the process, inevitably change it. Culture cannot be something static, and its development is hard to control. The only way for a culture to survive is to successfully adapt (Scheffler 2007, 107). Admitting newborns is not necessarily compatible with preservation or control.
To rescue the Cultural Preservation Disanalogy one would either have to show that children can and may be stopped from changing national culture radically, or that the kind of change children bring is compatible with it. Perhaps by shaping their life plans in line with the key values of the national culture? It is questionable that this would work, but more importantly it would be profoundly illiberal. 23 Children should be allowed to formulate and pursue their own plans in life and should not be comprehensively enrolled in a cultural project (although some influence is unavoidable). Unless one is willing to give up on a key commitment of liberal political thought, this route is unavailable.
Is admitting newborns compatible with the preservation of culture, given that they are raised within the national culture and migrants already have a culture? This is too stark an opposition between the two. Children bring their own backgrounds too: they are raised with different ideas and they are confronted by a plurality of ideas growing up. When children make their own plans in life in both reference and opposition to the culture of their parents, change happens. They interpret, adapt, and alter the national culture to a point where it may no longer be recognizable to earlier generations as their culture.
I would claim, admittedly speculatively, that some cultural changes that come from within, may have a larger impact on the cultural fabric of society than waves of migration. 24 To think that a culture can or should go unchanged is a 'misunderstanding of the nature and prospect of culture preservation' (Scheffler 2007, 104): a culture cannot be kept static and its development cannot be controlled. Migrants, taking their own personal background, do the same as newborns: they define their life plans in reference and in opposition to the national culture, in the light of their own personal background. 25 Both migrants and newborns cause discontinuity in reference to the societal culture, creating some continuity and common reference point. 26 Both will considerably influence the development of national culture, and are potentially a threat to the preservation -or control over the development -of national culture.
Some migrants may even strengthen certain values some consider part of the national culture. Imagine someone who regrets, say, the erosion of traditional family values and respect for authority. They might blame this development in part on mass migration, but many of these changes have been accomplished by this person's children 23 To what extent comprehensive enrolment is permissible is a point of discussion (e.g. Clayton 2006: ch. 3;2012; Brighouse and Swift 2014) but each liberal view accepts considerable constraints. 24 One could think of secularization, attitudes to same-sex marriage, the sexual revolution, or more recently the large difference in support in the UK for Brexit between age groups. 25 Miller agrees: 'immigration, in other words, is likely to change a society's public culture rather than destroy it ' Miller (2005a, 200). 26 Sandelind (2015: 497) writes '[i]t is far from clear that adding people via immigration would have a larger impact than both internal migration and the birth of new generations.' and grandchildren. Migrants holding conservative views probably share her worries. These migrants, in other words, could very well be this person's allies in protecting the kind of society she believes desirable. This holds generally: certain migrants may very well be allies or enemies of a national culture depending on one's interpretation of what the national culture consists in, just as newborns may turn out to be allies or enemies on some but not other interpretations of what the national culture is. What about the instrumental, solidarity-based argument for the right to exclude? One could argue that, even if a national culture cannot be protected, the belief that we share a common project is essential for social justice to work. This is a concern about the feasibility of social justice: because it is so demanding, one might argue, a sense of 'us' is essential for justice to work (Miller 1995: ch. 3). The disanalogy between newborns and migrants would, then, go as follows: Solidarity Disanalogy. Newborns are perceived as being part of the in-group, whereas migrants are not. In-group solidarity is necessary for social justice to function. Having too many outsiders in society will lead to a lack of support for social justice. Therefore, states have a right to treat migrants and newborns differently.
If the empirical premise in this argument is true (but see Abizadeh 2005;Pevnick 2009;Blake 2013, 105-106) this disanalogy picks out a genuine difference between migrants and newborns. However, is the difference one that is morally relevant? The concern is that even if it is true, it is only true because it is made true, that is, states 'reproduce and maintain the character of national identities and contribute significantly to the fundamental character of the relationship between those who share national identity' (Axelsen 2013, 463). States make it so that living up to principles of justice is only feasible amongst the citizens of a nation. Similarly, states need the right to exclude because an influx of large numbers of migrants may undermine solidarity. But this is also made true: we are, at least partially, taught to show solidarity mainly with people of our national community. To say to the migrant that she cannot acquire citizenship, while children can, springs from the fact that we (as a community) decide to show solidarity with children of citizens and not with people of another nationality.
The state's past failure to expand the circle of solidarity is used as a reason to exclude the migrant. It seems unfair to ask the prospective migrant to shoulder the burden of society's failure to make its capacity of solidarity more inclusive. The fact that states (and others) actively encourage and teach their population to show solidarity only (or at least primarily) with their co-nationals and their children, and not with foreigners is not a morally relevant difference between migrants and newborns, unless an underlying justification for differential solidarity is available.

Preservation of political culture
If the protection of national culture understood in a thick sense cannot justify differential treatment, could an appeal to the need of protection of -or control over the development of -the public culture of a society do the job? Perhaps the values underlying a political culture are more constant over time, and instilling respect for these values in children is not problematic. We could rephrase the argument offered above to defend such a position, appealing to the importance of protecting our public (perhaps liberal, democratic) culture from changes caused by migration 27 newborns will be raised with the values of our political community, whereas migrants may not share them.
Political culture disanalogy: granting citizenship to newborns is compatible with (or necessary for) the protection of the national political culture, whereas granting citizenship to migrants is not.
Does the disanalogy succeed in this case? It does not hold for all migrants: some do not share liberal democratic values, others do. Because this notion of culture is much thinner, it is much more likely that there are non-citizens who do. If the disanalogy holds, it holds as a proxy: one would have to assume (or show) that being born or raised a child of current citizens is a good proxy for support of the public culture of a society. However, some migrants do share the relevant values (and others are willing to embrace them), some newborns will grow up to reject them, perhaps because they were raised by citizens hostile to these public values. The state has an interest in limiting the number of unreasonable citizens, home-grown or not. Taking the country of birth as a proxy for support of the public culture of a society is impermissibly imprecise.
To take it a step further, if states may legitimately refuse citizenship to those migrants likely to hold unreasonable political views, may they also refuse citizenship to those newborns who are likely to disavow the values underlying the public culture? This is not what states actually do; rather, they try to secure support for public values through public education. States could of course subject prospective citizens to similar compulsory education, for example, citizenship-classes that could serve the same purpose of trying to secure support for public institutions. If one lacks justification for excluding newborns with a high statistical probability of not supporting the values on which public institutions rest, one lacks justification for excluding migrants on similar grounds. There are many reasonable prospective migrants, and many unreasonable citizens. Migrants and newborns are not that different in this respect.
The protection of a public culture cannot ground differential treatment. Miller agrees that the public culture argument does justify discrimination between reasonable and unreasonable migrants at the point of entry: [O]ur understanding of national cultures in recent years has primarily involved subscription to a set of political principles, together perhaps with some familiarity with the history and customs of the country in question. So understood national culture cannot provide a strong rationale for discrimination at the point of entry, since it can be argued that immigrants will quickly adapt to the new political environment in which they find themselves, and can also be required to familiarize themselves with the aspects for the local way of life as a condition for admission to citizenship (Miller 2007, 229).
Discrimination between newborns and migrants falls under the scope of this claim as well. Support for the political culture provides no basis for discriminating between migrants at the point of entry and newborns.
The arguments here have focused on preservation, understood as control over the development of the culture of the receiving society. These are instances, one may argue, of political self-determination: the capacity of a collective to determine what direction they are going. I have not shown that an appeal to self-determination cannot ground any justification for differential treatment, but what the arguments in this section do try to show is that when it comes to preservation or control over the development of national culture (in the thin or thick sense), migrants by plane and by stork do not significantly differ. Both provide risks and opportunities for cultural self-determination in very much the same way.

Sustainability based disanalogy argument
Defenders of the right to exclude have appealed to the importance limiting the global population for sustainability reasons. This argument, appealed to by both statists (Rawls 1999;Miller 2007;Meijers 2017) as well as more cosmopolitan minded egalitarians (Barry 1992), runs as follows: P1.The size of the global population should be contained (for sustainability and intergenerational justice reasons); P2.Holding states responsible for their own population's size is necessary to contain global demographic growth (Miller 2005a, 102;Rawls 1999, 118); C1:We should hold states responsible for their own demographic growth. P3.The right to exclude is necessary for states to take responsibility for their own demographic growth; C2.States have the right to exclude.
To show that this argument applies to migrants but not to newborns, one would have to defend the following disanalogy: Sustainability disanalogy: not having the right to exclude undermines sustainability, whereas granting citizenship to newborns does not.
Before offering some reasons to doubt that the argument for exclusion is correct, let us assume it is 28 and focus on the disanalogy. Can a state say to a prospective migrant: we have the right to refuse you, because not having this would undermine our capacity to limit the size of our population, which would, in turn, be bad for the size of the global population?
Countries that attempt to limit their population size (using permissible means) could refuse migrants on these grounds without being inconsistent. Countries that have a steady or shrinking population and are not trying to return to demographic growth could do so, too. However, many of the world's most popular destination countries have, by their own estimates, a demographic deficit, due to shrinking and/or ageing populations. The EU -for example -has been pondering ways to increase fertility levels in order to address these issues. The European Commission gave 'policy directions' to member states to 'support demographic renewal through better conditions for families and improved reconciliation of working and family life' (Europe Commission 2014, 10) in order to meet the challenges of demographic ageing. Individual destination countries report that they consider their fertility rates too low, and many have reported the intention to use policies to raise fertility levels (UN 2013). Can we exclude one and not 28 There are reasons to doubt P2. There is evidence that migration to richer countries leads to lower fertility globally.
First, migrants' fertility levels converge with the local fertility levels (in many rich destination countries lower than in most source countries). Second, migration to low-fertility countries impacts fertility levels in source countries: fertility norms travel. See Beine, Docquier, and Schiff (2013).
the other? It is inconsistent to refuse migrants for reasons of population size, but do nothing to control the influx of migrants 'by stork'. Such a state would be like the tennis club mentioned in the introduction: refusing the migrant membership with the argument that the courts are overbooked, while at the same time trying to increase membership elsewhere. This tells us something about what kind of arguments actual states can and cannot (consistently) use. Is this an unfair move against statism as a theoretical position, using this contingent fact about states (i.e. that they would prefer more newborns) to undermine this argument? Miller explicitly says that the arguments he offers: 'do apply to many liberal democracies that are currently having to decide how to respond to potentially very large flows of immigrants from less economically developed societies' (Miller 2005a, 199). Exactly these liberal democracies that are confronted with -and complain about -low fertility. For them, the argument from sustainability is unavailable as a justification for differential treatment here.
A state that does what is required with regard to limiting its own population, can consistently refuse migrants in order to make sure that other states cannot externalize their failure to limit demographic growth. But only if certain other conditions are met. Simply limiting one's own demographic growth to acceptable levels (whatever these are exactly) is a necessary but not a sufficient condition for legitimate differential treatment. An example will help. Replacementratia's fertility levels are below replacement rate, and it refuses new migrants on the grounds under discussion: it wants to set the incentives for global sustainably straight. However, Replacementratia does not act on its duties of minimal global justice, supports an unfair international trade system and externalizes the costs of its polluting industries to other countries: it fails to alleviate (and actively contributes) to global poverty, and it violates principles of peripheral 29 global justice. It supports unjust regimes that allow and facilitate the suppression of women in society. Given that poverty and gender inequality are important determinants of fertility (Meijers 2016) Replacementratia is indirectly contributing to global population pressure by not doing its duty, and hence it cannot, at least not without acting in an incoherent way, exclude migrants on global population growth grounds.

Conclusion
This paper asked whether differential treatment between migrants by plane and migrants by stork can be justified. Can statists justify the practice that states refuse citizenship to one but not the other? Is it possible to offer reasons to the migrant requesting citizenship, and asking why she is not let in whereas plenty of newborns are?
Several arguments that seemed relatively plausible from the outset offer limited and conditional justification for differential treatment. Two reasons survive scrutiny: sustainability-based differential treatment, and parenthood-based differential treatment. Those states that are actively trying to limit their own population size and act on their duties of global justice (or humanitarian duties), can offer a coherent reason for exclusion. For most states this argument is not available. Many destination societies are refusing migrants without doing their fair share, even on less demanding statist standards, internationally. And many destination states regret low fertility numbers amongst current citizens. Differential treatment, even with statist arguments in hand, is hard to justify on these grounds.
In addition, the parental argument succeeds in justifying differential treatment. Statists can appeal to the interest current citizens have in parenting the child, and sharing citizenship with it. A state can say to the prospective migrant: 'This newcomer (the child) is granted citizenship because her parents have an interest in her acquiring it, whereas no current members of our society have a similarly important interest in sharing citizenship with you.' This argument has an interesting feature: the child is not entitled to citizenship, but the parents are entitled to their child gaining citizenship. Acquiring citizenship, the right to full standing in a particular society, is reduced to a derivative right: a newborn is entitled to citizenship to a particular state only because (and only if) it is in the interest of current citizens. This has, as a counterintuitive conclusion that unwanted children are owed very little: they have no stronger claim to the assistance of the state than any other outsider. To make this implication palatable (by requiring acceptable standards of international adoptions), statists will have to posit stronger duties to nonmembers. For example, by appealing to the vulnerability or dependency of the newborn. And, as I've shown in this paper, it is hard to do so consistently in a nondiscriminating fashion: the duties to the unwanted child will spill over to other outsiders. If statists want to avoid this implication, more severe restrictions on the right to exclude would seem inevitable.
Do these arguments hold any value for debates amongst cosmopolitans on migration? The sustainability-based reason for exclusion is available to those holding a cosmopolitan position as well. If the facts hold up, a cosmopolitan right to exclude as well as differential treatment might be justified on global sustainability reasons. The second, parenthood-based, argument is more difficult to reconcile with cosmopolitan views. This effectively depends on how important one thinks the right to parent is. To justify inclusion of newborns and exclusion of migrants, one would have to show that the interest people have in parenting is weightier than the interest migrants have in acquiring citizenship. It is hard to imagine that the interest in parenting could outweigh the interests of refugees, if partiality for compatriots is taken out of the equation. The same counts for many people migrating for economic reasons. For a cosmopolitan, the parental argument for differential treatment is weaker, because it depends on the statist claim that partiality for compatriots is justified (or required). For those cosmopolitans who embrace open borders this is, of course, much less of a problem: they are not interested in defending differential treatment anyway.
At least one big question about the parental argument remains. Is the right to citizenship not too important to be dependent on a derivative right? Intuitively many will find that there is something troubling about a state replying to a prospective citizen by saying: 'our citizens don't have an interest in sharing citizenship with you, therefore you do not get it.' Those with cosmopolitan intuitions like myself, will think that the fact that some people are more wanted as fellow-citizens is a morally arbitrary fact about a person, which shouldn't determine the kind of life one has. This, of course, is a well-known implication of the statist view: accidents of birth determine how well one's life goes. Statists may simply accept this as an implication of their view. It does seem to come at some intuitive costs. One implication is particularly hard to stomach: we do not owe citizenship to unwanted children. But this is ultimately not a question of the internal coherency of the view, but an external cosmopolitan concern about the plausibility of the statist commitment. If statists are troubled by these consequences, they will have to conclude that the argument for differential treatment has implications that are intuitive evidence against statism or, at least about the grounds and the limits of the right to exclude.