Normative Assumptions

I begin by laying out the normative assumptions that frame my inquiry. Whether one endorses the conventional view of migration ethics depends heavily on one’s theory of justice at both the domestic and global levels, a controversial topic much debated in political theory. These debates have focused on 1) the scope, and 2) the nature, of our duties of justice to others. The scope question asks: among whom do duties of justice apply? Three positions have emerged. Statists see the domestic state as the domain within which demanding duties of distributive justice (beyond the protection of basic human rights) apply, including duties to ensure equality of opportunity, equal treatment, and to provide social welfare goods (Blake Reference Blake2001; Nagel Reference Nagel2005; Sangiovanni Reference Sangiovanni2007). Cosmopolitans hold that the same demanding duties of justice that apply domestically should also apply at a global scale (Beitz Reference Beitz1999; Caney Reference Caney2005; Pogge Reference Pogge1989). More recently, some theorists have argued for a “hybrid” view, according to which distinctive and demanding principles of distributive justice apply within the state, but other, somewhat weaker, principles of justice should apply to global social practices, e.g., of international trade or investment (James Reference James2012; Julius Reference Julius2006; Valentini Reference Valentini2011).

A related debate concerns the precise nature of our duties of justice: is justice a matter of the proper distribution of goods and opportunities, or is it in addition, or perhaps instead, concerned with eradicating relationships of power, domination, and inferior social status among people? While distributive theorists hold that justice concerns the distribution of valuable resources within a target population (Arneson Reference Arneson1989, Cohen Reference Cohen1989, Dworkin Reference Dworkin2002), relational theorists hold that justice more fundamentally concerns the dismantling of objectionable social hierarchies (Anderson Reference Anderson1999; Scheffler Reference Scheffler2003; Fourie, Schuppert, and Wallimann-Helmer Reference Fourie, Schuppert and Wallimann-Helmer2015). On the latter view, a just society should be one in which people interact on an equal footing. No one should take precedence over others, dominate or exploit them, or treat them with contempt or disregard. Of course, it may be that people can relate on an equal footing only if goods and opportunities are distributed in certain ways, but many relational theorists see the demand for just distribution as deriving its importance from a broader ideal of human social relations. The distributive and relational approaches to justice are not mutually exclusive—one can combine them—and they each come in statist, cosmopolitan, or hybrid forms, as figure 1 illustrates.

Figure 1

Positions on domestic and global justice

The conventional view of migration ethics sits most naturally with a statist or hybrid account of justice, which sees domestic states as having stronger justice obligations, in at least some areas, to their own citizens than they do to foreigners. I myself favor a hybrid view that incorporates both distributive and relational concerns: while I believe there are distinctively demanding distributive requirements that apply domestically, I also think there are significant requirements of socioeconomic fairness in global investment and trade, including some that regulate relative inequalities between states (Stilz Reference Stilz2019, 15–17). Further, I hold that egalitarian social relations are important in addition to fair material distributions, both domestically and globally. We should eradicate relations of domination, exploitation, and caste-like social stratification, in favor of relationships “that are, in certain crucial respects at least, unstructured by differences of rank, power, or status” (Scheffler Reference Scheffler2005, 17).

Since not everyone shares my assumptions, however, I develop my policy arguments in an ecumenical mode, showing that my conclusions can be supported by other positions in figure 1, including some that reject the more controversial aspects of the conventional view of migration ethics, like priority for the state’s inhabitants. Thus, those who reject my assumptions about global justice may be able to travel a considerable way with my argument. My outreach is not universal: libertarians and market liberals who do not share any commitment to social justice will not be compelled by my arguments.

This article has two main targets. First, I seek to persuade statists, who believe that distributive justice requires priority for the domestic poor, that their concerns are best addressed through strengthening the social safety net, not through restricting immigration. Second, I seek to persuade cosmopolitans and hybrid theorists, who believe we have significant duties of distributive justice to the global poor, not to strictly prioritize more open migration over strengthening redistributive and social schemes in the host society. Doing so may play into the hands of market liberals, who hope to permanently weaken the power of the state to interfere in economic life.

I engage especially in ecumenical outreach to distributive cosmopolitans, who hold that it is morally most important to benefit the globally worst off. Distributive cosmopolitans may be tempted to see increased low-skilled migration as an improvement from a justice perspective, even if it harms some workers in wealthy countries, erodes domestic social rights or labor protections, exacerbates domestic inequality, or creates objectionable social hierarchies, so long as increased low-skilled migration decreases distributive inequalities in the globe as a whole. Still, I argue that even distributive cosmopolitans have reason to support my policy proposals.

Access without Rights: Unauthorized Migrants and Guestworkers

Recall our opening question: is it permissible for the state to use migration restriction as a technique of labor market protection? I start by examining schemes that aim to protect native workers by ensuring that if foreign workers access a host-state’s labor market, they do so without full rights, including the right to free employment, to public services and social benefits, to remain permanently, and be granted citizenship. These restrictions are deemed to benefit domestic workers by ensuring that foreign workers can be hired only in specific sectors, so they do not compete directly with domestic workers, and do not burden the welfare state (Ruhs Reference Ruhs2013, 45). Many economists, concerned about alleviating global poverty, embrace temporary work programs with restricted rights, seeing them as a feasible way to allow as many of the global poor as possible to profit from labor markets in rich countries (Chang Reference Chang2001; Pritchett Reference Pritchett2006; Milanovic Reference Milanovic2016; Rodrik Reference Rodrik2011; for philosophers, see Arneson Reference Arneson2018; Brock Reference Brock2020; Pevnick Reference Pevnick2011).

My first argument is that if foreign workers are granted access to a host country’s labor market, they must be given full rights, including the right to remain on its territory, social protections, and a path to citizenship. No migrant should be forced into temporary or unauthorized status.

Access without rights is the status quo in the United States, which has been characterized for the last thirty to forty years by high levels of unauthorized migration. This creates a large population of foreign-born workers subject to a continual threat of removal, although only a small percentage of them will be deported. As Andrew Elmore puts it, “this informal policy allows U.S. employers to employ a sub-class of millions of workers during times of high labor demand, and permits their exclusion during times of economic anxiety” (Elmore Reference Elmore2006, 558). Many scholars note that the immigration status of unauthorized workers is a powerful tool in the hands of employers, who can use it as a pretext for firing workers who organize unions, demand workplace protections, or agitate for higher wages (Bacon Reference Bacon2008, 6; Lee Reference Lee2019). The first task of any proposal for economic migration to the United States is to reform this exploitative informal labor system.

The most popular alternative to unauthorized migration is the admission of low-skilled workers to the United States as guestworkers. Guestworker programs were proposed in comprehensive immigration reform bills introduced in Congress in 2006 and 2007, and in the 2013 “Gang of Eight” bill that passed the Senate but failed in the House. Proponents of guestworker programs argue that these programs have a good chance of reducing the undocumented population in the United States; that they fit the needs of low-skilled migrants, who often prefer temporary or circular relocation patterns; and that—if well-designed—they could increase the well-being of some of the world’s least well-off people.

Yet like unauthorized migration, guestworker programs face the charge of exploitation (Fine and Ypi Reference Fine and Ypi2016, ch. 8). From 1942–1964, more than 400,000 Mexican farm workers came to work in the United States under the Bracero program, and guestworker programs also existed at the time for Caribbean workers (Hahamovitch Reference Hahamovitch2013). The United States continues to have guestworker programs today, including the H2–A visa program for seasonal agricultural workers, which has increased rapidly in recent years (257,667 visas issued in 2019) and the H2–B program for low-skilled non-agricultural workers (capped at 66,000 visas/year). There is also the H1–B program for skilled workers.

The Bracero program, the Caribbean guestworker program, and the H–2 programs have all been criticized for their abusive labor practices. Indeed, the Southern Poverty Law Center’s report on U.S. temporary labor programs was titled “Close to Slavery.” On paper, these programs were not unattractive: they extended foreign workers many legal protections. Braceros were to be paid “the prevailing wage” and were entitled to housing and food at minimum standards, and safe working conditions (Calavita Reference Calavita2010, 19). The H–2 programs allow for similar protections (“Close to Slavery: Guestworker Programs in the United States” 2013).

In practice, however, these attractive legal protections have gone unenforced, because foreign workers “could not invoke [them] without risk of deportation” (Hahamovitch Reference Hahamovitch2013, 2). Since they operate outside the free labor system, guestworkers can be deported if they lose their jobs. So the standards that are supposed to govern guestworkers’ working conditions, hours, wages, and living accommodations are often disregarded, since employers can fire workers who appeal to them (Calavita Reference Calavita2010, 25). By creating a vulnerable, exploitable temporary workforce, guestworker programs also adversely impact domestic workers, weakening labor protections and lowering wages in key sectors.

I believe that temporary guestworker programs should be abolished (here I repudiate my own earlier view; see Stilz Reference Stilz2010). While this claim is not novel (see Bacon Reference Bacon2008; Lenard and Straehle Reference Lenard and Straehle2012; “Close to Slavery: Guestworker Programs in the United States” 2013), it is contested. Some theorists endorse guestworker programs as a tool for reducing global poverty, arguing that their problems can be ameliorated through reforms (Brock Reference Brock2020; Ruhs Reference Ruhs2013; Pevnick Reference Pevnick2011, 178). They hold that guestworker programs are superior to other development tools (such as foreign aid) because they directly increase migrants’ incomes, rather than benefitting often-corrupt governments. Guestworker programs also benefit the sending society through skill transfer and especially through remittances, which is now the largest source of external finance in many developing countries (Mosley and Singer Reference Mosley and Singer2015).

A common reform proposal (Ruhs Reference Ruhs2013, 174; Brock Reference Brock2020, 161; Pevnick Reference Pevnick2011) is to suggest that guestworker visas should be made portable, so that guestworkers are no longer tied to a single employer, but are allowed to compete openly for jobs within certain pre-specified occupational sectors. This would allow workers who lose their job or who prefer alternative employment the ability to get a new job. Some proposals would grant guestworkers a lengthy period in which to find new employment (Rodriguez Reference Rodriguez2007, 36). Some would afford guestworkers rights to bring their families and to convert to permanent resident status after several years (Motomura Reference Motomura2014, 227).

The main tension in these reform proposals, however, is that in seeking to limit the abuses associated with guestworker programs, the reforms make guestworkers more like permanent immigrants (Lenard and Straehle Reference Lenard and Straehle2012). To count as guestworkers, migrants must have only temporary access to the state’s labor market, and there must be strong incentives for them to return home. Yet the policies necessary to effectively prevent abuse of guestworkers increase their access to the labor market and weaken their incentives to return. These proposals thus undercut one of the main attractions of guestworker programs, from the perspective of employers and the receiving society—that the programs will provide access to foreign low-skilled labor without increasing permanent immigration.

I am not opposed to these reform proposals—in fact, I endorse them, and would go beyond them. I doubt that the abuses inherent in guestworker programs can be corrected through piecemeal reforms. If we are to reliably protect foreign workers against abuse, we must create conditions in which they can claim their legal rights without fear (Lenard and Straehle Reference Lenard and Straehle2012, 219). This depends on recognizing their secure right to remain on the state’s territory. Anything less will continue to facilitate foreign workers’ exploitation, and will further contribute to labor degradation in the host country.

Proponents of temporary labor migration may counter that rather than abolishing guestworker programs, a state could adopt more robust regulatory oversight. But a strong finding from the history of guestworker programs is that regulatory mechanisms have failed to function well: “rather than protect[ing] workers, U.S. officials … served growers’ interests” (Hahamovitch Reference Hahamovitch2013, 3). Plausibly, this is because foreign workers were deprived of all political representation, whereas their employers formed a powerful interest group in domestic politics. This gave U.S. officials little incentive to respond to their concerns. Efforts in 2012 by the U.S. Department of Labor to strengthen wage regulations and labor protections for H2–B workers failed when their hostile employers lobbied members of Congress to prevent the new regulations being put into effect (“Close to Slavery: Guestworker Programs in the United States” 2013, 16). So long as the host-country government is in no way accountable to migrants’ interests, it is unlikely that regulatory reforms will achieve their desired ends.

These reflections support Michael Walzer’s contention that “political justice” requires a path to citizenship for all those who “live within [a state’s] territory, work within the local economy, and are subject to local law” (Walzer Reference Walzer2008, 59). Granting migrants democratic rights allows them to hold officials accountable: without this, host-country protection is likely to remain weak.Footnote ³ Beyond these instrumental concerns, there are also intrinsic reasons to grant migrants a path to citizenship: a permanent class of “denizens” barred from political rights would undermine the social ethos of equality fundamental to a democratic society, characterized by an absence of caste distinctions (Kolodny Reference Kolodny2014).

Proponents of temporary labor migration may further cite the “numbers versus rights” tradeoff: the more rights labor migrants are given, the fewer will be accepted by rich countries (Ruhs Reference Ruhs2013; Brock Reference Brock2020, 148). If foreign laborers waive their rights, this reduces their cost, inducing more rich-country employers to hire them. Overall, the effect is to reduce global poverty to a greater extent than programs that granted full rights-protection would do. Martin Ruhs argues for allowing migrants to waive their rights, pointing to the importance of the workers’ free choice in this matter. The fact that workers voluntarily accept employment even in countries that severely restrict migrants’ rights “can be seen as an indication that many workers are willing to tolerate, at least temporarily, a trade-off between better economic outcomes and fewer rights” (Ruhs Reference Ruhs2013, 128).

Yet one may respond to these proponents by noting that an appeal to workers’ free choice is insufficient to legitimate a bargain when that bargain is made from an objectionable set of options, and the terms of the agreement disproportionately benefit the more powerful party. Indeed, as Ruhs emphasizes, sending countries are afraid of insisting on the rights of their own workers for fear rich-country employers will recruit elsewhere. Migrant consent does not suffice to legitimate an unfair agreement in which rich country employers leverage their bargaining power to impose unreasonable terms.

Still, the numbers versus rights problem may remain, since it does not rest wholly on the significance of migrant consent. We might instead say that giving foreign workers more rights may prove counterproductive, in that a large number of badly-off people will be prevented from accessing employment that could benefit them.

I offer two replies to this concern, one more ecumenical and one less so. In an ecumenical vein, I believe that the views on domestic and global justice surveyed in the normative assumptions section should share in an “overlapping consensus” that guestworker programs are morally problematic. Motivating that consensus seems most difficult with respect to the distributive cosmopolitan, who sees it as morally most important to benefit the globally worst off. A distributive cosmopolitan may see large temporary guestworker programs as the best politically feasible measure to lessen global distributive inequalities (Arneson Reference Arneson2018).

But while guestworker programs will have short-run consequences that strike distributive cosmopolitans as desirable, they must also consider the long-run consequences of such programs for their global justice goals. Guestworker policies not only benefit extremely poor people; they also enhance rich-country employers’ ability to exploit a vulnerable workforce, leading entire occupational sectors to become dependent on this mode of labor procurement, in ways that can be politically difficult to undo. Large guestworker programs will significantly increase the wealth, bargaining power, and political influence of business interests in wealthy countries. These less desirable effects may make future egalitarian reforms more difficult to accomplish, short-circuiting the path to global distributive justice in the long-run. These effects need to be weighed against short-run distributional improvements, and could plausibly outweigh them, motivating even a distributive cosmopolitan to share in an overlapping consensus that rejects guestworker programs.

My less ecumenical response appeals to the thought—characteristic of a “statist” or “hybrid” view of global justice—that greater distributive equality at the global level is not the only thing that matters. Because domestic states have stronger justice obligations, in at least some areas, to their own citizens, we should object to the impact that guestworker programs have on the justice of domestic distributions, by undermining domestic wages, working conditions, and labor rights in key sectors. A destination country has the right to insist on just terms of employment domestically, even if this does not produce a fairer distribution of income and wealth at the global level.

Further, as previously noted, I believe egalitarian social relations are also important, in addition to fair material distributions. We should place considerable weight on the avoidance of hierarchical relations of domination and oppression among those who regularly interact. Plausibly, this give us reason to object to programs that place foreign workers into relationships of exploitation and inferior status.

Thus, guestworker programs can be rejected as morally unacceptable from several different global justice perspectives. If this rejection succeeds, then the terms of economic migration should be politically regulated, and they should include social and labor protections, the right to remain, and a path to citizenship. Even if fewer workers are enabled to migrate, those who do will migrate in fairer conditions.

Open (or More Open) Borders?

Thus, if foreign workers are allowed to migrate, it should be on a permanent basis. Policy options for increasing permanent low-skilled migration lie on a continuum: while the current legal immigration system offers few permanent visas for low-skilled workers,Footnote ⁴ the United States might dramatically increase the number of employment visas offered, without abandoning immigration control. Or, more radically, one might argue for removing restrictions on labor migration entirely, as open borders advocates propose (Caplan Reference Caplan2019; Clemens Reference Clemens2011; Lee Reference Lee2019; Oberman Reference Oberman2015; Van der Vossen and Brennan Reference Van der Vossen and Brennan2018). Gallup, which conducts a world poll asking adults if they would like to move permanently to another country, estimates that under an open borders regime, U.S. population would rise by 46% (Gallup 2018). Should the United States open its borders to permanent low-skilled migration? If so, how open should it be?

The conventional view of migration ethics holds that wealthy states have an obligation to consider migrants’ interests in their immigration policies, which grounds a presumption in favor of openness where migration threatens no significant harms to the state’s constituents. So the question is whether permanent low-skilled migration would cause such harms, and if so, whether these harms could be mitigated by appropriate policies, or prevented only by limiting migration. In what follows, I explore worries about two possible harms. The first is that low-skilled migration may adversely impact less-skilled domestic workers, by lowering wages or increasing unemployment. The second is that it may worsen troubling patterns of domestic inequality, exacerbating relative disparities in ways that threaten egalitarian social and political relations.

As background, I note that there is significant evidence that the United States is severely failing in the duty to ensure a fair distribution of income, wealth, and opportunity for its citizens. Since the mid-1970s, working-class Americans have seen few gains from economic growth. Almost all of these gains have gone to the wealthy, and to a lesser extent, the professional classes, while working-class Americans have seen their wages stagnate (Milanovic Reference Milanovic2016; Deaton Reference Deaton2013). Globalized trade is in part responsible for this inequality, creating winners and losers in the contemporary economy (Autor, Dorn, and Hanson Reference Autor, Dorn and Hanson2016; Acemoglu et al. Reference Acemoglu, Autor, Dorn, Hanson and Price2016). Deindustrialization has caused low-skilled manufacturing sectors of the economy to move abroad (Milkman Reference Milkman2020, 67). This has led to loss of the secure, high-paying jobs once fundamental to the working class. To the extent these jobs have been replaced, it is with insecure, temporary, or part-time service-sector employment. This radical labor market restructuring has disproportionately affected non-college-educated U.S.-born workers (Milkman Reference Milkman2020, 68).

An important driver of this process, beginning in the late 1970s, was a strong business-led effort to reduce the power of organized labor in the United States. Union density is now at an all-time low: 10.3% of American workers belong to a union, a far cry from the 35% peak unionization rates of the 1950s. Private-sector unions are nearly dead, with a particularly low membership rate of 6.2%. Falling levels of unionization have been a driver of stagnant wages (Massey Reference Massey2007, 164; Milkman Reference Milkman2020, 74).

Economic inequality has been exacerbated by lower corporate, income, and inheritance taxes and welfare state retrenchment—the product of Republican tax-cutting agendas and the adoption, by Democrats, of “Third Way” policies. The result is a level of economic inequality in the United States that is higher than at any time since the 1920s (Piketty and Saez Reference Piketty and Saez2003). The domestic working class also suffers from a lack of political representation: mainstream political parties have been disinclined to respond to their concerns (Gilens and Page Reference Gilens and Page2014).

Since immigration grew dramatically during this period of rising inequality, one might see a correlation between the two phenomena. Indeed, studies show that low-skilled workers are less favorable to an open immigration policy than high-skilled workers are (Scheve and Slaughter Reference Scheve and Slaughter2001; Mayda Reference Mayda2006; Hainmueller and Hiscox Reference Hainmueller and Hiscox2010). But there is little evidence that immigration has been a primary driver of these developments (National Academies of Sciences 2016). David Card estimates that immigration can account for a relatively small share (4%–6%) of the rise in wage inequality in the United States over the past twenty-five years (Card Reference Card2009). Qualitative work by sociologist Ruth Milkman agrees that while there is a correlation between inequality and immigration, the relationship is not causal (Milkman Reference Milkman2020). She holds that it was the economic reforms of the 1980s that “degraded” many low-skilled U.S. jobs, causing native-born workers to leave these positions, after which economic migrants were recruited to take them.

Still, there are concerns about how a liberalized migration policy might interact with the troubling scenario already faced by domestic workers. Would it further exacerbate their economic precarity? The wage and employment effects of immigration are hotly disputed among economists. There is a consensus that immigration confers net economic benefits on the United States as a whole. But it may also create winners and losers, with losers perhaps concentrated among low-skilled native workers. As a matter of economic theory, how immigration impacts these workers depends on whether low-skilled immigrants are substitutes or complements to native workers. If immigrants are perfect substitutes, they are likely to displace native workers and lower their wages (National Academies of Sciences 2016, 2). If immigrants are complements, with different skills than native workers, they might instead improve low-skilled natives’ job prospects.

This question has been much debated in recent years, and many economists believe there is a significant degree of complementarity between low-skilled migrants and natives (Ottaviano and Peri Reference Ottaviano and Peri2012). The intuitive idea is that low-skilled native workers’ better communication abilities afford them access to different jobs, to which they reallocate their skills. Peri and Sparber argue that “immigrants have a comparative advantage in occupations requiring manual labor and tasks, while less educated native-born workers have an advantage in jobs demanding communication skills” (Peri and Sparber Reference Peri and Sparber2009, 136). Thus, when foreign-born construction workers, say, flow into the market, native-born workers can move to being, e.g., construction supervisors, as additional positions are created to serve the needs of the expanded workforce (National Academies of Sciences 2016, 198). On this view, new immigrants tend to compete mainly with previous cohorts of immigrants, and any negative effect they have on wages and working conditions is largely to the detriment of immigrants in similar circumstances (Ottaviano and Peri Reference Ottaviano and Peri2012; Abizadeh, Pandey, and Abizadeh Reference Abizadeh, Pandey and Abizadeh2015). If low-skilled natives and immigrants are imperfect substitutes or complements, then increased low-skilled migration is likely to have much less—perhaps even no—impact on the wages of native-born unskilled workers.

Lending some credence to this theory, empirical work on local labor markets has generally failed to find that immigration causes large negative wage or employment effects for native workers. While these studies show a range of effects, the effects are nearly always small. Card estimates that immigration from 1985–1990 reduced real wages by 1%–3% in low-skilled occupations in U.S. metropolitan areas with the highest migration levels, with the largest impacts on previous cohorts of immigrants (Card Reference Card2001). Other studies have found zero or even slightly positive wage impacts.

Further studies have leveraged natural experiments. Well-known research by David Card conducted after the Mariel boatlift to Miami concluded that the influx of migrants did not significantly lower wages or increase unemployment for Miamians (Card Reference Card1990). That case is continuing to be litigated in the empirical literature.Footnote ⁵ Some natural experiment studies find more negative results: Monras investigates a surge in immigration related to the Mexican peso crisis in 1994–1995, and finds that a 1% increase in labor supply due to immigration reduced the wages of native non-Hispanic workers by 0.7% in the short run (Monras Reference Monras2020).

Beyond regional studies and natural experiments, economists have taken two other approaches to estimating wage impacts. The “skill cell” method groups workers by skill or occupation in an attempt to isolate impacts on native workers who are close substitutes for migrants. This method tends to find more sizable negative effects, but may not adequately capture complementarity (Bansak, Simpson, and Zavodny Reference Bansak, Simpson and Zavodny2020, 182). Finally, some researchers have used production functions to simulate the overall impact of immigration on the economy, with estimates that range from somewhat negative impacts on the least skilled workers to slightly positive impacts, depending on assumptions about substitutability and complementarity and division of skill groups (Borjas Reference Borjas2003; Reference Borjas2014; Ottaviano and Peri Reference Ottaviano and Peri2012; National Academies of Sciences 2016).

On balance, the empirical studies support the view that existing immigration flows have had, at most, a modest negative impact on the least-skilled native workers. To the extent they exist, these impacts appear to be concentrated on particular groups—high school dropouts, low-skilled African-American workers, and previous cohorts of low-skilled immigrants (National Academies of Sciences 2016, 241).

There are reasons for caution in interpreting these results. First, though the bulk of the evidence suggests that negative wage impacts are modest, that does not mean they are non-existent. Given fifty years of stagnant wages in the United States, even a small decline in wages for the most vulnerable native workers might have a dramatic impact on their welfare. For example, a recent study shows that just 39% of Americans have enough savings to cover a $1,000 emergency expense (Konish Reference Konish2021). For a family of four just above the U.S. poverty line of $26,000, a 2.5% wage loss would be $650, an amount that would make a significant difference to the family’s security.Footnote ⁶

Second, since current levels of immigrant admissions are highly restricted, it is unclear how evidence about current wage impacts—such as those shown in the studies considered here—bears on our assessment of much more open borders. The National Academy of Sciences provides a helpful cross-study comparison, which shows that estimates of the wage effects of an immigrant inflow that increases labor supply by 1 percent range from a –1.7% decline in the least skilled workers’ wages to a 0.3% increase (National Academies of Sciences 2016). Suppose that an open borders policy were to create migration flows on the order of the last global era of migration, which raised the New World labor force by one-third (Hatton and Williamson Reference Hatton and Williamson1998, 3). Given the range of current empirical estimates, and assuming that effects are linear, this could lead to anything from a –56.1% drop in the least-skilled workers’ wages to a 9.9% increase. That is a very large window!

Finally, economic historians who work on the age of migration, such as Timothy Hatton and Jeffrey Williamson, and Claudia Goldin, estimate that unrestricted migration in that period did lower wages for the least-skilled, with the biggest impact on those places which received the most migrants (Goldin Reference Goldin, Goldin and Liebcap1994, 225; Hatton and Williamson Reference Hatton and Williamson1998, 113).

These reasons for caution do not, in my view, undercut the case for a more open policy toward low-skilled labor, so much as they tell in favor of taking a measured approach to it. I argue that the state should liberalize low-skilled migration gradually, pairing it with policies to compensate low-skilled native workers for possible negative impacts, and standing ready to re-impose migration restrictions, at least temporarily, should significant economic harms eventuate. In particular, possible adverse effects on low-skilled workers should be addressed through changes to the social policies that accompany increased migration. All economists agree that open migration will produce domestic gains that more than suffice to offset any losses to low-skilled native workers (Bansak, Simpson, and Zavodny Reference Bansak, Simpson and Zavodny2020, 12; Borjas Reference Borjas2014, 6–7). Absent intervention, these gains will accrue principally to the capital-owning and professional classes, who disproportionately benefit from the cheaper labor that migrants provide. As it opens its borders, then, the state should tax away a significant portion of these elite gains and transfer them to its working class in the form of enhanced social rights and opportunities, designed to cushion them against possible losses.

Why can the state be asked to restructure itself in this way? After all, reforms are costly. Yet we must ask whether the state has any legitimate interest in maintaining status quo distributive policies. The evidence of inequality and precarity presented earlier suggests the status quo is unjust: the costs of institutional change are costs the state is already morally required to bear. Further, whether there will be “trade-offs” between the interests of migrants and the native working class is entirely an artifact of the particular mix of distributive policies the state adopts. If the state structures its policies to ensure the domestic gains from migration are fairly distributed, no harms to native workers need occur. The state’s duty to give some weight to migrants’ interests gives it additional reason to favor this mix of policies.

My approach assumes that, in principle, the harms caused by economic competition could provide a reason to compensate low-skilled workers. Some contest that assumption (Huemer Reference Huemer2010, 438). They say that domestic low-skilled workers have no legitimate claim to their jobs or level of wages. An interest in not suffering so-called “competition harms” is not a legitimate moral interest: “society admits no right, either legal or moral, in the disappointed competitors to immunity from this kind of suffering” (Mill Reference Mill2008).

Consider the following case:

Van der Vossen and Brennan argue that even if people like George suffer as a result of Jorge’s arrival, this simply shows that George was “in effect enjoying an economic rent from closed borders—[his] income was artificially inflated compared to what the market otherwise would have supported” (Van der Vossen and Brennan Reference Van der Vossen and Brennan2018, 34).

Is it true that people lack any moral complaint about competition harms? I believe this is so only in special scenarios: namely, when those harms occur within an institutional framework that ensures that the fundamental interests of those who lose out from competition are well protected, and they have access to a range of other options and opportunities to secure a good life for themselves. In this scenario, the harmed individual’s complaint is considerably diminished. Though the loss of a job may cause her disruption, that disruption is unlikely to greatly diminish her lifetime prospects. Under these conditions, temporary displacement is a reasonable burden to ask people to bear for the sake of economic innovation and lower consumer prices.

When such background institutions are absent or defective, I think individuals do have a moral complaint about suffering competition harms. Even if greater economic competition, on net, benefits society as a whole, those aggregate benefits may not outweigh concentrated harms to certain groups. Moreover, given the facts about economic precarity outlined earlier, the current social framework in the United States seems insufficiently robust to preclude all complaint about competition harms. The economic losers could reasonably reject an open immigration policy that caused competition harms to them, unless it was paired with guarantees for their social and economic opportunity rights that would sufficiently mitigate these harms.

This baseline for assessing morally relevant harm invokes a broader theory of the just social distribution of benefits and burdens (Barry Reference Barry2014). As canvassed in the section on normative assumptions, there are a range of views on what would constitute a just distribution of benefits and burdens at the domestic and global level. Yet in an ecumenical vein, I argue that none of the views in figure 1 should favor open borders on laissez-faire terms if redistributive policies are available that would mitigate the negative impacts of open migration, should there be such, on the working class of rich countries. Adherents of statist or hybrid views like my own, which place weight on egalitarian distributions within the state, will clearly have reason to pair open migration with stronger social and economic protections for domestic workers.

It may however seem that a distributive cosmopolitan should disagree, holding that migration policy should place greatest weight on benefiting the globally least-advantaged. These are more likely to be foreign workers than domestic ones, and open borders could significantly improve their position, achieving an important cosmopolitan goal. Yet distributive cosmopolitans should consider: who should justly bear the burdens of benefiting the globally least-advantaged? Supposing open migration were to create significant losses, should those burdens lie where they would initially fall, on the domestic working class of rich countries, or should they be shifted towards their domestic elites? These elites are already better off—and so more capable of shouldering such burdens—and they can be expected to gain even further from open migration. Surely a distributive cosmopolitan ought to prefer policies that would require these elites to redistribute their gains to ensure that protections are provided to their domestic working classes: this would lead to a more egalitarian distribution overall. Thus, theorists of many different stripes will have reason to support pairing more open borders with the deepening of domestic labor and social rights and economic redistribution.

Social Hierarchy and Caste Stigmas

Let me turn to a second concern about a more open labor migration policy: it may worsen troubling patterns of domestic inequality. Beyond possible absolute effects on native wages and employment, open migration can be expected to contribute to rising income gaps between the upper and lower classes in the host state (Card Reference Card2009). This is because the wages, wealth, and education of low-skilled immigrants will be lower than that of natives. From a distributive cosmopolitan perspective, this increase in domestic inequality might seem an improvement. After all, even if they are on the lowest rungs of the social ladder in the new country, immigrants’ incomes and prospects will be higher than they would be had they stayed home, decreasing global inequality overall (Milanovic Reference Milanovic2016). But from a relational perspective, this distributional improvement may be accompanied by an important worsening. While “location-based” differences between people in different countries have decreased, “class-based” inequalities between people who share the same country have increased (Milanovic Reference Milanovic2016). The worry is that this may lead to objectionably hierarchical relations between people in the host society: relations of domination, exploitation, or caste-like social stratification.

Is this a matter of principled moral concern? A standard way of answering “no” is to suggest that it is morally arbitrary that an inequality between A and B would be less troubling when A and B are separated by a border than when they share the same society. From a distributive cosmopolitan perspective, this is indeed baffling. But from a relational perspective, it is less so. Relational theorists emphasize that a just society is one in which ongoing social relations have an egalitarian, nonhierarchical character. The concern is that when significant material inequalities become embedded in a context of dense social interactions, as they often are in a domestic state, they may give rise to objectionable social hierarchies.

As noted earlier, the relational theorist need not take the view that global inequalities are unobjectionable. There might be independent reasons, sourced in a theory of just distribution, to object to them (thus, relational considerations can be combined with distributive concerns in a “pluralist” account of justice). In addition, as already noted, there are clearly many cross-border relationships that raise relational worries: structural relationships in the global economy frequently foster exploitation or oppression. Plausibly, relational principles will require mitigating these global inequalities as well as domestic ones. All the relational theorist need claim, for my purposes here, is that significant material inequalities are especially likely to be concerning when they occur within a domestic state.

To put the point sharply, consider Michael Piore’s comment that on a conventional economic model of open migration, “every American family should have a foreign maid” (Piore Reference Piore1980, 30). There is clearly a difference in the American family’s relationship with their foreign maid, compared to their relationships with foreigners beyond borders in general. The foreign maid stands in personal dependence upon this family, and that relation is likely to be characterized by significant asymmetries of power. If not regulated, relations of this kind may easily lead to oppression, or give rise to attitudes of superciliousness and servility. Since the extent to which a material inequality can be expected to create a hierarchy of superiority and inferiority between A and B varies with the density of the social interactions that the two sustain, a relational theorist can explain why an inequality is often more troubling when it occurs within a domestic society. This is because that inequality is likely to become embedded in patterns of interaction that frequently give rise to negative shared expectations about how subordinate classes of people are to be treated.

An important worry about open migration for the low-skilled, then, is that if not paired with appropriate social protections, it could lead to the formation of an underclass. A well-known fact about low-skilled migrants is that they tend to work in a distinct set of jobs, often jobs that the native labor force refuses to accept (Piore Reference Piore1980, 3). In the U.S. context, Milkman calls these “brown collar jobs”: low-skilled migration, she argues, leads to a “pattern of occupational segregation [which] involves the overrepresentation of foreign born workers in poorly paid, physically demanding, menial, and often dangerous jobs” (Milkman Reference Milkman2020, 20). Indeed, it is this pattern of occupational segregation that explains the phenomenon of economic complementarity considered earlier. It is because immigrants are occupationally segregated that they compete only to a limited extent with native workers. As Piore observes, a feature of immigrant jobs is that they carry low social status: “they often involve hard or unpleasant working conditions and considerable insecurity, they seldom offer chances of advancement toward better-paying, more attractive job opportunities, they are usually performed in an unstructured work environment, and involve an informal, highly personalistic relationship between supervisor and subordinate” (Piore Reference Piore1980, 17). As he puts it, these are the “bottom positions in the social hierarchy.”

A relational theorist should be concerned that unless it is offset by egalitarian background institutions, occupational segregation may facilitate exploitation and caste-like social stratification. This is especially likely to occur when low-status jobs in a social hierarchy are correlated with observable, differential traits, such as race, ethnicity, or language. Racial and caste prejudices often gather around economic distinctions that track socially recognizable traits. Martin Delany, the nineteenth-century African American theorist, notes that:

Delany holds that persistent economic inequality, particularly where it tracks membership in a recognized social group, often leads to social and racial stigma:

When a recognized group persistently occupies the position of being another group’s servants, a degraded social status is commonly attributed to them, with accompanying negative expectations about how they are to be treated. They are less likely to be considered social equals, accepted into friendships or other social associations, and they may suffer from stereotypes of inferiority, which can further justify their exclusion from desirable jobs and opportunities.Footnote ⁷ Indeed, Delany interprets the ideology of racial inferiority as a post hoc justification for established practices of economic exploitation (Delany Reference Delany2014, 5–6).

Of course, we also have reason to object to relational harms at the global level. Still, even if open migration on laissez-faire terms might enable workers to exit exploitative relations abroad, it remains unclear why replacing these relations with more proximate exploitation combined with caste-like social stratification should be considered a overall gain from a relational perspective. If greater openness to low-skilled migration is not to create relational harms in the host society, it is especially important that it occur within an egalitarian basic structure. Robust labor guarantees are necessary to secure foreign workers against exploitation due to dependence on their native employers. And robust redistributive and social guarantees are necessary to ensure social mobility for immigrant children, to guard against the formation of caste stigmas towards them. As Delany puts it, the best remedy for social inferiority is “that some colored men and women, in like proportion to the whites, should be qualified in all the attainments possessed by them” (Delany Reference Delany2014, 12).

One might worry that social mobility for immigrant children could exacerbate wage and employment impacts, making it more likely that they are substitutes, rather than complements, for native workers. Yet this concern is misplaced, since intergenerational social mobility occurs over a long time horizon, and it replicates the skill mix of the pre-existing population, rather than changing it. Immigration is most likely to have adverse effects on native workers when, in the short run, it shifts the relative mix of skills in the economy: i.e., when immigration suddenly increases the unskilled workforce without increasing the skilled workforce and capital stock by a similar amount. Since equality of opportunity policies are designed to ensure that the children of immigrants are qualified in proportions similar to the rest of the population, it does not affect the relative skill mix. Instead, these policies simply enlarge the overall economy, just as natural population growth does (Lewis Reference Lewis2017; National Academies of Sciences 2016, 147). And because equality of opportunity policies take effect over the long run, the capital stock can adjust: more businesses open to serve a larger population, and provided skill ratios stay the same, wages should not be affected (Borjas Reference Borjas2014, 65).

Defenders of open migration who agree that egalitarian social relations are important should be concerned to combat social hierarchy and caste stigmas. Recall that on the conventional view of migration ethics, migration policy should take both the state's constituents and immigrants' interests into account, though it should give some priority to the interests of constituents over immigrants in cases of conflict. Durable social hierarchies harm both citizens and immigrants, by undermining egalitarian relations in the host society, and stigmatizing immigrants at the bottom of these hierarchies. Thus, concerns about relative disparities, like the concern about negative absolute impacts for domestic workers, support pairing a liberalized migration policy with economic and social guarantees of an ambitious kind.

Redistributing the Gains from Migration

The main theme of my discussion has been that it would be a mistake to divorce a concern for a more open migration policy from a concern with social and labor rights and redistribution within the host country. To put the point sharply, I worry that many cosmopolitans are tempted to make this mistake. Open borders is favored by an odd coalition of bedfellows: market liberals and libertarians, on the one hand, and cosmopolitans, on the other. Cosmopolitans’ willingness to move towards libertarian positions is reflected in their critique of the domestic state, and in their embrace of freedom-based arguments for open borders.Footnote ⁸

To my mind, the persuasive case for (more) open borders appeals not to a fundamental right to freedom, but rather to the egalitarian idea that relaxing border controls might help reduce global inequalities. Yet whether open borders will reduce global class disparities is more fraught than it seems: openness is unlikely to achieve egalitarian ends if it proceeds on laissez-faire terms. In a laissez-faire scenario, domestic employers will be the principal beneficiaries of the domestic immigration surplus, in the form of higher profits due to cheaper labor costs. A larger pool of low-skilled labor will also give employers more bargaining power over wages and the terms of work. This increase in rich country elites’ income and wealth will predictably strengthen their political and social power, both domestically and around the globe.

Cosmopolitans should be concerned about these power disparities. Given this, it may be misguided to grant strict priority to economic openness over strengthening host countries’ social and redistributive schemes. Doing so may favor market liberal outcomes. Along with advocating more open migration, a cosmopolitan should seek to limit and redistribute the gains that rich-country elites can be expected to derive from it. They should support pairing greater openness with redistributive and social reforms in the host country.

More controversially, I believe that statist or hybrid theorists should support something stronger, at least in the U.S. context: other things equal, liberalization of low-skilled migration should be conditioned on the adoption of stronger social and labor rights and greater egalitarian redistribution. Conditioning should be conceived on a sliding scale: given the economic evidence canvassed earlier, some modest liberalization of low-skilled migration seems acceptable even absent ambitious social reforms. But the larger the potential harms associated with open migration, the more demanding the social protections that need to be in place before that degree of liberalization is permitted.Footnote ⁹ This conditioning argument relies on controversial premises, including a willingness to grant some degree of priority to the state’s inhabitants vis à vis foreigners, and to place significant weight on relational egalitarian concerns. Not all the views canvassed in figure 1 will go so far. Still, I hope that even distributive cosmopolitans might support the weaker claim that open borders should be paired with domestic social reforms. Egalitarians of many stripes should be pushing for both, even if not all will wish to condition the former on the latter.

What egalitarian policies might be needed? First, liberalization of low-skilled migration should be accompanied by a major government investment in education. In response to globalized trade, Ken Scheve and Matt Slaughter recommend building a “lifelong ladder of opportunity” to enable rich-country citizens to adapt to competitive pressures. In their view, this should include 1) state-subsidized early childhood education for every U.S. child, from birth to kindergarten; 2) guaranteed federal funding for higher education; and 3) for those without a college degree, a special “training” credit of $10,000 every decade that might be used to gain skills in their occupational category throughout their lifetimes (Scheve and Slaughter Reference Scheve2018). If undertaken, this would represent the largest government investment in human capital formation in U.S. history.

In terms of our previous discussion, this education policy would have two desirable effects. First, it would enhance complementarity for the native-born workforce, cushioning them against possible competition harms from open immigration. The idea is that as the pool of foreign-born low-skilled laborers grows, native workers can move to occupying different jobs, which often require communication, quantitative, or supervisory skills. Second, educational investment would facilitate equality of opportunity for the children of immigrants. This is necessary to avoid their being “trapped” in the low-skilled jobs of their parents, in ways that will solidify persistent social hierarchies. If children of immigrant groups have equal or near-equal chances with natives of attaining professional or managerial positions, group-based stereotypes of inferiority and degraded social status are less likely to develop.

While educational investment is important, it is not enough. Some workers lack the raw capacities—communication skills, quantitative abilities, comfort with technology, social facility, and managerial qualifications—necessary to occupy higher-skilled roles. What to do for them? Educational expenditures could be accompanied by a government guarantee to act as an employer of last resort (Atkinson Reference Atkinson2015, 141–43). Such a program might guarantee employment at a certain number of hours a week in infrastructure development, childcare and eldercare services, and other sectors where investment is needed (Atkinson Reference Atkinson2015). This would be a permanent, ongoing measure designed to protect domestic workers (including minorities and children of immigrants) against the pressures of global openness, guaranteeing them stable employment and a productive role in society. Once naturalized, first-generation immigrants would also become eligible for guaranteed employment (currently the waiting period for long-term permanent residents to naturalize is five years).

Finally, these policies need to be accompanied by labor market reforms. Raising the minimum wage and extending unemployment protections are essential to ensure that immigrant workers are not overly dependent on their native employers for money. When immigrant workers teeter on the edge of impoverishment, this enhances their employers’ power, facilitating their exploitation. Changes to the minimum wage should be combined with reforms that strengthen the power of unions in American life, increasing the bargaining power of low-status workers. Twenty-eight states now have right-to-work laws that enable workers to opt out of union dues paying even while gaining the benefits of union representation, making it difficult to organize new unions or maintain membership in unions that exist. Union-enhancing reforms should include the repeal of right-to-work laws, measures that punish private employers for union avoidance tactics, and the revitalization of the strike.

One might raise a concern about the cost of these reforms: would they come at the expense of other social programs? To a significant extent, the proposed reforms need not derogate from other programs: instead, they can be funded by the domestic gains that will accrue from more open migration. As noted, absent intervention, those gains will fall primarily to domestic elites: open migration can be expected to increase firms’ profits, by lowering their labor costs; and it will reduce the price of immigrant services, making it cheaper for wealthy professionals, say, to have their houses cleaned and lawns mowed. Funding reform programs through higher corporate, income, and wealth taxes will target those people most likely to reap efficiency gains, lessening the impact on other social priorities. This also makes domestic social reforms a more cost-effective way of addressing global inequality than proposals that do not target efficiency gains, like simply taxing the domestic rich and redistributing the proceeds to the world’s most needy.

Of course, a critic may object that such redistributive and social reforms are unlikely to happen anytime soon. What we may face in reality is a “second-best” choice between open and closed borders. I am not convinced that these proposals are infeasible: support is increasing in the United States for enhanced redistributive measures, and many in the Democratic Party defend policies not unlike the ones outlined earlier. In any case, it is not obvious that enhanced social protections are less politically feasible than much more open borders, currently not embraced by any mainstream party.

Still, suppose the critic is correct: in practice, we can expect that, if adopted, more open migration will have to proceed on laissez-faire terms, for lack of political will. Should borders be opened anyway?

Many economists argue for separating open immigration policies from redistribution: redistributive concerns do not, in their view, justify migration restriction. Rather, we should first liberalize migration and only then attempt to implement redistributive transfers (Pritchett Reference Pritchett2006, 95; Chang Reference Chang2008).

I disagree with this approach: by allowing free movement without accompanying egalitarian reforms, we risk cementing market liberal outcomes. Theorists who share a commitment to social justice should instead prioritize the robust protection of the social, labor, and economic rights of all workers, both immigrant and native-born. To achieve that, we need to place as much, if not more, weight on strengthening existing social policies as on increasing global openness. Faced with a policy choice between open and closed borders, we ought to reframe the agenda to include the option of pairing more open borders with enhanced redistribution.