Immigration, Backlash, and Democracy

Abstract

How do considerations related to backlash affect the desirability of pursuing otherwise justified immigration policies? This paper argues that backlash-related considerations bear on immigration policy decisions in ways that are both more powerful and complicated than typically recognized. The standard possibility, the egalitarian backlash argument, endorses immigration restrictions in order to protect support for egalitarian distributive institutions. The paper shows that this account does not, by itself, provide a convincing rationale for restricting immigration because such diminished support is (a) likely outweighed by the benefits of more permissive immigration policies and (b) caused by the objectionable preferences of citizens. However, the paper develops an alternative account of the relevance of backlash-related considerations, the democratic backlash argument, which holds that increased levels of immigration threaten to contribute to undermining democratic institutions. This argument provides a more powerful rationale for restricting immigration, one that can—under identified conditions—justify immigration restrictions.

Much of the normative literature on immigration argues for policies that are far more permissive than the status quo. The goal of the present paper is to ask: have advocates of such policies given adequate consideration to the backlash that these policies are likely to provoke if enacted in the developed democracies? Think of backlash as action that imposes social costs and is taken, consciously or not, in response to a disliked policy. With respect to immigration policy, backlash may stem from, among other things, sincere disagreement about what justice requires, xenophobia, ethnic antagonism, and/or self-interest. This paper examines considerations related to backlash to see whether, when, and how they might affect the desirability of pursuing otherwise justified immigration policies. It is an attempt to engage with opponents of more permissive immigration policies by charitably reconstructing, and then exploring the force of their backlash-related concerns.

The relevance of such concerns depends on the type of immigration restrictions that can be independently justified. Accordingly, the discussion separately explores the force of considerations related to backlash assuming (1) that open borders are otherwise justified and (2) that states have a right of self-determination that creates a permission to restrict immigration. This division of the argument is useful because the threshold that considerations related to backlash would need to clear to be relevant is different in the two cases. If open borders arguments are otherwise correct, then backlash-related considerations would need to show that the costs to citizens of immigration exceed the benefits to potential immigrants. By contrast, if states have a right of self-determination that includes a permission to restrict immigration, then backlash-related considerations would only need to show that citizens have an important interest in restricting immigration, not necessarily one that—by itself—outweighs the interest that potential immigrants have in entering. Since I cannot settle the background question, I separately explore the implications of backlash-related concerns in each case.

The first two sections of the paper, then, proceed under the assumption that, in a world of full compliance, restrictions on immigration would not be justified (i.e., that the required policy would be open borders). I begin with the most commonly discussed possibility related to backlash. This egalitarian backlash argument holds that we should restrict immigration because otherwise support for egalitarian distributive institutions that are themselves required by justice will substantially dissipate. Even assuming that the costs that this argument describes will come to pass, I contend that this argument does not provide a convincing rationale for restricting immigration because (1) it does not explain why such costs outweigh the benefits of more permissive immigration policies and (2) the relevant costs only arise because of the objectionable behavior of citizens. The rest of the paper explores whether there are backlash-related arguments that do not suffer from these two shortcomings.

The paper develops an alternative possibility, which I call the democratic backlash argument, according to which large increases in immigration pose a threat to the stability of democratic institutions and should, for that reason, be avoided. I contend that this argument provides a more powerful pragmatic rationale for restricting immigration because it identifies costs of permissive immigration policies that are more significant than those at stake in the egalitarian backlash argument. The implication is not that borders should be fully closed or that the status quo is justified, but just that—given realistic behavioral assumptions—there are pragmatic considerations related to sustaining democracy that speak in favor of restrictions, even given the assumption that open borders are otherwise justified.

The paper’s final section assumes, instead, that states have a right of self-determination that includes a permission to restrict immigration. It then shows that it is, in principle, possible for backlash-related considerations to contribute to a justification of—and not just a pragmatic rationale for—immigration restrictions. The broad implication is that backlash-related considerations bear on the immigration policy that we ought to accept in ways that are more powerful, and more complicated, than typically recognized.

THE EGALITARIAN BACKLASH ARGUMENT

It is commonly argued that there is a tension between satisfying egalitarian principles of distributive justice within a political community and permissive immigration policies. Indeed, a wealth of empirical evidence supports the claim that increased levels of immigration strongly reduce citizens’ support for egalitarian (or redistributive) measures (e.g., Alesina, Miano, and Stantcheva 2018; Alesina, Murard, and Rapoport 2021; Dahlberg, Edmark, and Lundqvist 2012; Tabellini 2020). This appears to be a robust finding—one that has been established in many studies, with different social scientific tools, and in many different receiving countries. ¹

Partly as a result of such evidence, many have been tempted by the following kind of argument:

P1. A just society requires an egalitarian basic structure.

P2. Citizens will only be willing to support an egalitarian basic structure if they share a common national identity with their fellow citizens.

P3. Immigration disrupts the shared national identity of citizens, undermining support for an egalitarian basic structure.

Conclusion: Justice requires the restriction of immigration.

This is an egalitarian backlash argument. It suggests that, in practice, there is a trade-off between two goals, both of which are typically embraced by liberal egalitarians: permissive immigration policies and an egalitarian basic structure. The implication is that those who are unwilling to abandon an egalitarian basic structure (i.e., those who accept P1) must concede to the backlash by accepting restrictive immigration policies (for examples of positions in this vicinity, see Gibney 1999, 174; Miller 1995; 2005). ²

The rest of the section evaluates this argument. The next subsection explores the truth of the empirical premises (P2 and P3) and presents a reformulation of the argument that is more consistent with existing evidence. The following subsection contends that the reformulated argument fails to provide adequate reason to adopt restrictive immigration policies.

The Empirical Claims

I begin by examining the potential points of empirical vulnerability—P2 and P3.

P2 claims that citizens with a strong national identity will be more supportive of egalitarian institutions. The immediate problem for P2 is that there is abundant cross-country empirical evidence that conflicts with it. As Shayo (2009, 158) writes, “people who identify more strongly with their nation prefer a lower level of redistribution than people with low levels of identification and similar income.” This is a robust relationship, which holds in almost all of the major receiving countries. Moreover, the effect is very strong relative to income—moving from very weak to very strong identification with one’s nation is “equivalent in terms of attitudes toward redistribution, to having one’s household income multiplied by a factor of between 1.5 and 3 in most western democracies” (Shayo 2009, 160). One explanation for this relationship is that, for individuals who identify primarily with the nation, the relative appeal of redistributive programs is relatively low compared with their appeal for those who identify more strongly with their social class (Shayo 2009; 2020). As Shayo puts it, “identifying with the lower class increases support for redistribution, whereas identifying with the nation tends to reduce it” (Shayo 2009, 168).

If the preceding is correct, the egalitarian backlash argument needs to be recast in roughly the following terms:

P1. A just society requires an egalitarian basic structure.

P2’. The working class will only be able to effectively demand an egalitarian basic structure if they are united.

P3’. Immigration allows for the disruption of the shared class identity of working-class citizens.

Conclusion: Justice requires the restriction of immigration.

Whereas traditional formulations revolve around liberal nationalist commitments, this revisionist account is more consistent with positions, such as those in the socialist tradition, that focus on class conflict’s egalitarian potential. This theoretical account is consistent with the argument that high levels of immigration prevented the United States from developing the same level of class consciousness, and therefore support for redistribution, that existed in many European countries with lower levels of immigration (e.g., Lipset and Marks 2000).

While the first two premises are straightforward, the third may require some motivation. The following mechanism may account for P3’. Globalization has caused economic anxiety and rising inequality that has no obviously sufficient policy response since they are driven, in important part, by technological conditions, as well as an increasingly globalized labor market, both of which are at least somewhat beyond the control of national governments. ³ Where immigration levels and discontent with such consequences of globalization are simultaneously high, enterprising politicians have a tempting opportunity to gain politically by running on an anti-immigration platform (Abramowitz and McCoy 2019; Rodrik 2018, 24). Populists seize on economic anxiety to build a cleavage (often, rooted in ethnic antagonism) between native and non-native working-class citizens. This is one explanation of how immigration, under present conditions, leads to a disruption of the shared identity of working-class citizens and thereby interferes with their effectively demanding an egalitarian basic structure (Alesina, Murard, and Rapoport 2021). ⁴

The Argument’s Normative Force

Having reformulated the egalitarian backlash argument so that it better fits the existing evidence, I now identify its two critical shortcomings. First, the egalitarian backlash argument identifies a relevant cost of immigration—it putatively weakens support for an egalitarian basic structure. However, since those from poor countries have a tremendous amount to gain from accessing the productive economies of wealthy states (see, e.g., Clemens 2011), it is not obvious that the bad consequences associated with a less robustly egalitarian basic structure are sufficient to justify restrictions on immigration. Differently put: even if the egalitarian backlash argument is correct that a robust egalitarian basic structure and permissive immigration policy are incompatible, the argument fails—at least, on its own—to provide reason to resolve the conflict in favor of an egalitarian basic structure (Pevnick 2009; similarly, Holtug 2022, chap. 6). Thus, even granting the somewhat controversial empirical premises on which it depends, the egalitarian backlash argument fails to provide even a convincing pragmatic rationale for restricting immigration.

There are, to be sure, other kinds of arguments in the literature that may be able to fill this gap. For instance, many have argued that there are reasons for states to weigh the interests of their own citizens more heavily than the interests of foreigners (e.g., Rawls 1999). It is true that one could invoke such arguments in order to address the first critical shortcoming of the egalitarian backlash argument. The difficulty with this move—from the perspective of the current piece—is that it makes considerations related to backlash superfluous. For immigration restrictions could, on such accounts, be justified just by insisting that the benefits to would-be migrants are of little consequence and that preventing them from moving will be advantageous to certain groups of citizens who, perhaps, deserve special attention (e.g., the least well-off citizens with whom immigrants may compete in the labor market) (Macedo 2007; 2020). No reference to backlash would be needed.

The argument’s second critical shortcoming concerns the reasons for the trade-off between more permissive immigration policies and support for an egalitarian basic structure. As we saw in the preceding discussion, the evidence suggests that working-class citizens do not support egalitarian institutions when those institutions also confer benefits on newly arrived, and ethnically different, immigrants (Alesina, Baqir, and Easterly 1999, 1244; similarly, see Alesina, Murard, and Rapoport 2021; Card, Dustmann, and Preston 2012). The relevant opposition to egalitarian institutions is driven by citizens’ susceptibility to ethnocentric political appeals, especially in the face of the challenges presented by globalization. Yet it is surely objectionable for existing citizens to (a) insist that they will, because of outgroup hostility, refuse to support egalitarian institutions if too many ethnically different immigrants will benefit from them and then (b) appeal to the subsequent weakening of egalitarian institutions as a reason to justify restrictions. The fact that citizens’ objectionable bias is responsible for the existence of the conflict between egalitarian institutions and permissive immigration policies would—all-else-equal—seem to count against resolving the conflict by accepting significant restrictions on immigration. ⁵

The implication of this section’s discussion is that while the egalitarian backlash argument identifies a trade-off between high levels of immigration and an egalitarian basic structure, the argument—taken by itself—fails to explain why we should resolve that trade-off by restricting immigration. It fails because (1) it cannot explain why the cost that it associates with more permissive immigration policies outweighs the important benefits associated with such policies and (2) that cost only arises because of the unjustified behavior of citizens. The rest of the paper explores whether there is a backlash-based argument for restrictions that, at least under specified assumptions, is not susceptible to these concerns.

THE DEMOCRATIC BACKLASH ARGUMENT

This section explores a different backlash argument that potentially circumvents the concern that the benefits associated with restricting immigration appear insufficiently significant. In order to motivate this reformulation, I begin by describing some recent empirical evidence about the relationship between ethnic antagonism and support for democratic institutions.

In the United States, a remarkably large number of citizens claim to be willing to flout rudimentary democratic norms. ⁶

• ◾ Half of Republican and Republican-leaning citizens agree or strongly agree that the traditional American way of life is disappearing so fast that we may have to use force to save it.

• ◾ About one-third of such citizens do not think that it is important for government to prevent private actors from engaging in politically motivated violence.

• ◾ About one-third of such citizens deny that it is essential for government to refrain from interfering with journalists.

These sentiments are “grounded in real political values—specifically, and overwhelmingly, in Republicans’ ethnocentric concerns about the political and social role of immigrants” (Bartels 2020, 22752). ⁷ Indeed, controlling for other differences, respondents are far less likely to accept democratic norms if they agree to statements such as:

• ◾ Things have changed so much that I feel like a stranger in my own country;

• ◾ Immigrants get more than their fair share of government resources; or

• ◾ Speaking English is essential for being a true American.

There are reasons to worry that the relationship between concern about immigration and disparagement of democratic norms is widespread (Norris and Inglehart 2019, 195–200). Far-right parties in Western Europe, which are typically built around a combination of ethnocentrism, anti-immigrant platforms, and a dismissal of traditional democratic commitments, have gained substantial support in recent years (Ivarsflaten 2008; Rodrik 2018, 24). There is, furthermore, extensive evidence of a strong relationship between high levels of immigration and vote share for far-right parties, as well as evidence that the former is an important cause of the latter (Barone et al. 2016; Dustmann, Vasiljeva, and Damm 2019; Halla, Wagner, and Zweimüller 2017; Harmon 2018; Otto and Steinhardt 2014; Tabellini 2020). The effect is substantial; for instance, Dustmann, Vasiljeva, and Damm (2019) exploit the quasi-random assignment of refugees to municipalities in Denmark to show that, in all but the most urban areas, a 1% increase in refugee settlement led to more than a 1% gain in national vote share for far-right political parties (e.g., the Danish People’s Party). This is a sizable effect given that the parties in question had an overall vote share of under 8% in parliamentary elections (Dustmann, Vasiljeva, and Damm 2019, 2036). Similar evidence exists about the impact of non-Western immigrants who are not refugees (Harmon 2018). These are remarkably robust findings, drawn from a range of countries and time periods and estimated via appropriate, modern social scientific methods. The relationship between immigration and support for political parties who aim to circumvent democratic norms is of a piece with the mechanisms discussed in the previous section—political entrepreneurs take advantage of citizens’ anxiety and susceptibility to ethnocentrism to stoke fears about immigration and thereby build support for populist parties. ⁸

From this perspective, the following democratic backlash argument may appear tempting.

P1”. Justice requires a democratic form of government.

P2”. Democracy is endangered in an environment in which ordinary citizens do not accept democratic procedural norms (including freedom of the press and the importance of peacefully resolving political disagreement).

P3”. Immigration contributes to an environment in which enterprising politicians can gain politically by undermining democratic norms.

P4”. In a competitive system, it is unrealistic not to expect enterprising politicians to emerge who are willing to undermine such norms for political gain.

Conclusion: Justice requires the restriction of immigration.

In the relatively recent past, one might have reasonably discounted such an argument as implausibly alarmist, for wealthy democracies have a long history of stability (Przeworski et al. 2000) and are often thought of as “self-enforcing” (Fearon 2011). Yet, in recent years, we have learned that even if wealthy democracies are unlikely to succumb to coup or other kinds of dramatic regime overthrow, they are susceptible to backsliding—that is, to a slow eroding of democratic institutions (e.g., Bermeo 2016; Levitsky and Ziblatt 2018; Walter 2022).

Importantly, leaders bring backsliding about, in part, by undermining institutions that are central to electoral systems’ ordinary mechanisms of accountability (Bermeo 2016, 12). They, for instance, allow for the criminal prosecution of journalists, establish wide-ranging defamation laws, block websites critical of the government, preclude discussion of particular issues, capture the judiciary, prohibit certain parties from standing for election, and tolerate threats and attacks on members of the political opposition. The weaker citizens’ commitments to democratic norms are, the easier it is to take such steps. For our purposes, the crucial point is that the empirical evidence described above suggests that citizens are more willing to accept and even support these tactics in the presence of high levels of immigration. Thus, it at least plausible to think that P2” and P3” are true. While it is still objectionable for political leaders to undermine democratic institutions, the argument supposes that there are reasons to expect political leaders to emerge who will nevertheless be willing to do so (P4”). In light of that expectation, there are reasons for the rest of us to accept immigration restrictions.

Relative to the egalitarian backlash argument, this type of backlash argument has been paid much less attention in the normative literature on immigration. ⁹ Yet, if P2” and P3” are true, this argument is far more powerful. A central problem with the previous version of the argument was that the benefits associated with restricting immigration (i.e., a more egalitarian basic structure) seemed—in the absence of some further argument for prioritizing the interests of citizens—too small to justify the costs (e.g., the benefits to potential immigrants associated with accessing a productive and well-functioning labor market).

The democratic backlash argument is on much stronger footing in this respect. The central reason for this is that a stable democracy—one in which elections peacefully adjudicate conflict over who has the right to rule—is a foundational good. In the absence of an effective mechanism for resolving such conflict, we will need to worry about significantly higher levels of human rights violations (since the opposition will be forced to the streets, and the government will be tempted to persecute them). Furthermore, relative to the competitive authoritarian systems that backsliding threatens to bring about, democracies do a better job of incentivizing leaders to provide benefits to the population as a whole (rather than just to a select group of elites who are in a position to overthrow the authoritarian leader) (Bueno de Mesquita et al. 2005). ¹⁰ The implications are both more far-reaching, and more fundamental, than those at stake in the support of redistributive programs—for they suggest that we should expect democracies to do a better job of providing rule of law, protecting human rights, facilitating relatively efficient market systems, and so forth. ¹¹

The more foundational and far-reaching nature of these benefits has an important further implication for the argument—namely, the very benefits that drive so many to move to the developed democracies (e.g., a reliable property rights system, the protection of human rights, and a well-functioning labor market) are threatened by the collapse of democracy in a way that they are not by the collapse of redistributive programs. While one may reasonably say, then, that the benefits to existing citizens associated with an egalitarian basic structure may be outweighed by the benefits to immigrants associated with accessing a productive labor market, at some margin, it will no longer make sense to say that the benefits associated with maintaining a democratic political system are outweighed by the benefits to potential immigrants of accessing the economy. The preceding suggests that the democratic backlash argument is less susceptible to the first of the two criticisms that led us to set aside the egalitarian backlash argument.

It does, however, appear to be susceptible to the second objection—that the relevant trade-off only arises because of the objectionable behavior of citizens. One may, after all, object that it is audacious, even indecent, for existing citizens to (a) insist that they will, because of outgroup hostility, undermine democratic institutions if too many immigrants arrive and then (b) appeal to the resulting bad consequences to justify restrictions. They are, one might say, taking democracy hostage to win a policy battle over immigration. Nevertheless, at some margin, the argument provides an important—even if purely pragmatic—reason to oppose liberalization of immigration restrictions. Indeed, it even suggests a limit on the number of refugees that states should accept (for related discussion, see Ruhs 2022). ¹²

Once we appreciate the significance of the democratic backlash argument, a number of important questions emerge. How much immigration is compatible with retaining a robustly democratic system? Under what conditions is P3” true? Are there admissions policies that tend to make P3” more or less true? For instance, the evidence of backlash does not appear to extend to programs that prioritize highly skilled migrants—can such programs be expanded without courting backlash? To what extent is democracy, in fact, threatened when citizens reject democratic procedural norms? While these are, in the first instance, empirical questions, resolving them is critical to furnishing an all-things-considered judgment about the appropriate immigration policy to pursue. For present purposes, however, the main point is that the democratic backlash argument provides considerations that speak in favor of immigration restrictions, even assuming that open borders would be required in a world of full compliance. ¹³

Let us take stock. The egalitarian backlash argument failed for two reasons: the costs of immigration that it identified seemed insufficient to counterbalance the connected benefits and, furthermore, those costs only arose because of the objectionable behavior of citizens. The democratic backlash argument proved to be more powerful insofar as it is less susceptible to the first of these two challenges. Yet, as we have seen, the costs that it identifies also arise because of the objectionable behavior of citizens. The discussion to this point, then, is consistent with the common view that backlash arguments may be able to provide pragmatic, but not justificatory, reasons for immigration restrictions (e.g., Carens 1992, 31; Stilz 2019, 201). The next section explores whether it is possible to identify circumstances in which considerations related to backlash may help justify immigration restrictions.

DEMOCRATIC BACKLASH GIVEN A COMMITMENT TO SELF-DETERMINATION

This section, in contrast to the previous two, adopts the assumption that states have a right of self-determination that potentially includes a permission to restrict immigration. Even if the state has such a right, it does not follow that restrictions based on any rationale are acceptable. For instance, the United States may have a right to restrict immigration without it following that citizens may permissibly initiate a blanket ban on Muslim immigrants (Blake 2003, 232–3; 2019, 130; Akhtar 2022; cf. Walzer 1983). Such a policy is objectionable, even assuming such a right, both because (a) it is motivated by the objectionable bias of citizens and (b) it is not credible that it significantly advances a legitimate interest of citizens. At least as we have developed it so far, however, the democratic backlash argument is—with respect to the former—similar to the ban on Muslim immigrants.

This section, by contrast, describes a democratic backlash-based reason for restricting immigration that—under specified circumstances—is linked to a legitimate interest of citizens and is not motivated by an objectionable bias. The idea, then, is to identify a set of conditions under which such backlash-related considerations can contribute not merely to a pragmatic rationale for limiting immigration, but to a justification for such restrictions. From that perspective, this section can be interpreted as a possibility proof. While I try to argue that the relevant conditions are not implausible, they are, as we will see, demanding. One value of the discussion is that laying bare the relevant conditions helps to clarify evaluation of backlash-related arguments.

Elections and the Democratic Equilibrium

The argument will hinge on the claim that immigration can, under certain circumstances, undermine the reasons that political elites have for complying with electoral institutions and, so, threaten the stability of democratic regimes. Hence, I begin by providing an account of how electoral regimes incentivize key actors to comply with electoral results.

Although it is typically taken for granted in the political theory literature, the fact that those who hold political power often allow reasonably fair elections to proceed and step down from power when they lose is, in some sense, surely surprising. What explains such compliance with democratic norms?

Some may hold elections and comply with the results because they are intrinsically committed to democratic norms. Unavoidably, however, a system as dependent on individual ambition as representative democracy will sometimes elevate people to high office who either lack such intrinsic commitments altogether or possess them, but with insufficient strength to override their temptations. Thus, if there are not powerful instrumental reasons for well-placed individuals to comply with democratic norms, the system is unlikely to be stable.

Fortunately, the existence of elections can help facilitate an equilibrium in which reasonably fair elections are held and competitors comply with their results. The effect of elections on the behavior of those out of power is simplest to see. If you strongly disapprove of the government and have no officially sanctioned way to access power, then you have strong reasons to work to overthrow the regime. However, doing so is costly, dangerous, and carries with it very uncertain prospects for success. Thus, if you can have confidence that reasonably fair elections will be held, you will likely prefer to compete with the governing party through that channel. In this way, the existence of regular elections weakens the reasons that those out of power have for working to overthrow the regime in some extra-legal manner. They are typically better off focusing their efforts on electoral victory. Thus, where the opposition is given a reasonable chance of gaining office through electoral competition, they will often have good reason to comply with the rules of the game—even when they lose.

Connectedly, the most straightforward self-interested reason for incumbents to hold elections is that it will incentivize their opponents to abandon efforts at extra-legal regime change—which, if successful, are much more costly for incumbents than removal via election (since the former are likely to involve execution, imprisonment, seizure of assets, etc.). But why might incumbents hold fair elections if their victory is not assured? It could be that the incumbent prefers to compete electorally, even given some elevated chance of losing, just because the benefit of decreasing the attractiveness of extra-legal attempts at regime overthrow outweighs the cost associated with the additional likelihood of being removed from office. This consideration could lead rulers to create, or accede to, electoral institutions. ¹⁴

In addition, given the stability-based benefits associated with electoral victory just noted, the incumbents who will be particularly inclined to cancel elections will be those who lack popular support. To see this, consider the limit case: there is no reason for an incumbent who knows that they will win the election to cancel it. The implication, though, is that canceling elections is tantamount to signaling that one’s support is fragile, which can help the opposition to coordinate on rebellion (Fearon 2011). Given the high costs associated with being removed from office in a coup or rebellion, incumbents may often prefer to hold, and comply with, elections—even when they expect to lose—than to run the risk of such costs by simultaneously (a) signaling the fragility of their support and (b) eliminating the opposition’s opportunities for peaceful competition.

Finally, where electoral institutions are already in place, they make it easier for members of the incumbent’s own party to resist his efforts at regime change. Elite allies of the incumbent (e.g., lower-level elected officials in the party) often have a strong stake in the existing system, for they hold privileged positions within it and often expect meaningful opportunities to compete for higher-level positions in the future. Such individuals, furthermore, will often need to worry that they would be perceived as dangerous competitors in any non-electoral system that the incumbent attempts to establish. Indeed, new authoritarian leaders often purge political figures with a competing basis of public support (Bueno de Mesquita and Smith 2011, chap. 3). We should not assume, then, that even purely self-interested co-partisans will necessarily support an incumbent’s attempt to undermine the existing political regime. In an electoral regime, many of these officials rely for their power, at least directly, on their constituents, rather than on the very individual trying to seize power. This gives them a freer hand to resist the incumbent’s efforts at altering the political regime than they would have in a typical non-electoral regime (e.g., a personalist dictatorship). In this way, the existence of elections makes it more difficult for the incumbent to get support from elites within their own party for undermining the existing regime from within.

The preceding, then, explains both why those who are out of office may prefer to compete electorally and why incumbents may have powerful self-interested reasons to hold elections (even when the prospects of victory are slim) and to comply with the results (even when they lose). Furthermore, if the incumbents recognize the strong reasons that exist for holding elections even when expecting defeat, this should increase their confidence that their competitors—once in office—will also hold elections, which should make incumbents less uneasy about giving up office. The upshot is that the presence of elections can help facilitate a peaceful equilibrium. Roughly speaking, then, what allows electoral systems to do so is that they provide those who lose elections with a reasonable hope of future office holding opportunities. As Przeworski (1999, 50) explains, “Democracy lasts when it offers an opportunity to the conflicting forces to advance their interests within the institutional framework.” While there are many reasons to value electoral systems, their contribution to regime stability and, through that, social peace is one of their signal virtues (for broader discussion, see Landa and Pevnick 2023, chap. 6).

Immigration as a Threat to the Democratic Equilibrium

While the preceding suggests that there are often strong reasons for competitors to play by the rules of electoral systems, such reasons will not always win out—that much is made obvious by today’s many examples of democratic backsliding.

The possibility of changes to immigration policy can, under certain circumstances, weaken the ability of elections to facilitate peaceful transitions of power. To see why, imagine a political system like that of the United States with a large state that has been a reliable supporter of a particular party in Presidential elections. However, due to a large number of immigrants, who—upon gaining the right to vote—overwhelmingly support the opposition party, electoral margins are becoming closer. If the opposition party wins control of government nationally, they are expected to pass legislation that will have the effect of substantially increasing immigration to the state. Since this policy is expected to increase the number of supporters of the party passing the legislation, this will significantly threaten the other party’s electoral chances in the state and, as a result, nationally.

From this perspective, there is a potential similarity between gerrymandering and immigration policy. Gerrymandering is the practice of manipulating the boundaries of political competition so as to advantage one’s own party. One might, for instance, pack opposition voters into a single district in order to waste their votes and, thereby, enable one’s own party to win a number of surrounding districts. If immigrants’ political behavior is relatively predictable, then much the same can, in principle, be achieved through immigration policy. Instead of altering the borders that define political competition, the government could invite more like-minded individuals into the relevant districts and extend voting rights to them. ¹⁵ Both are ways for politicians to control the terms of the very contests that structure their access to power.

The relationship between gerrymandering and immigration is not lost on political elites. A well-known example comes from Malaysia, where the governing party extended voting rights to undocumented immigrants (often by providing them with falsified documentation). The “long-term goal” was to “ensure the political hegemony of Malay-based parties such as the UMNO” and opposition leaders worried that they could, as a result, be reduced to a permanent minority (Sadiq 2005, 109). Similarly, Boss Tweed of Tammany Hall was infamous for vote fraud—one aspect of which was high-speed naturalization proceedings during elections, in which the Tammany Society paid fees, expedited paperwork, bribed judges, and secured witnesses (some of whom would appear dozens of times in a day) for would-be supporters (Connable and Silberfarb 1967, 154). These cases are well known because of their reliance on illicit means. It is much harder to say to what extent parties try to use the ordinary legislative process to alter immigration policy in electorally advantageous ways, for there will always be perfectly plausible, nonpolitical, reasons that they can give (and, indeed, also accept) for favoring politically useful immigration policies. It is instructive, however, that it is no simple task to think of political parties that favor immigration policies that can be expected to harm their electoral chances. Like gerrymandering, then, immigration policy can be seen—among other things—as a tool to alter the terms of political competition so as to advantage one’s own party. ¹⁶

Return now to the example described above. If parties do not expect to be able to efficiently realign to appeal to the new median voter, then fear of the new immigration policy (like fear of gerrymandering) will importantly weaken the reasons that the party that would lose out from it has to hold fair elections in the future (if they are in power) or to comply with electoral results (if they lose). This is because such immigration would weaken their expectation of future opportunities to hold office. However, as we have seen, the expectation of such opportunities is critical to democracy’s ability to facilitate peaceful rotation of power. As the literature on civil conflict stresses, people are especially prone to fight if they see their power slipping away with little hope of regaining it in the future, which most often happens for demographic reasons, including those related to migration (Walter 2022, 63). The important point is that immigration policy decisions, or even just the possibility of such decisions, may threaten the ongoing stability of otherwise well-functioning electoral systems. ¹⁷

This threat depends, as I just noted, on there being an inability to efficiently realign around a new median voter. There are two reasons to think that that condition may often hold. First, voters’ political identification is quite durable (Bartels 2010)—indeed, one’s voting behavior is strongly predicted by the political identification of one’s parents. So, once voters have developed a partisan identification, they will often resist quickly running to a different political party just because that party alters its policy platform. To the contrary, there is considerable evidence that, on many issues, citizens’ policy views depend on their partisan affiliation, rather than—as traditionally assumed—the other way around (e.g., Lenz 2013). Second, there is also considerable friction within parties that prevents them from efficiently moving to the position of the median voter. In the U.S. system, important sources of friction include the need to compete in primaries and the need to be responsive to donor constituencies (e.g., Kujala 2020).

Moreover, even if the parties are able to realign around a new median voter, there are two reasons that this may be insufficient to facilitate stability. First, the stability of the electoral system depends not just on parties’ long-run chances to hold office, but also on the expectations of individual political elites, who necessarily have shorter time horizons. Thus, the ability of political parties to converge on the new median voter in the long run may not be a sufficient response to concerns about instability. Second, even if the parties efficiently realigned around a new median voter, this realignment could leave a substantial minority of citizens alienated by the existing options and, so, amenable to the appeal of an enterprising populist figure—the kind of political actor who succeeds by undermining democratic institutions. For these reasons, it strikes me as imprudent to dismiss the threat that demographic change can pose to the political system by expecting political parties to always be able to realign around a new median voter without significant costs.

A Backlash-Related Justification of Immigration Restrictions?

Now consider a slightly different kind of case involving a closely divided democracy. Imagine that both parties recognize that were there to be an election in a year in which conditions were sufficiently unfavorable, their adversaries could subsequently, and unilaterally, pass immigration policy that would tip the balance of power. Assume, furthermore, perhaps for the reasons noted above, that they could not expect to be able to costlessly realign around the new median voter. The parties may reasonably fear such an outcome because it would undermine—in the relatively long run—their chance to hold office and ability to meaningfully affect policy outcomes. This case is different, then, from the example presented in the preceding section because in the present case we are imagining the parties to be symmetrically positioned: there are immigration policies that either could pass that would undermine the political position of their rival. One way that such a situation could arise is if (a) party affiliation were largely a function of economic position and (b) one party favored a skills-based immigration policy aimed at recruiting high-earners, whereas the other favored policies that prioritized less-advantaged potential migrants (perhaps via a policy that prioritizes family reunification).

Imagine, further, that both parties are conditionally committed democrats: their preferred outcome is that they, and their adversaries, comply with democratic norms (meaning, here, compliance with unwelcome electoral results and the refusal to use immigration policy as a political tool). However, both parties would rather renege on those democratic norms than uphold them in a situation in which their adversaries renege. Given such motivations, the example takes the form of an assurance dilemma, which can be illustrated via the following payoff matrix (Table 1).

Table 1. Higher # = Preferred Outcome (A, B)

There are two pure-strategy equilibria here—one in which both parties comply with democratic norms and one in which both renege. The central challenge when facing an assurance dilemma lies in figuring out how to organize around the socially optimal equilibrium. The parties need to find a way to credibly commit to compliance.

One possibility is for the parties to agree, before such an election occurs, to a constitutional amendment that prohibits admitting more than a relatively small number of immigrants each year. The parties should find such an amendment attractive for three reasons. First, it provides them with assurance that their opponent will not, in the event of a large electoral victory, use immigration policy to lock in their political advantage. This is important not because their opponent would, in fact, prefer to do so. Since the opponents are conditionally committed democrats, they would only do so in the presence of a compelling reason for thinking that the other side would. The assurance is important, instead, because it provides the parties with a way to credibly signal to one another that they will comply with democratic standards. In the absence of such a signal, there is a risk of ending up with the inferior equilibrium.

Second, the constitutional amendment assures their opposition that, even in the wake of a huge defeat, the shape of the political community will not be restructured, via immigration policy, in a way that locks in their disadvantage. This is also important to the winning party because, by lowering the stakes, it makes it safer for the losing party to peacefully accept such a defeat. Again, the point is not that the winning party would have otherwise passed such an immigration policy; instead, the constitutional amendment works to assure the losing party that this will not happen, which helps to stabilize the socially optimal equilibrium.

Third, the mere possibility of one of the parties making political use of immigration could threaten conciliatory politics. I—as a member of Party A—may prefer compliance. However, I may be concerned that Party B is not, in fact, committed to democratic norms and will renege, given a sufficiently large electoral victory. In light of that, I may work to pre-emptively undermine fair democratic competition. Meanwhile, members of Party B may interpret my efforts as a preference for reneging, which—in turn—may lead them to do the same (thus reinforcing my initial doubts). In this way, the mere possibility that a party could use immigration for political gain can make it more difficult to sustain conciliatory politics. It may encourage citizens see their political rivals as existential threats, rather than as fellow participants in an ongoing scheme of social cooperation with whom they disagree on important substantive matters. Citizens may, therefore, have an interest in removing the possibility of unilateral changes to immigration policy from the table, even in cases in which parties are highly unlikely to win the big electoral victories that would be needed to enact them. This is important because it shows that the case for restrictions does not depend on actually observing instances in which immigration policy is successfully used for political gain; instead, that may be an off the path of play possibility that nevertheless pollutes political life.

Critically, the considerations in favor of the amendment do not involve the parties threatening to refuse to comply with democratic norms if they do not get their way on substantive policy measures. To the contrary, the parties favor the amendment precisely because it allows them to effectively reassure one other that their collective decisions will be made via ordinary democratic procedures (e.g., persuasion and mobilization), rather than via end runs around them. To see this, it may be helpful to imagine a version of the case in which both parties are unconditionally committed to democratic norms. Even here, the parties will be uncertain whether their adversaries are committed to such norms. Because of this uncertainty, the three reasons for accepting the constitutional amendment given above would still apply. Thus, the argument does not depend on there being any limits on the parties’ commitment to democracy. Rather, it depends on uncertainty about the extent of one another’s commitments—which is a basic feature of political life.

Thus, the constitutional amendment can be thought of as facilitating a credible commitment to “disarmament” (i.e., a refusal to use immigration policies for political advantage). The amendment, then, is a way of safeguarding the project of democratic self-determination. Surely, it is implicit in the idea that the state has a right of democratic self-determination that the factions that compose it can take meaningful steps to reassure one another that they are, indeed, committed to compliance with democratic norms and procedures. ¹⁸

Assuming, then, that states have a right of self-determination, this appears to be a case in which a democratic backlash argument provides a justification—and not merely a pragmatic reason—for immigration restrictions. This is because, unlike in the previous versions of the backlash argument, the argument explains how restrictions help advance legitimate interests of citizens (in regime stability, social peace, and a more conciliatory politics) without requiring us to assume that anybody acts (or wishes to act) inappropriately. Instead, the crucial feature of the example is that while the parties wish to comply with democratic norms, they are uncertain about one another’s intentions and, so, need a way to make this commitment credible to one another. As a result, each may reasonably see the benefit of taking the possibility of using immigration policy as a political tool off the table by constitutionally prohibiting the risky immigration policies. This, then, is a backlash-related argument for restrictions that (a) identifies legitimate interests that are advanced by restrictions and (b) is not motivated by citizens’ unjustified bias. Thus, under the specified circumstances, backlash-related considerations appear to furnish a justification, and not just a pragmatic rationale, for immigration restrictions. ¹⁹

This argument has strayed a considerable distance from the backlash argument with which we began. Does it still make sense to think of it as a “backlash” argument? I defined backlash, at the outset, as action that imposes social costs and is taken, consciously not, in response to a disliked policy. In the setting under consideration, think of the disliked policy as Party A’s potential use of immigration policy for political ends. The social costs imposed in “response” to this policy include Party B’s unwillingness to comply with electoral results and/or hold fair elections. The “backlash” may precede the change to immigration policy, for it is taken in strategic anticipation of it. Yet this may well be a feature of the more familiar cases, as well—the crucial point is just that what gets the argument for restrictions off the ground is anxiety about the potential political use of immigration and the costs that may be imposed in response. In that sense, the argument hinges on actions that impose social costs and are taken in response to immigration policy. So, the case fits the template of backlash arguments, at least as I have defined them here. ²⁰

My claim, then, is that we have identified a way in which considerations related to backlash could help furnish a justification of immigration restrictions. Before commenting on the relevance of this in-principle possibility, I briefly consider two potential objections—each of which contains a kernel of truth, even if—as I will argue—neither undermines the section’s main claim.

First, one may object that there are many other threats to democratic stability—some of which may be more significant than immigration. Potential examples include the partisan nature of the information environment, institutional rules that systematically favor certain segments of the population, and the broader ability of incumbents to, in effect, control the rules of political competition (e.g., through gerrymandering, and control over rules of campaign finance) (Landa and Pevnick 2022). While I do not dispute the claim that other sources of instability may be more significant, it does not follow from the fact that democracies face other threats to stability that citizens should be regarded as powerless to defend themselves from immigration-related threats.

Second, somebody may rightly point out that limiting immigration is not the only way to prevent the identified threat to political stability. Moreover, other strategies may not force us to sacrifice the welfare and freedom-related gains associated with immigration. One possibility is to delegate the design of immigration policy to the supermajority decision of a committee constituted by members of all of the society’s major political parties. This would be analogous to the use of interparty commissions to handle redistricting. Such an arrangement could help prevent parties from using immigration policy as a political tool, defuse concern about the asymmetric political effect of immigration decisions, and—by doing so—mitigate democratic backlash. Insofar as conflicts about immigration are, in important part, proxy battles over political power, rather than merely sincere substantive disagreements about the costs and benefits of immigration, such an arrangement may lower the temperature of disputes surrounding immigration policy. A second possibility is to insist that migrants wait a longer period of time before being granted the right to vote. This may make the political use of immigration policy less tempting by (a) increasing the uncertainty surrounding migrants’ likely political behavior and (b) moving the potential political benefits beyond the window that many individual elected officials will consider relevant. Both of these institutional strategies are well worth considering. However, since there can be reasonable doubts about their likely efficacy, and because they will come with their own downsides, I see no grounds for asserting that a self-determining state must prefer them to immigration restrictions.

The Argument’s Relevance

The previous subsection identified a setting in which backlash-related considerations can contribute to a justification of immigration restrictions. This setting requires that:

1. 1. States have a right of self-determination,

2. 2. Immigrants’ political behavior is sufficiently predictable,

3. 3. It is difficult for parties to efficiently, or without significant costs, realign around a new median voter, and

4. 4. The factions are motivated by apprehension about the uncertain behavior of their opponents.

Does the requirement that these conditions hold render the argument irrelevant to the immigration debate in actual democracies?

With respect to (1), there is already a large literature debating whether states have a right of self-determination that gives them a permission to restrict immigration. ²¹ I cannot, in this context, add anything meaningful to it. Turning to (2), the predictability of immigrants’ political behavior will plainly vary from case to case and is, furthermore, endogenous to the particular policy that is passed. So, I do not think that there is much general that can be said here, aside from noting that people’s political behavior is often fairly predictable given knowledge of underlying demographic information. ²² Meanwhile, while (3) depends on specific characteristics of the political system in question, I have already provided reasons for thinking that it may hold in some existing democracies. Thus, I set (1–3) aside here.

Let us consider, then, (4). The example that I used to motivate the justification abstracts entirely from ethnic antagonism (or other forms of objectionable bias) in order to clearly identify how political anxiety about immigration policy can independently threaten political stability and conciliatory politics. However, while mechanisms related to political anxiety and ethnic antagonism are analytically separable, we should expect them to coexist in practice. In part, this is because, as we have noted, politicians who are anxious about the political implications of immigration have reason to drum up opposition to immigration by exploiting people’s susceptibility to ethnic antagonism. Thus, we are unlikely to come across an actual case that mirrors our hypothetical example—one in which the factions are motivated purely by anxiety about the uncertain behavior of the other side, with objectionable motivations entirely absent.

This raises the difficult question of whether the argument retains justificatory power in the more complicated, and more realistic, scenario in which both kinds of motivations for restricting immigration (i.e., those related to ethnic antagonism and uncertainty) are present. While I will not try to defend it, it seems to me that the most reasonable position is that the justification is germane if accepting restrictions on immigration may be important to stabilizing the democratic system and, so, facilitating social peace independent of citizens’ objectionable motivations. However, for the reason described in the preceding paragraph, we should expect that when this is true, it will also be true that many will oppose immigration for straightforwardly objectionable reasons. If this position were correct, the argument would retain relevance in circumstances that are more complex and multifaceted than the example used to motivate it. ²³

The further, applied, question of interest is, surely, whether the circumstances in today’s highly polarized democracies satisfy the standard laid out in the previous paragraph. It certainly seems plausible—say, in the United States—that (a) there is a lot of anxiety about the political implications of immigration (this is quite palpable), (b) such anxiety is contributing to legitimate risks to the ability of the electoral system to facilitate peaceful exchange of power, and (c) the costs associated with a breakdown of peaceful exchange of power and subsequent eroding of the democratic system are sufficiently significant that they could provide a justifiable reason to restrict immigration (if we also assume that (1–3) above obtain). Yet moving beyond the claim that this seems like it may be a “plausible” description of circumstances in the United States would depend on a host of complicated and context-specific empirical considerations—as well as difficult matters of judgment. Just how much of the distrust in the political system is driven by immigration? How realistic are the threats of democratic backsliding? How costly might such backsliding be? I cannot explore these questions here. My more limited ambition is to defend the claim that there are circumstances under which a version of the democratic backlash argument can, in principle, provide a justification of immigration restrictions that is not vulnerable to either of the shortcomings that face the more traditional, egalitarian backlash argument.

CONCLUSION

The goal of this essay has been to explore the power of backlash-related reasons for immigration restrictions in the developed democracies, particularly in the context of recent evidence and events. I have argued that the most common backlash-related arguments for restricting immigration—those that hinge on egalitarian distributive considerations—fail to provide a convincing rationale for restrictions. The central difficulties facing such arguments are (1) their inability to explain why we should regard the costs associated with undermining egalitarian institutions as outweighing the benefits of more permissive immigration policies and (2) the fact that it is the objectionable behavior of citizens that leads to the erosion of redistributive institutions. These difficulties suggest the force of an alternative rendering of the backlash argument, according to which the case for restricting immigration hinges on threats to democracy, rather than threats to distributive justice. Since the benefits of stable democracy are quite significant, they cannot be as easily overwhelmed by the substantial benefits associated with more permissive immigration policies. Moreover, we have seen that there is a version of the democratic backlash argument that provides a rationale for restrictions that does not depend on anybody acting objectionably. Under certain circumstances, then, the democratic backlash argument can provide important reasons to restrict immigration, even while circumventing the difficulties facing the egalitarian backlash argument. One important implication is that the case for increased levels of immigration depends on identifying proposals that are less likely to create backlash toward, and abandonment of, democratic institutions.