Joseph Carens argues that a state should not have the power to deport persons who are members of the society ruled by that state. People will disagree about what social membership means. Some may emphasize a shared national identity, others will consider the extent to which migrants mix socially with the natives. Carens focuses on time of residence: that, he says, is the proper measure of a moral claim to stay. I agree. But states have other concerns and duties that go beyond, and sometimes conflict with, individual moral claims. These other concerns and duties, therefore, need to be taken into account in deciding on the right public policies about irregular migration.

First consider the harms caused by irregular migration. Carens discusses, and broadly dismisses, the harm irregular migration causes to the integrity of the legal system and also the charge of unfairness vis-à-vis would-be regular immigrants who wait in line while others jump the queue. He does not consider the harm caused to the state’s tax revenue and capacity to enforce decent minimum standards of employment, housing, or health. Carens rightly says that non-enforcement of immigration controls does not prove state complicity in irregular migration. Yet democratic states should still be held responsible for failing to regulate the economy in such a way that it does not attract irregular migration. Large irregular migration may significantly damage the state’s efforts to provide social justice and should therefore not be tolerated. This does not imply that irregular migrants must not be regularized, but that governments ought to act preventively by creating conditions that make irregular migration less attractive both for employers and migrant workers. In the European experience, from which Carens suggests Americans can learn, the more comprehensively regulated Scandinavian welfare states attract considerably less irregular migration than the neoliberal labor markets in Britain and Ireland, or the unregulated shadow economies in Mediterranean countries.

Carens’s moral argument provides an answer to the question, “Why regularize?” Public policies, however, must address the additional question of how to regularize. Carens proposes individual, rolling regularization after five to seven years of residence. Yet the U.S. and European governments have instead declared collective amnesties with a specific deadline for registration. Which is preferable?

Periodic amnesties with intervals of five to seven years between them would presumably provide opportunities for everyone who, according to Carens, has an individual claim to regularization, thus solving the moral issue. Collective amnesties, however, have highly problematic side-effects. Processing masses of applications in a short period makes it much more likely that regularization will be over-inclusive of newcomers and under-inclusive of longer-term residents who qualify but cannot file the required documents in time. Where such amnesties occur more or less predictably every four or five years—as they have in Italy since 1986—they clearly create incentive for future irregular migration. In contrast, individual regularization that happens continuously and for smaller numbers offers fewer opportunities for abuse and fewer perverse incentives.

If Carens’s proposal would make better public policy, why do more governments not adopt it? There are two reasons. First, collective amnesties can be a necessary fresh start in response to immigration emergencies. Still, without individual regularization the same problems will build up again. The second, and more essential, reason why governments prefer collective amnesties is an electoral one. Since such amnesties are always coupled with promises of harsher sanctions, they can be sold to pro-immigration as well as anti-immigration voters. Although rolling regularizations make better public policy, politicians will find them much harder to defend.

Behind this dilemma lies a deeper problem—a mismatch between citizens’ moral intuitions about individual migrants, which are well captured in Carens’s argument, and their political views about migration. The same person may support a neighbor or coworker threatened by a deportation order, but vote for politicians who promise to tighten laws against irregular migrants. In my native Austria, even mayors from the two large anti-immigrant parties have opposed the deportation of rejected asylum seekers who had lived in their towns for many years. Carens’s idea that migrants become social members through living the lives of ordinary residents is probably widely shared but does not translate into an inclusive view of the larger society governed by national laws.

Neighborhoods are societies of co-residents bound together through shared experiences of everyday encounters. The political community, however, is not imagined as a society of this kind, but as a community of destiny that must control its borders in order to survive. The United States is, in New School political scientist Aristide Zolberg’s words, “a nation by design” that has generally welcomed immigrants, whereas European “nations by descent” have been much less inclined to see immigrants as future citizens. Despite this difference, voters on both sides of the Atlantic broadly share the view that immigrants have to be invited rather than invite themselves. Overstaying may lead to social membership at the local level, but not political membership at the national level.

This distinction between social and political membership is not an objection to Carens’s proposal but an explanation of the resistance it will face despite its moral and political reasonableness. The problem of how to deal with irregular migrants will stay with us until American and European nations learn to imagine themselves as societies with much more widely open borders. Europe had to do so when on May 1, 2004 large numbers of irregular migrants from Central Eastern Europe were suddenly transformed into European Union citizens with rights of free movement (although not yet free access to employment) in all member states. Nobody called this an “amnesty” then.