Should Cities Control Immigration Policy?

Avner de-Shalit wants cities to have their own immigration policies. On a radical reading, this would transfer control over immigrant admissions from states to cities. But can cities choose the immigrants they prefer on economic or cultural grounds, or does this dis-criminate unfairly against those judged to be less desirable? I argue that de-Shalit fails to apply the luck egalitarian principle consistently when discussing immigrant admissions. I also claim that there is a tension between seeing cities as the bearers of distinct cultural ethoses, and therefore as bulwarks against the homogenising effects of globalization, and disbarring them from carrying out culturally selective immigration policies. De-Shalit ’ s own preferred model of the immigrant-friendly city – Amsterdam – appears to lack any distinct ethos, other than an ethos of welcome and cultural blending. Moreover, democratic states also have a legitimate interest in controlling immigration. They must be concerned about the consequences for social justice of admitting migrants and the political effects that follow when the migrants become citizens themselves. They must also consider the environmental impacts of population growth. Cities should play a major role in integrating immigrants, but not in admitting them.

I am going to leave aside many of the questions raised in Avner de-Shalit's rich and provocative book in order to focus on his claim that cities should be able to make their own immigration policies. 1 In today's world, as we know, controlling immigration is normally seen as a matter that rests entirely in the hands of the state (the EU provides a partial exception), and debates have raged over whether states are entitled to exclude migrants in the first place, and then if so, on what grounds and subject to what limits. Yet on reflection, we might think that cities should be given a significant role in immigration policy. After all, the city is the place where most migrants choose to live, so it is city-dwellers who stand in the front line to reap the benefits and shoulder the burdens that immigration brings with it. Shouldn't a principle of subsidiarity apply here as in many other areas of public policy? My remarks will be divided into two main sections. In Section 1, I will look at immigration through the eyes of the city and ask why cities might want to attract or repel immigrants and how such policies could be justified. I claim that de-Shalit presents us with two different models of the city to contemplate, and the answer may depend on which model we find more appealing. Then in Section 2, I will look at the reasons that states have for wanting to retain control over immigration and why, therefore, they will be reluctant to pass this power down to the city level. I begin by asking what exactly it would mean for cities to have their own immigration policies.

De-Shalit on City-Based Immigrant-Selection Policies
There are two main ways in which a city might seek to control migration flows, one considerably more radical than the other. The less radical alternative would leave the initial decision to select and admit migrants in the hands of the state. Once the selection has occurred, however, cities would be given the opportunity to attract the number and type of migrants that they wanted from the national pool. They could do this by offering incentives of various kindsguaranteed jobs, low-cost housing, cultural facilities, and so forth. Migrants would then be presented with a choice between urban 'packages' on arrivalof course without a guarantee that everyone would get the package they most preferred, since cities could choose who to host.
The more radical alternative would place primary selection of migrants in the hands of cities: London would have its own immigration policy, Birmingham and Manchester likewise. UK immigration policy would then simply be an aggregate of these citylevel policies: a person turning up at the national border would need to produce an entry visa showing that one or other city had offered them admission. They would then be directed to the city that had chosen them and asked to take up residence there at least for some time.
Although de-Shalit is not always explicit about which of these alternatives he is considering, the general position advanced in the book suggests that it is the second, more radical, version that he favours. Although cities would not control their own borders physicallymovement in and out should be free for everyone, barring criminals or others posing a risk to public safetythey would issue visas to prospective immigrants that granted them residence rights (pp. [3][4]. So migrants entering the state would be required to settle in the city that had selected them, at least for a period of years. De-Shalit does not say what would happen to those who used their right of free movement to migrate to another city. Perhaps he would rely on immigrants being held in place by economic incentives and disincentives (such as loss of welfare benefits if they decided to move elsewhere).
Regardless of whether cities attract migrants by granting them entry visas or by offering inducements to those whom the state has admitted, the issue of the criteria that can legitimately be used in selection arises. De-Shalit first approaches this by considering what could justify refusing someone entry and argues that criminality is the only such ground. 'Any other selective refusal system would be discriminatory on the basis of brute luck, discriminating against those who happen to lack the skills which are needed in a particular city or who happen to hold different sets of beliefs. People should not suffer because of their bad brute luck' (p. 49). He continues the theme on the following page when arguing that immigrants who are turned away because they lack useful skills or other desirable attributes have no reason to accept their exclusion as legitimate. 'Consider a person who is very talented as a tailor and less talented, say, at math. In the nineteenth century he would have been welcome in many cities; today, through no fault of his own, his skill is not needed' (p. 50).
But if the luck egalitarian principle is going to be applied in this way to the case of entry refusal, simple logic seems to require that it should also be applied to those who will be selected under an admission or incentive system. Even if not refused entry outright, the unfortunate tailor can complain that he is unfairly being excluded from whatever set of benefits the city is offering to attract the immigrants it wants. Luck egalitarianism, to paraphrase Weber, is not a cab to be taken at will. 2 Nevertheless de-Shalit seems to think that an appeal to 'the freedom, health and well-being of citizens' can provide grounds for selecting in favour of immigrants who can provide vital services like health and education. The argument here is couched in terms of comparative need: the needs of the (less essential) migrants who are turned down can be outweighed by the needs of children and the elderly (p. 50). It's not clear, though, that this could deal with the complaint of the unchosen tailor, who could argue that it's through no fault of his that the city now needs teaching or nursing skills rather than tailoring.
This way of interpreting the luck egalitarian principle strikes me as peculiar. The tailor cannot, of course, be held responsible for the city having the job requirements that it does, but he can be held responsible, at least to some extent, for failing to develop the skills he will need to have to be attractive as an immigrant. De-Shalit attempts to rule out a response of this kind, saying that 'without entering the debate about whether or not we really freely choose what skills to have, my point is that at the moment of migration, the immigrant lacks a needed skill, and this is, indeed, something the immigrant did not choose to lack' (p. 49). But this rests on an implausible way of understanding the luck egalitarian principle, which would count as relevant only choices made at the point at which a particular allocation of resources is being made and ignore previous choices that will determine the attributes that a person possesses at that point in time. We would need to be given some special reason to think that the tailor should be exempted from responsibility for decisions made prior to the moment of migration. But my main point is that however the luck egalitarian principle is applied to this case, it must be applied consistently and without regard to the particular grounds on which the city might choose its newcomers. Using social need as the criterion still favours the lucky ones who possess the relevant skills over the unlucky ones who don't.
De-Shalit suggests a further reason to justify the selective encouragement of particular categories of immigrant: that they 'can easily integrate and later on contribute in the effort to integrate future immigrants' (p. 50). He cites evidence showing that when immigrants arrive and settle, they tend to do best if they are able to link up with established communities that share their religious and cultural values, easing the transition to their new social environment. So this is a consequentialist argument that shows that a city that later on expects to welcome immigrants of type T now has good reason to select in favour of those who display type T cultural traits. As he puts it, 'the justification for selective invitations is therefore based on the capability of those selected to contribute to the successful integration of potential additional immigrants in the near future' (p. 51).
The sociological basis for this argument is almost certainly sound, but its normative status is puzzling. A city that has already decided that it wants to attract type T migrants in future has good reason now to welcome type T migrants as trailblazers. 3 But what is the basis for its long-term preference for migrants of type T? Unless cultural selection can be independently justified, it seems that the argument speaks merely for an immigration policy that is consistent over time: if you begin by attracting migrants who are Sikhs, say, then it makes sense to continue attracting them in future, since they will be more likely to integrate. But what can justify the decision to single out Sikhs as the group you wish to attract? Even on a wider interpretation of luck egalitarianism than the one that de-Shalit favours, it must be brute bad luck to be a Muslim or a Hindu wanting to move to a city that has adopted a policy that favours Sikhs. So it seems that the principle of justice that rules out cultural selection as morally arbitrary collides with his consequentialist argument that it makes sense to attract immigrants now who will help later immigrants to integrate.
Setting luck egalitarianism aside, therefore, we can ask directly whether it is legitimate for cities to favour certain immigrants over others on cultural grounds. This must depend on how we view the city as a social and political community. In earlier work (jointly with Daniel Bell), de-Shalit argued persuasively that many of the world's leading cities each had their own special 'ethos' and moreover that this distinctiveness was something to be treasured: . . .cities have been increasingly the mechanism by which people oppose globalization and its tendency to flatten cultures into sameness. Many cities invest thought, time, and money in protecting their unique ethos and preserving it through policies of design and architecture and through the way people use the cities and interact with them. Arguably, not all cities do this and some may simply surrender to the demands of globalization. But the idea that cities can and should promote their particular ways of life does not arouse much controversy: even defenders of liberal neutrality at the level of the state tend to allow for the public expression of particularity at the level of cities. 4 This view of the city (which it should be noted is both empirical and normative) can certainly be found in the present book, albeit modified by the observation that 'the very ethoses of the cities which their current inhabitants want to retain were, in fact, shaped by, among other things, interaction with newcomers' (p. 36). But it coexists with another view of the city in which its internal relations are conceived more on the model of Gesellschaft (association) than of Gemeinschaft (community). Here what is highlighted are the economic and cooperative interactions that city life involves, with culture reduced to expectations about how others will behave: In many of the economic activities we do we also assume the actions of others in the city. When I plan my day, I assume that I will be able to buy a loaf of bread in the morning and prepare a sandwich for lunch. I take it for granted that somebody will wake up very early and travel to the bakery and bake the bread; another person will drive the bus on which the baker will travel to the bakery; the police will ensure that traffic is smooth, that if the lights need fixing somebody will do this, and so on (p. 89).
Whether deliberately chosen or not, the example of the baker inevitably conjures up Adam Smith's account of how social coordination is possible on the basis of mutual advantage (and convergent expectations)the famous 'invisible hand'without the need for a shared cultural or normative framework. 5 All I need to know about the baker is that he can bake (good) bread and that he can be relied on to rise early enough to bake it for me. His cultural background, his values, his sense of identity are all irrelevant. On this second view of the city, then, we have no reason to be concerned about the cultural impact that immigration might have and no reason to consider cultural factors when deciding which immigrants we want to attract.
The question that arises, therefore, is whether de-Shalit understands the city in this Smithian way or whether his view is more communitarian, or 'Hegelian' as he suggests on p. 102. 6 A city seen as a community with a distinctive ethos could still accept immigrants, but it would expect them to adjust their outlook and their behaviour to fit in with the prevailing ethos. There are points at which such a view appears to be endorsed 7 : A city can decide about its public culture (e.g. it is Catholic) and yet tolerate people who espouse different cultures, especially if their cultural expression is most manifest at home. Seventeenth-century century Amsterdam tolerated Jews even though the city was Christian. The question an immigrant should be asked is not whether she is ready to convert but rather whether she is willing to live in coexistence with the dominant culture. This is a reasonable threshold which nearly all immigrants would accept (p. 38).
To help clarify de-Shalit's position, we can consider the contrast he draws between the cities of Jerusalem, Berlin, and Amsterdam in the final chapter of the book. Each of these is presented as exemplifying one possible (and by implication legitimate) way of including immigrants. However, he reserves the highest praise for Amsterdam, which he describes as a model in which the native inhabitants positively seek out and immerse themselves in the cultures that immigrants bring with them, to the extent that they may actually begin to participate in some of the immigrants' practices. By way of example, he cites an Amsterdam resident who claims that 'in the morning I am Christian because this is how I was born and because at work I work with Christians; but in the afternoon or evening I am Muslim because I celebrate my next-door neighbours' holidays with them, and I do not do this like an anthropologist would do but as one who tries to internalize their customs'; p. 113). He calls this the milkshake model of the city, in which 'people mix and assimilate each other's personalities, values, norms, and customs' (p. 148).
If this is indeed an accurate depiction of life in the city today, then 21st-century Amsterdam is clearly very different from 17th-century Amsterdam, which according to de-Shalit had a dominant public culture while allowing minorities to practise their own cultures in private. Instead, cultures blend in such a way that it is hard to say that the city any longer has a distinct ethosother than the ethos of curiosity and experimentation implicit in the blending process itself (Amsterdam did not feature as a case study in The Spirit of Cities). In other words, if we imagine several such milkshake cities, they are going to look very much alike; at most, their colour and flavour are going to be a little different depending on the cultural ingredients that are blended. Moreover, there would be no reason for such a city to pursue a policy of selecting people from particular cultural backgrounds to join; random selection would be enough to ensure a sufficiently diverse set of ingredients for the milkshake.
Things look different, however, in the case of a city that rejects the milkshake model, preferring instead, in Hegelian spirit, to preserve its cultural distinctiveness over time. For it faces three broad options when it comes to admitting immigrants (mixed strategies are also possible). Option 1 would be to select only those whose cultural values already reflect those prevalent in the cityfor example, they speak the same language or practise the same religion. Option 2 would be to admit immigrants without regard to their culture but then to engage in an active policy of cultural assimilation, so that over time they converge with the native born. Option 3 is to admit immigrants, but on the condition that they do not attempt to alter the city's public culture, instead confining their activities to the private spherefollowing the older Amsterdam model. Each of these options poses a challenge to contemporary liberal values. Option 1 would be widely condemned as discriminatory at the point of entry; Option 2 would be seen as oppressive, since it curtails the freedom of the immigrant groups to pursue their own cultural development; while Option 3 appears to contravene liberal equality, since it reinforces the privileged status of the city's established public culture with no concessions to the incoming minority groups.
De-Shalit does not discuss these options explicitly, but we can deduce roughly how he would respond to them. One of his examples, Jerusalem, can be seen as a lessthan-fully successful attempt to implement Option 1, since although it does not have the formal capacity to select its incomers, it piggybacks on Israel's discriminatory Law of Return, and 'Jerusalem is the most popular destination among Jews who migrate to Israel' (p. 122). (It is less than fully successful since many Jews also leave, so the Muslim population is growing at a faster rate than the Jewish.) Although de-Shalit regards current relationships within the city as far from idealhe says the two main communities, Jews and Muslims, live alongside each other 'like two exhausted boxers' with no attempt to integrate in terms of values or identityhe does not challenge the immigration policy that privileges Jews. Thus in a case where a city is strongly infused with a dominant culture, he seems willing to allow a selection policy that favours immigrants from that cultural background in order to preserve its character. Option 1, then, is permissible, though clearly suboptimal, for de-Shalit.
Option 2, however, is ruled out. De-Shalit is in favour of assimilation, but only when it takes the form of mutual assimilation as displayed in today's Amsterdam, where native inhabitants and immigrants want to explore one another's cultures and adopt parts of them, thereby producing a blended milkshake. He does not say exactly why he dislikes one-sided assimilation: perhaps he thinks that it can never be wholly successful (immigrants try but ultimately fail to pass themselves off as natives), or perhaps he thinks that to demand unilateral assimilation infringes the right of the immigrants to preserve their own culture.
Option 3 is also not discussed explicitly, though on the basis of his remarks about 17th-century Amsterdam cited above, de-Shalit would not find it objectionable. What implications would it have for a city's immigration policy? First, favour immigrants who are content with private forms of cultural expression. Second, try to avoid recruiting too many immigrants from the same background lest they consolidate into a political bloc and begin making demands on public space and public forms of cultural expression. Third, discourage the formation of immigrant enclaves in which their culture becomes dominant. I do not mean to suggest that de-Shalit would favour such policies. But they are the price that would need to be paid if a city wants to admit substantial numbers of immigrants who do not share its prevailing 'spirit' while still ensuring that that spirit remains dominant. This highlights the tension between seeing the city as a bulwark against the culturally levelling effects of globalization and treating it as the solution to the many pressing economic and other problems that cause people to migrate.

Why States Want to Retain Overall Control of Immigration Policy
So far I have been exploring the reasons why cities might want to enact their own immigration policies. I now turn to examine the reasons why states are likely to resist such a demandespecially in its more radical versionand insist on retaining control over immigration. Interestingly, up until the modern age, it was cities rather than states who were more likely to have proactive immigration policies. This was partly because cities needed to draw in migrants in order to sustain their populations, and partly because their economic prosperity often relied on attracting artisans and merchants from elsewhere, so when groups with these desirable attributes were expelled from other placessuch as the Huguenots from France and the Jews from the Iberian peninsulathey were warmly received in cities such as Amsterdam, Berlin, London, and Venice. However, this sometimes ran counter to the wishes of state authorities, as an historian of the period explains: In Western Europe, the political autonomy and 'freedom' most cities enjoyed implied that they could pursue their own policies of attraction, restriction, and expulsion with regard to immigration. Urban ordinances designed to stimulate the influx of those considered valuable and useful and to prevent the entry of 'idle strangers', are therefore a familiar and recurring theme to anyone studying late medieval or early modern urban policies. It is worth remarking that these often ran counter to the wishes and endeavours of central authorities, for whom considerations of social stability and political control tended to predominate over the economic considerations that motivated urban authorities. National restrictions on labour mobility in Tudor England, which were inspired mainly by concerns for societal stability, were for instance actively circumvented by urban authorities seeking to recruit an adequate labour supply. 8 We can hear echoes of this conflict today in the case of so-called 'sanctuary cities' that try to resist the deportation of unauthorised migrants by, for example, refusing to pass information about them to state authorities. 9 But why should states feel that they have to get involved if cities are happy to continue to host the migrants, who they presumably see as valued members of their community? Why, more generally, should states be unwilling to pass control of immigration entirely over to cities?
At this point, it is worth reminding ourselves briefly about why democratic states in particular should want to control immigration. 10 A central reason is their interest in self-determination, understood as the wish on the part of citizens collectively to decide the future shape and direction of their society. On the assumption that immigrants are likely to arrive with values and cultural traits acquired in their societies of origin, their presence, in significant numbers, will change the receiving society in two main ways. First, it will require new initiatives in public policy to meet the needs and legitimate demands of the new arrivals. Education, housing, and health and social service policies will all have to be expanded and adjusted to accommodate them. Religious and cultural differences will necessitate changes in the law and in public provision to ensure that immigrants are treated fairly. None of this need be problematic, but the scale of change may be considerable. Then second, when the immigrants themselves qualify for citizenship, they alter the composition of the demos itself, and this is likely to affect the decisions that the citizen body will make in the future. Insofar as present-day citizens have an interest in continuity over timeand this is plainly one important aspect of self-determinationthey have reason to look ahead and ask how today's immigration policy will influence the choices of tomorrow's enlarged demos.
What bearing do these general considerations have on the relationship between cities and immigrants? In Chapter 2, de-Shalit canvasses a range of reasons for granting newly arrived immigrants what he calls city-zenship, i.e. voting rights in the authority that governs the city they have settled in. These reasons are persuasive, and I agree with de-Shalit that it is desirable for immigrants to become integrated politically at the local level as soon as possible. But these city-zens will in due course qualify for full citizenship, including the right to vote in national elections. We agree again that it would be both ethically and politically unacceptable to deny them this right, though we have different views about the value and legitimacy of citizenship tests as a precondition of access (see pp. 93-94). At this point, their votes will have an influence on public policy throughout the state, not just on the city they have moved to. So their having the vote is a matter that concerns the society as a whole. Take the more radical version of city-based immigration policy where each city is allowed to choose which immigrants are admitted to the state. If they are subsequently given the vote in national elections, as we agree they must be, then this creates a kind of externality for everywhere except the city that chose to admit them.
A possible reply here is that no single group of immigrants will have a significant impact on national policy, so the problem I have identified is only theoretical. 11 Even so, the demands on public policy that immigration creates have to be considered. Whose responsibility will it be to meet these demands? At one extreme, we might imagine giving cities full control of their own education, health, housing budgets, etc., which they would presumably need to fund through local taxation. In the practice of most contemporary democracies, howeverand leaving aside the complications introduced by federal arrangementswe find that both policymaking and financing are largely carried out at state level, and there are good reasons for this. Citizens rightly expect that the level of provision should be roughly the same wherever they live: the schools and hospitals should be equally good, unemployment and disability benefit should be standardised, and so forththis is what social justice requires. Unavoidably this means some level of cross-subsidisation between different areas of the country, for reasons of demography, etc. Suppose now that cities are given the right to select immigrants for admission to the state. Depending on the profile of the people who are selected, this might impose significant demands on national budgets. The immigrants might, for example, have special educational or medical needs and in the short term at least be unable to find work that pays well enough to allow them to cover these costs through taxation (as we well know, states try to avoid this possibility by selecting immigrants who are already highly skilled and therefore likely to be fruitful taxpayers).
The interests of city and state may also diverge over the environmental impact of immigration. This is a topic that is rarely discussed in the philosophical literature on migration, and that remains the case here. 12 When cities swell through immigration, they also tend to expand outwards, encroaching on what had previously been protected as green belt land (a striking example is provided by the Ontario Green Belt around Toronto, now under increasing pressure from a city of six million people, of whom approximately half are immigrants). 13 Although the loss of wildlife habitats and green spaces may be felt most acutely by the city's own inhabitants, it will also have an impact on national policies of environmental conservation. More significantly still, states will be concerned about the additional greenhouse gas emissions that immigration will inevitably produce, given that the majority of immigrants move from a developing to a developed country. It is an inconvenient truth that when a person makes such a transition, their carbon footprint expands as they adjust to the ecologically destructive lifestyle that prevails in the latterinconvenient for anyone who cares about the environment but also sees migration as essential to solving global poverty. 14 Of course, states can and should take steps to counteract these effects, by reducing everyone's footprint. But they will be hampered in their attempts to meet emission targets if they relinquish control over the numbers of people who arrive and the places where they settle.
Might cities themselves take on the responsibility for environmental policy? Where states fail to act effectively to control emissions or protect ecologically sensitive sites, there is a strong argument that cities or regions are morally required to step into the breach, as several are now doing. 15 But this is clearly a second-best solution: states are the bodies that commit themselves to emissions targets in international treaties, and they are also the bodies with the power to force business and consumers to make the necessary changes to their behaviour. So they must remain the primary actors if an effective environmental policy is to be pursued. And this, I have suggested, entails retaining overall control of immigration.
The concerns I have raised about placing immigration policy in the hands of cities obviously apply most forcefully to the radical alternative that would allow cities to select their future city-zens at the point of entry. But even in the case of the less radical proposal that would permit them to be proactive in encouraging immigrants to come and settle by offering incentives, there might still be reason for the state to want to exercise some overall control. Social and political integration is a legitimate public concern, and this may provide reasons to discourage immigrants from similar backgrounds from clustering in very large numbers. De-Shalit's observation that the presence of one cohort of immigrants can help to integrate the cohorts that follow is correct, but only up to a point; it is equally important that immigrants should be encouraged to interact with nonimmigrants, which they will have less incentive to do if they are drawn to a neighbourhood populated almost entirely by people from their own ethnic community. 16 If the worst comes to pass and race riots break out, it is the state that bears the ultimate responsibility to restore peace.

Conclusion
To conclude, Cities and Immigration, as one might expect of its author, is a generous book that includes some inspiring descriptions of cities that have warmly embraced their newcomers. Certainly, leaving cities with plenty of scope for pursuing their own integration policies makes a lot of sense. So does the idea that becoming a city-zen, with the right to take part in city politics, is a good first step to becoming a citizen. I have expressed more scepticism about giving cities powers of immigrant selection. This applies particularly strongly in the case of refugees, not discussed as a separate category in the book, where the costs of hosting the newcomers may for some time be greater than the amount they are able to contribute, making some burden sharing between towns and cities a practical necessity. Cities do have a legitimate interest in the characteristics of the people who come to join themparticularly if they reject the idea of becoming a milkshake. But this has to be set against the overall responsibility of states to promote social justice and a sustainable collective lifestyle within their territories.