About this Issue

How do immigrants affect the institutions in the countries that they move to? It’s fundamentally an empirical question, but it’s complicated: It’s hard to analyze without an account of which institutions truly matter, and how, and how they relate to one another. The values and beliefs of the immigrant population itself will also vary with time and place. As a result, the lessons learned in one country may or may not translate to another—but perhaps they do.

This month’s lead essay is by Prof. Benjamin Powell, co-author with the Cato Institute’s Alex Nowrasteh of Wretched Refuse: The Political Economy of Immigration and Institutions (Cambridge University Press: 2020). Responding to him will be Prof. Peter Skerry of Boston College, Prof. Ilya Somin of George Mason University, and Prof. Eric Kaufmann of Birkbeck College, University of London; each will write a response essay. After all have written, discussion will follow through the end of the month. Comments are enabled as well, and we invite readers to join the discussion.

Lead Essay

Would Immigrants Destroy our Freedoms?

If the United States returned to a policy of essentially unrestricted immigration, would a much larger number of immigrants lead to a decrease in our economic freedoms?

Cato economist Alex Nowrasteh and I explored this question among others in our recent book, Wretched Refuse: the Political Economy of Immigration and Institutions.[1] Economists estimate that the global gains to humans’ welfare would be massive if destination countries eliminated their current restrictive immigration policies and allowed for an unrestricted number of people to immigrate. Estimates range from an increase of 67% to 150% of total global output.[2] However, these gains are predicated on destination countries maintaining their high level of productivity.

Some social scientists have embraced what is coming to be known as “the new economic case for immigration restrictions.” These scholars posit that some of the factors responsible for the low productivity in immigrants’ origin countries might “migrate” with the immigrants and decrease the productivity of destination countries. These factors include the norms and beliefs of the immigrants that shape formal and informal institutions. Our book is the most comprehensive empirical examination of the new economic case for immigration restrictions, and it considers multiple factors related to productivity that immigration could influence.

This essay focuses on economic freedom because it is one example of an important institution related to productivity that immigrants could erode. Hundreds of papers have been written finding a positive relationship between economic freedom and desirable outcomes using the Economic Freedom of the World Annual Report over the last 25 years.[3] This index measures economic freedom across five main areas: The size of government; protection of private property rights; sound money; freedom to trade internationally; and regulation of businesses, labor markets, and credit markets.

My choice to focus on immigrants’ impact on economic freedom should also be of interest to conservatives and libertarians, independent from any impact economic freedom has on productivity. The fear that immigrants might erode our freedoms is a common one among conservatives and libertarians and can be traced back to fears of some of the founding fathers. For example, in 1783, Thomas Jefferson wrote that immigrants “will bring with them the principles of the governments they leave, imbibed in their early youth; or, if able to throw them off, it will be in exchange for an unbounded licentiousness, passing, as is usual, from one extreme to another. It would be a miracle were they to stop precisely at the point of temperate liberty.”[4]

Do immigrants bring with them the principles of the governments that they leave behind? If so, do they hold onto those principles long enough to impact the institutions of governance in destination countries? It is certainly plausible that immigrants from less free countries could decrease the economic freedom of destination countries. But it is also plausible that instead, they assimilate their beliefs about the proper role of government to match the views of the native-born population. A positive selection bias might also lead those people who most value the increased freedoms in destination countries to immigrate. This selection bias could even lead to increases in economic freedom in destination countries. There are multiple and conflicting channels of how immigrants might, or might not, impact the economic freedoms in destination countries. The netting of these effects is ultimately an empirical question. Let’s review some evidence that Alex and I explain in more detail in Wretched Refuse.

Some Evidence from U.S. History[5]

Prior to 1922, immigration to the United States was largely unrestricted, with the notable exception of bars on immigrants from East Asia, Africa, and other parts of Asia. In the century leading up to 1922, the average annual number of immigrants to the United States amounted to about 0.66 percent of the U.S. population. Immigration to the United States was most restrictive between 1922 and 1967. During this time, the average inflow of immigrants amounted to only 0.14 percent of the population per year. Since immigration laws were partially liberalized in 1968, the average inflow of immigrants doubled to about 0.3 percent of the population per year, but it still remained below half the level of that prior to 1922.

Unfortunately, we do not have a great measure of overall economic freedom for this entire period, but federal expenditures as a percent GDP can serve as a decent proxy. These expenditures increased more than 302 percent in the 45 years that immigration was most restricted. In the 45 years preceding the 1922 restrictions they grew less than 118 percent and in the 45 years following their partial liberalization they grew 8.7 percent.[6] This negative correlation is, of course, not causation. Two World Wars, a Great Depression, and numerous other factors are at play. But as a first cut, it is suggestive evidence that larger numbers of immigrant arrivals were not associated with decreased economic freedom in the past.

What about socialist immigrants during the period of unrestricted immigration? It is true that a large percentage of American socialist intellectuals were immigrants and even published their propaganda in foreign-language periodicals. But a socialist movement never grew in the United States to the extent it did in these immigrants’ origin countries. Many immigrants pursued their American dreams rather than socialist activism once they arrived in the United States. For instance, one list of 797 exiled socialists from Germany who arrived in the United States and were taken care of by the Socialist Labor Party records that only 191 of them joined the party—an attrition rate of 76 percent.

Immigrants did participate in labor activity, but because of the United States’ open immigration policy, ethnic, linguistic, and religious fractionalizations were increased. As a result, unionization rates remained low because collective action in heterogeneous communities is more difficult than it is in homogeneous communities. Thus unions, one of the main organized groups that lobby and vote for a larger and more interventionist welfare state, were smaller than they would have been in a less diverse America. Furthermore, voters were less supportive of redistribution, welfare, and government-supplied services when the consuming population was more heterogeneous. As a result, federal outlays, state expenditures, and local tax rates were negatively correlated with the size of the immigrant population.

Much more could be written on this topic. We spend an entire chapter of Wretched Refuse on it and only scratch the surface. However, even if one believes that unrestricted immigration in U.S. history didn’t lead to a decrease in economic freedom, that doesn’t mean unrestricted immigration wouldn’t reduce economic freedoms today. Today there is a large welfare state, information and transportation costs are lower, and immigrants’ origin countries differ, among other factors. Perhaps greater immigration today would impact our freedom differently than it did in our past.

Some Evidence Across Countries Today[7]

The first study coauthors and I did on immigration’s impact on economic freedom examined a broad cross-section of countries over two decades. We examined how initial stocks of immigrants in 1990, that were accumulated over many decades, and/or the subsequent 20-year inflow of immigrants, were associated with a country’s level of economic freedom in 2011. The immigrant share of the population in the 110 countries we studied ranged from a low of 0.03 percent of the population (China) to a high of 77 percent of the population (Kuwait).

We ran 32 separate regressions that looked at initial immigrant stocks and inflows individually, together, and interacting with each other. We ran baseline results that only controlled for countries’ initial level of economic freedom and more fully specified regressions that also controlled for countries’ income levels and political institutions. In our 32 regressions we did not find a single instance where immigration was associated with decreased economic freedom and also statistically significant. Instead, in half our regressions we found that greater immigration was associated with increases in economic freedom that were both statistically significant and meaningful in magnitude. For instance, in one specification we found that a one standard-deviation larger immigration stock in 1990 increases economic freedom by 0.34 points in 2011. Using an estimate for the impact of economic freedom on growth, that suggests that an increase in the immigrant share of this magnitude will generate a 0.45 percentage point higher long-run annual growth rate.

There are obvious limitations to using cross-country studies to infer whether greatly expanded immigration would decrease our freedoms in the United States today. A standard concern with cross-country studies is, of course, endogeneity–the fact that increases in economic freedom improve economic outcomes, and that immigrants are attracted to economic opportunities, could mean that immigration is not causing economic freedom to increase, rather, immigration is responding to it. Using initial stocks of immigrants that were accumulated over many decades partially alleviates endogeneity concerns (it is not likely they came in anticipation of freedom increasing later) but not completely. Furthermore, cross-country studies tell us the overall effect across all of the countries, perhaps immigration improves economic freedom in some countries while harming it in others. Perhaps most importantly, existing stocks and flows of immigration are the result of managed immigration policies that are highly restricted and thus impact both the mix and quantity of immigrants. Even if exiting immigrant stocks and flows increase economic freedom, perhaps unrestricted flows would decrease economic freedom.

Some Evidence from Recent Mass Migrations[8]

Individual case studies of modern mass migrations are an alternative way to examine how greatly increased immigration can impact economic freedom. We examine two of them, Israel and Jordan, in Wretched Refuse. Israel’s population increased by 20 percent during the 1990s due to an influx of immigrants from the Soviet Union. Jordan’s population increased about 10 percent due to immigration of Kuwaiti-Palestinians who were displaced by Iraq’s invasion of Kuwait and the subsequent Gulf War.

The mass migration to Israel occurred when Soviet emigration restrictions were lifted. Israel’s “Law of Return” allows any Jewish person to immigrate to Israel and receive instant citizenship, including full voting rights, upon arrival. The Law of Return defines Jewish people who qualify quite broadly, so the resulting immigration from the Soviet Union was mostly not religiously or linguistically homogeneous to the existing Israeli population. They also arrived from a country with a 70-year history of little economic or political freedom. If the beliefs responsible for the lack of freedom in the Soviet Union migrated with the immigrants, they could have used their political rights to undermine Israel’s economic freedom.

Instead, Israel’s economic freedom increased 45 percent during the decade following the mass migration. As a result, Israel catapulted from the 92nd freest economy in the world in 1990 to the 54th freest in 2000. In research with coauthors, and in the book, I argue that the immigrants caused this increase in economic freedom. One way we established causality was to employ a synthetic control methodology to create a counterfactual Israel to measure the real Israel’s changes against. We find that the real Israel improved more than its synthetic counterpart to a statistically significant extent. We also document the political activity of the immigrants from the former Soviet Union and argue that they were a political force for greater economic freedom.

Jordan’s case is similar. Due to a quirk of Jordanian law, the Kuwaiti-Palestinians received political rights, including the right to vote, upon arrival. Like Israel, Jordan also improved its economic freedom following the mass migration. Jordan improved its economic freedom score by nearly 34 percent and rose in the rankings from 65th to 42nd freest country between 1990 and 2000. The same synthetic control methodology also establishes that this is a larger improvement than counterfactual versions of Jordan that did not experience an immigration surge. We also describe how the political influence of the immigrants tipped the balance away from Transjordanians, who were largely opposed to liberalization, and towards Jordanian-Palestinians, who favored liberalization.

The advantages and disadvantages of the Israel and Jordan studies are similar. Neither had truly free immigration policies. Instead, each only had a free immigration policy that applied to certain people: Jews in the case of Israel and Kuwaiti-Palestinians in the case of Jordan. However, both benefit from having mass immigrations, from countries with low economic freedom, that stem from a clearly exogenous shock. Thus, endogeneity concerns are largely absent and are instead replaced with concerns about the external validity of these cases for other countries.


There is no Q.E.D. here. Nor is there one in Wretched Refuse (But you should still buy the book!). There are limitations to each way we examine the relationship between immigration and changes in economic freedom. These same limitations are present to varying degrees in the numerous other ways Alex Nowrasteh and I empirically investigate the new economic case for immigration restrictions in Wretched Refuse. However, regardless of which way we examined the relationship, we never found evidence that immigration decreases economic freedom in destination countries. That said, our historical investigation, cross-country findings, and case studies, cannot rule out the possibility that, in particular cases, immigration from one or more origin countries to a particular destination country could decrease economic freedom. However, empirical conjectures require empirical evidence. Until such evidence emerges, conservatives and libertarians should not oppose greater immigration because of its threat to our freedom. Similarly, economists should continue to estimate global gains from unrestricted immigration without assuming immigrants would undermine the factors that make destination countries productive.


[1] Nowrasteh, Alex and Benjamin Powell. (2020). Wretched Refuse? The Political Economy of Immigration and Institutions. Cambridge: Cambridge University Press.

[2] Clemens, M. (2011). “Economics and emigration: Trillion-dollar bills on the sidewalk?” Journal of Economic Perspectives 25(3), 83-106.

[3] Hall, J. and Lawson, R. (2013). “Economic freedom of the world: An accounting of the literature.” Contemporary Economic Policy 32: 1-19.

[4] Jefferson, T. (1787). Notes on the State of Virginia. London, John Stockdale, Chapter 8.

[5] Chapter 9 of Wretched Refuse covers this topic in much greater detail.

[6] The results are similar if calculated as real per capita expenditures instead of as a percent of GDP.

[7] Chapter 5 of Wretched Refuse covers the cross country evidence in much greater detail. Some of the main empirical work that the chapter is based on can also be found in: Clark, J., Lawson, R., Nowrasteh, A., Powell, B. and Murphy, R. (2015). Does immigration impact institutions?. Public Choice, 163(3-4), 321-335.

[8] Chapters 10 and 11 of Wretched Refuse cover these examples in greater detail. Those chapters are based on work published in: Powell, B., Clark, J., and Nowrasteh, A. (2017). “Does mass immigration destroy institutions? 1990s Israel as a natural experiment.” Journal of Economic Behavior & Organization 141, 83-95 and Nowrasteh, A., Forrester, A. and Blondin, C. (2020). “How mass immigration affects countries with weak economic institutions: A natural experiment in Jordan.” World Bank Economic Review 34 (2), 533-549.

Response Essays

How to Counter Possible Political Dangers of Immigration Without Barring Migrants

There are many possible justifications for immigration restrictions. But perhaps the most potent is the fear that too much immigration of the wrong kind might kill the goose that laid the golden eggs that make a nation attractive to migrants in the first place. Immigrants who do not value liberal democratic institutions or do not understand them might end up destroying them. The same goes for other economic and political institutions that promote liberty and prosperity. In the worst-case scenario, immigrants from nations with horrendous governments might simply replicate those same regimes in their new homes.

Alex Nowrasteh and Benjamin Powell’s book Wretched Refuse? The Political Economy of Immigration and Institutions is the most thorough and compelling refutation of such fears to date. As the authors painstakingly show, far from damaging the institutions of free societies, immigration poses little threat to them, and sometimes it makes them stronger.

I agree with nearly all of their major points and analysis. So I will focus on two important related issues that the authors largely leave out. First, Nowrasteh and Powell’s book focuses almost entirely on possible negative effects caused by the immigrants themselves. But it is also possible that institutional damage can be inflicted by natives’ reactions to immigration, even if those reactions are irrational or unnecessary. Thus, some argue that we need to restrict immigration not to protect natives from immigrants, but to protect natives against themselves.

Second, what if Nowrasteh and Powell are wrong, at least in some cases, and immigration does damage political institutions after all? Even in that worst-case scenario, immigration restrictions still might be unjustified. Instead, we should first consider whether the problem can be alleviated by less draconian measures.

The Problem of Nativist Backlash

In recent years, the danger posed by authoritarian nationalists has attracted renewed attention, as strong illiberal, anti-immigrant movements have arisen in both the United States and many European nations. In the United States, “ethnic antagonism” toward immigrants and racial minorities is the strongest predictor of anti-democratic attitudes among Republican voters. In Europe, illiberal nationalist movements hostile to immigration have risen to power in Poland and Hungary and gained electoral strength in many other countries.

Some have argued that western nations must reduce immigration in order to take away an issue these movements could use to political advantage. In the most extreme scenario, anti-immigration backlash could lead to the degradation of liberal democratic institutions by authoritarian nationalists. Thus, conservative political commentator David Frum warns that, “if liberals won’t enforce borders, fascists will.” Others have echoed such concerns.

One flaw in the argument that immigration restrictions are needed to constrain nativist backlash is that survey data consistently shows that most people in both the United States and Europe consistently overestimate the true amount of immigration, and those most opposed to immigration overestimate the most. Given such widespread ignorance, we cannot assume that, say, a 10% reduction in immigration will lead to a parallel reduction in ethno-nationalist sentiment. Indeed, most nationalist voters might not even notice the difference.

Nativism is often caused by deeply rooted xenophobic attitudes, rather than by the actual number of immigrants. Hostility to immigration among natives often tends to be greatest in parts of the United States and other countries that have the fewest immigrants. Multiple studies find either that a relatively high percentage of immigrants in close proximity increases favorable attitudes towards immigrant groups among natives, or at least has no detectable effect.[1] In the 2016 British “Brexit” referendum on leaving the European Union, in which the “Leave” side emphasized immigration as a major issue, areas with low percentages of immigrants in the population were far more likely to support Brexit than those with high levels.

Even if increased immigration does cause dangerous backlash, we should apply narrower, “keyhole” solutions in preference to migration restrictions. Some public opinion research suggests that hostility to immigration increases not because of the actual number of immigrants, but due to a sense of “loss of control” over the immigration process.[2] Voters might feel a loss of control even if the actual number of immigrants has not increased significantly.

If this is indeed the cause of the problem, there is an obvious keyhole solution: make it easy to enter the country legally, but require migrants to do so in an “orderly” way, such as by registering at designated points of entry. Migrants would have strong incentives to comply with such a system, so long as controlled entry and registration are easy to do and result in rights to residency and employment. This approach relies on making legal migration easier rather than harder. The easier it is, the stronger the incentive of migrants to comply with registration requirements instead of trying to enter illegally.

Efficacy aside, the idea that we must restrict immigration in order to protect against native-born nationalists is morally perverse. It suggests we severely restrict the liberty and opportunity of innocent people in order to protect against wrongdoing by others. The backlash-prevention rationale for immigration restrictions is similar to 19th-century claims that we must allow southern whites to impose racial segregation on blacks in order to prevent whites from continuing to engage in violence and otherwise pose an ongoing threat to the Union. Immigration restrictions, in fact, have many similarities to domestic racial segregation. Both impose severe constraints on liberty and opportunity based on arbitrary circumstances of birth, and often also based on racial and ethnic prejudice.

Keyhole Solutions

Where migration creates genuine problems, it is often possible to deal with the issue by means of “keyhole solutions” that minimize the risk without barring migrants. Instead of applying a meat cleaver that inflicts great sacrifices on potential migrants and many natives, it is better to apply a scalpel.

Where keyhole solutions are inadequate, policymakers should consider tapping the vast wealth created by expanded migration to mitigate negative side-effects that cannot be addressed in other ways. As Ben Powell points out in his lead essay, economists estimate vast economic gains from abolishing immigration restrictions, roughly a doubling of world GDP if all such barriers were abolished. Some of that new wealth can, if necessary, be put to use alleviating negative side effects of migration itself.

I do not claim this approach can solve all conceivable problems potentially caused by immigration. There are likely to be extreme cases where it fails. But the framework can be effectively applied to a wide range of issues often seen as strong justifications for imposing migration restrictions.

That includes many potential harms to political institutions. Consider, for example, claims that immigration will overburden the welfare state. Evidence from both the United States and Europe indicates that this problem is overblown, as jurisdictions with more immigrants do not have higher per capita welfare spending than those with fewer, and the vast majority of immigrants contribute more to the public fisc than they take out. But to the extent this is a problem, there are a variety of keyhole solutions, most notably restricting immigrants’ eligibility for various welfare benefits, as many nations—including the United States—already do.

Similarly, we may worry that immigrants will vote for politicians who institute harmful policies, detrimental to freedom. If we are not careful, they might even help elect a president who has no regard for liberal democratic values, and regularly expresses admiration for foreign despots!

Extensive evidence compiled by Nowrasteh and Powell suggests that such scenarios are greatly overblown. But to the extent they are a real danger, there are obvious keyhole solutions that could mitigate the risk. In the United States, immigrants are already ineligible to become citizens with voting rights until they have resided in the country for at least five years and pass a civics test most native-born Americans would fail. If necessary, the residency period could be extended, and the civics test made more difficult.

In Chapter 6 of my book Free to Move: Foot Voting, Migration, and Political Freedom, I outline similar keyhole solutions for a variety of other potential negative side-effects of immigration, including environmental harms, the spread of illiberal cultural values, and more. As described above, there are possible keyhole solutions for the nativist backlash scenario as well.[3]

In cases where there is both a genuine problem and no feasible keyhole solution, we can instead tap the vast wealth created by migration to alleviate negative side effects. For example, if, contrary to the available evidence, we conclude that immigration increases crime rates, we can use some of the extra wealth created to increase the number of police on the streets, which social science evidence shows can reduce crime greatly.

There can still potentially be extreme cases where immigration causes harm that cannot be forestalled by either keyhole solutions or tapping wealth created by migration. Even in those cases, however, immigration restrictions would only be justified if the harm prevented is greater than the enormous harm that migration restrictions themselves inflict on both immigrants and natives.

Keyhole solutions and targeted expenditures may not work in every case. But, in combination with the enormous gains of free migration rights, and the powerful evidence compiled by Nowrasteh and Powell, they justify a strong presumption against immigration restrictions.


Ilya Somin is a law professor at George Mason University and author of Free to Move: Foot Voting, Migration and Political Freedom. Parts of this essay are adapted from the forthcoming revised edition of Free to Move.

[1] See, e.g., Joel S. Fetzer, Public Attitudes Toward Immigration in the United States, France, and Germany. (New York: Cambridge University Press, 2000) (no effect); Christopher G. Ellison, Heeju Shin, and David L. Leal, “The Contact Hypothesis and Attitudes Toward Latinos in the United States,” Social Science Quarterly 92 (2011) 938–58 (improves attitudes towards immigrants); Andrea Louise Campbell, Cara Wang, and Jack Citrin,“Racial Threat, Partisan Climate, and Direct Democracy: Contextual Effects in Three California Initiatives,” Political Behavior 28 (2006): 129–50 (no effect).; Shayerah Ilias, Katherine Fennelly and Christopher M. Federico, “American Attitudes Toward Guest Worker Policies.” International Migration Review 42 (2008): 741–66 (improves).

[2] See, e.g., Allison Harel, Stuart Soroka, and and Shanto Iyengar, “Locus of Control and Anti‐Immigrant Sentiment in Canada, the United States, and the United Kingdom,” Political Psychology 38 (2017): 245-60).

[3] I discuss these in more detail in the forthcoming revised edition of Free to Move.

Divided: Open Borders and the Political Fracturing of Societies

Benjamin Powell and Alex Nowrasteh advance a provocative thesis: that global welfare and economic freedom would be greatly enhanced if countries implemented an open borders policy. My contention is that the problem here is analogous to that highlighted by behavioral economists: that people are not rational economic utility-maximizing “econs,” except in narrow spheres of life. Rather, they are multifaceted individuals shaped by evolutionary psychology and culture. Once we factor in this broader range of human motivation, the costs of an open borders policy rapidly mount, leading to sharp declines in human welfare and even economic freedom.

Powell makes a good point that there are efficiencies to be gained by permitting labor to flow from where it is plentiful and cheap to where it is scarce and expensive. This “productivity equals price” logic for all factors of production makes perfect sense from the perspective of neoclassical economics and global wealth maximization. I also accept that multi-ethnicity in a society makes it more difficult for unions to organize, and that immigration undercuts labor’s bargaining power. All of this reduces the ability of powerful unions and left-wing parties to regulate business or redistribute, increasing economic freedom.

The panel-based cross-country model cited in the piece confirms this mechanism by showing that economic freedom tends to increase with immigration. The authors make a claim that more immigration leads to more economic freedom. This is plausible, though the converse, that economic freedom is what leads to more immigration, is equally credible. I cannot see that this endogeneity problem has been adequately addressed. One possibility would be to take subnational units so a cross-lagged modeling strategy can be used. Another is to consider an instrumental variable for immigration, such as some measure of humanitarianism or liberalism in the culture. But let us grant that there is a relationship. Does this mean we should open the borders? No.

Why not? First, even if the only thing that matters is economics, we have to account for the persistently negative effect of ethnic diversity (a major effect of immigration) on economic growth. In their famous paper on why ethnically homogeneous East Asian economies took off relative to ethnically heterogeneous African ones, Easterly and Levine (1999) write, “High levels of ethnic diversity are strongly linked to high black market premiums, poor financial development, low provision of infrastructure, and low levels of education.” Indeed, 25 to 40 percent of the difference in economic growth trajectories between East Asia and Africa could be attributed to this factor. With increased ethnic diversification, agreement on public goods such as roads, hospitals, or schools becomes a political issue.

Alberto Alesina and colleagues (1999) showed that at county and city level in America, even when controlling for education level, income, population size, and the proportion elderly, ethnic diversity was the most important predictor of low spending on useful public goods such as roads and garbage collection, and high spending on policing and incarceration. In one natural experiment, James Habyarimana and colleagues (2007) showed that when the Ugandan government cut funds for policing, ethnically homogeneous slums in Kampala, Uganda were able to organize self-policing while heterogeneous ones were not.

Numerous studies also show a trade-off between diversity and solidarity. As Robert Putnam wrote in 2007, trust and altruism decline in more diverse societies. A meta-analysis by Dinesen, Schaeffer, and Sonderskov (2020) for over 1,000 estimates of the impact of diversity on trust for 87 studies across western societies shows that this relationship to be robust. The decline of social capital is linked to, among other phenomena, political polarization, which has been increasing in the United States, Canada, and the wider West.

Let us pause to consider the key mechanisms at work here. Immigration tends to increase ethnic diversity. Different ethnic groups often have disparate economic outcomes. People tend to be attached not only to their own fate, but to that of their ethnic communities, and their politics reflects these sociotropic considerations. This is why wealthy African-Americans tend to vote for Democrats while many poor whites vote Republican. In multi-ethnic societies such as Trinidad, Nigeria, or Northern Ireland, parties form along ethnic rather than class lines, which may blunt the degree of economic redistribution, but ethnic party politics leads to sclerotic and crisis-prone governance that is inimical to economic growth. Multi-ethnic Kenya or Tanzania may have weak unions, but they are unlikely to build the social solidarity needed to efficiently locate schools and roads and tackle corruption.

Violent conflict also becomes more common. Côte d’Ivoire imported large numbers of Muslim workers from Burkina Faso and other Sahelian countries to work on its cocoa plantations in its verdant and relatively Christian south. In 1922, Muslims formed just 6 percent of the total population, but this had reached nearly 40 percent of the total by the time of the last census in 1998, over 10 points more than the Christians. In 2000, once the country democratized, a civil war broke out between the now largely Muslim north and mainly Christian south, with southern politicians claiming that many northerners were not Ivoirian and should not be voting. Ethno-demographic shifts were not the only cause, just as the increase in Sunni Muslims in Lebanon due to the exodus of Palestinian refugees from Israel was not the only cause of Lebanon’s disastrous 1975-90 civil war. Yet these demographic shifts cannot be discounted as a major factor in such conflicts. From Xinjiang in China to Kenya’s highlands to West Papua, what David Laitin and James Fearon (2011) call “sons of the soil” conflicts have regularly broken out between natives and migrants. This is hardly conducive to investment and growth.

In the West, the rise of populist right parties is linked to higher levels of immigration. In Europe, James Dennison and Andrew Geddes’s (2019) study of 10 countries between 2005 and 2016 found a significant correlation between changes in immigration levels, the salience of immigration for voters, and populist right voting. The United States is a partial exception to this rule insofar as universalist conservatism—a fusion of the religious right, neoconservatism and fiscal conservatism—among conservative media and political elites kept the immigration issue away from the center of federal politics. But, beginning with Pat Buchanan and Pete Wilson in the early 1990s, moving through to local anti-immigration ordinances in the 2000s, and then, via the grassroots Tea Party and congressional Republicans to Trump in 2015, immigration broke through as a central issue in American politics.

The counter-claim that diverse areas like major cities do not vote for populism, and that therefore diversity has nothing to do with the rise of populism, is misleading. Once one accounts for age, education and nonwhite share, there is very little difference in support for national populism between diverse regions and homogeneous ones. London, for instance, voted to remain in the European Union, but it is only 45 percent white British and is younger and has more residents with college degrees than the rest of England. 38 percent of the capital still voted Leave, and white working-class Londoners are as likely as rural white working-class voters to have voted Leave. In Eastern Europe, it is countries located inside the EU and thus exposed to potential immigration, like Hungary, rather than outside it, like Serbia, that have seen an upsurge of national populism.

The rise of populism is a prelude to a more intractable political polarization as the basis of politics shifts from class and economics (left vs right) to culture (nationalist vs cosmopolitan). In Britain, the Conservative Party now has a more working-class composition than the Labour Party, something absolutely unthinkable from the 19th century until 1997. In the United States, the Republicans are also becoming much more working-class, while the Democrats control the wealthiest ZIP codes in the country. The reason for this change is that attitudes to immigration have become more important than views on economic redistribution for voting.

Meta analyses of immigration attitudes, notably Hainmueller and Hopkins (2014), find that people’s views on immigration have virtually nothing to do with their income or employment situation. Instead, they are shaped by deep-seated psychological preferences. Whether one views difference as disorder or as interesting, whether one views change as loss or as stimulating, is half heritable. Trying to teach or shame these inclinations out of people only leads to resentment (Stenner 2005). We thus see radically different responses to immigration within the white population of western countries. The populist response to immigration triggers a moralistic politics among what Sobolewska and Ford (2020) term “identity liberals,” many of whom are younger, urban, and highly educated. This demonization of the white working class and populist conservatives, such as Brexiteers or Trump voters, in turn produces blowback against the elite, setting in motion a toxic ratchet of accusation and counter-accusation, a zero-sum conflict with no room for compromise.

How does this affect economic freedom and growth? A secondary reflex arising out of anti-immigration populism is for populists to sour on globalization, withdraw from international institutions, and move toward isolationism in foreign policy. This is because, as Yotam Margalit (2019) has shown, hostility to immigration and increased diversity tends to shape economic attitudes such as support for free trade. This is not a straightforward relationship: Brexit reduced Britain’s ties to Europe but opened it up, potentially, to ties to non-EU markets. Yet the rise of populism seems linked to more protectionism and mistrust of multilateral institutions, which can hardly be conducive to global growth.

There are some counter trends. Electorally, the rise of populism and shift of the basis of politics from economics to culture benefits the right, with the European left recording their worst results since 1945 in country after country as the white working class shifts right. This probably weakens the left and protects economic freedom in an age of surging inequality, but we also must bear in mind that left-wing anti-immigration voters change the character of the conservative parties they join. Since 2017, Britain’s Conservative Party has embraced high public spending and the “leveling up” of poorer regions because they get more of their seats from these areas than used to be the case under Thatcher. In the United States, fiscal conservatives with their tax cuts still dominate in Congress, but Trump did particularly well among left-leaning anti-immigration Republicans who support Social Security, Medicare, and infrastructure spending. Do not discount the possibility of a more Michael Lind/Theodore Roosevelt-style populist Republicanism emerging in the future and tacking left on economics but right on culture.

I am more of a believer in the virtues of a welfare state than Powell and Nowrasteh. I also think a world in which we must be as attached to strangers as to those we have special ties to is a worse one than the one we currently inhabit. But even if my objective function was merely to maximize global wealth and economic freedom, I think open borders is the wrong move. It will result in increasingly fractious host societies, stimulating populism, polarization, protectionism, and even civil violence.

There is no getting away from the fact that immigration levels in modern societies with welfare states and inclusive citizenship must be calibrated to the rate of assimilation. Only where immigration is relatively similar in ethnic composition to the host population or where native-born fertility is high enough to minimize the rate of change, as in the United States during the 1776-1840 and 1860-90 periods, can higher immigration work. It took 70 to 80 years for the ethnic diversity of the America of 1900 to fully melt. If anything, this suggests that immigration levels should fall rather than rise in the coming decades so that assimilation can reduce diversity to levels more conducive to national solidarity. When that process succeeds, a more generous policy can be considered once again.


Alesina, A., et al. (1999). “Public Goods and Ethnic Divisions.” Quarterly Journal of Economics 114(November): 1243–1284.

Dennison, J. (2019). “How Issue Salience Explains the Rise of the Populist Right in Western Europe.” International Journal of Public Opinion Research 32(3): 397-420.

Dinesen, P. T., et al. (2020). “Ethnic diversity and social trust: A narrative and meta-analytical review.” Annual Review of Political Science 23: 441-465.

Easterly, W. and R. Levine (1997). “Africa’s Growth Tragedy: Policies and Ethnic Divisions.” Quarterly Journal of Economics 111(4): 1203-1250.

Fearon, J. D. and D. D. Laitin (2011). “Sons of the Soil, Migrants, and Civil War.” World Development 39(2): 199-211.

Habyarimana, J., et al. (2007). “Why does ethnic diversity undermine public goods provision?” American Political Science Review 101(4): 709-725.

Hainmueller, J. and D. J. Hopkins (2014). “Public Attitudes Toward Immigration.” Annual Review of Political Science 17: 225-249.

Margalit, Y. (2019). “Economic insecurity and the causes of populism, reconsidered.” Journal of Economic Perspectives 33(4): 152-170.

Sobolewska, M. and R. Ford (2020). Brexitland: Identity, Diversity and the Reshaping of British Politics, Cambridge University Press.

Stenner, K. (2005). The authoritarian dynamic, Cambridge University Press.

What We Can Learn from Our Relatively Open Borders

Alex Nowrasteh and Benjamin Powell’s immigration research offers a decidedly fresh, invigorating perspective on a critical policy debate that has grown increasingly stale as it has become more and more salient in our politics. Yet I am doubtful how much their many insights will help Americans grapple with this perennially fraught issue. At a time when economic inequality has emerged as one of the more pressing concerns on the national agenda, our authors embrace what they identify as a utilitarian perspective on free immigration’s positive impact on economic efficiency. Yet they have little at all to say about how the economic gains they argue would follow from unconstrained immigration would be distributed. Or as the nineteenth-century British utilitarian Henry Sidgwick might ask: Would the anticipated increase in total utility from unrestricted immigration also result in an increase of average utility—or at least not a diminution of the utility of the average American? Additionally, and in light of our nation’s overdue appreciation of the continuing plight of African Americans, would the presumed increase in total utility from unrestricted immigration benefit them? Obviously not easy questions to consider, but also ones that our authors do not address.

Further, I would challenge Nowrasteh and Powell to elaborate on their postulate that governmental expenditures are inevitably and uniformly a threat to economic freedom. Putting aside the growth of such expenditures resulting from the New Deal’s response to the Depression, World War II certainly resulted in massive and costly governmental growth that was undeniably in defense of economic freedom. So were programs like the Marshall Plan in Western Europe and the G.I. Bill here at home. Then, too, the Cold War and nuclear competition with the Soviet Union required massive governmental expenditures that eventually led to greatly increased economic—as well as political—freedom around the world.

Turning to Nowrasteh and Powell’s intriguing case studies—what they refer to as quasi-natural experiments—of sudden, unpredicted, but delimited mass immigration to Israel and to Jordan, the relevance of their findings to the American context is questionable. After all, neither of these two small nation-states is comparable in size, history, or longevity to the United States. Nor, even more importantly, is the arrival and integration of Russian Jews in Israel after the demise of the Soviet Union comparable to the fate, say, of millions of Spanish-speaking immigrants to the United States during the same period. Not only do such Russians share an important cultural heritage—Judaism—with their new countrymen, but their embrace by Israel, however ambivalent, is nevertheless a critical component of the state’s raison d’etre.

In fact, the departure of Russian Jews from their home country and their arrival and settlement in Israel look more like what refugees experience than immigrants. And this is even more evident with the Kuwaiti Palestinians who arrived in Jordan in 1990 at the outbreak of the First Gulf War. Although the distinction between refugees and immigrants has typically and purposely been blurred by advocates seeking to augment the visibility and clout of both, the divergent circumstances and goals of two such populations inevitably impact outcomes. Notable among these: immigrants typically seek what the authors refer to as economic freedom; refugees by definition seek physical safety and political liberty. Such motives must impact the outcomes pointed to by our authors. Yet in their analyses they fail to acknowledge them.

Following up on the authors’ quasi-natural experimental approach, I would point to an important case that they overlook and is arguably more relevant than the Israeli or Jordanian experiences. I refer of course to America’s de facto experiment with illegal (or undocumented) immigration. For about thirty-five years now, since passage of the Immigration Reform and Control Act (IRCA) of 1986, the United States has in effect been experimenting with lax enforcement of its immigration laws. To be sure, one might widen this window to include the period between passage of the Immigration and Naturalization Act of 1965 and the 1986 measure, during which time immigration law enforcement was extremely lax and illegal immigration gradually but steadily increased—such that it surfaced in earnest in Washington during the Carter administration. The United States has experienced some illegal immigration for much of its history. But it has never risen to the numerical levels or political salience of the current period. This is not to suggest that as a nation we have de facto open borders. We clearly do not.

But neither have we adopted the one tool that would undeniably curtail illegal immigration: a blanket program of meaningful sanctions on those employing the undocumented. The laws that are on the books are sporadically and unevenly enforced both geographically and temporally. And yes, they are enforced on some groups more than others. Hispanics are undoubtedly under more scrutiny from immigration enforcement than South Asians. But then, there are proportionately more undocumented among Hispanics here than among South Asians. In any event, for about thirty years now the United States has been home to from 10 to 12 million or more undocumented immigrants, many of whom have eschewed the opportunity to pursue the path to legal status and eventual citizenship.

For better or worse, the millions of illegal immigrants have been the focus of much of our episodic immigration debate. And I would argue that our unresolved reckoning with undocumented immigration has been a surrogate for a broader, more substantive debate about our immigration policy more generally. Stop and consider the complaints one has been hearing for years now about the undocumented from many Americans: (1) they have broken the law by arriving or staying here without authorization; (2) they are a fiscal burden on schools, hospitals, and other locally based and funded services; (3) they are taking jobs away from native-born workers; (4) they do not speak and are not learning English; and (5) they are disproportionately involved in criminal activities.[1] With the exception of the first point, I would point out that these complaints—some of which have some validity, others not—are similar, if not identical, to those lodged against immigrants generally, especially poorly educated, unskilled immigrants and their families.

I do not believe that the ubiquitous and inevitable distinction between legal and illegal immigration is as valid or as helpful as it is widely taken to be. In field research and interviews with border patrol agents, I have invariably heard them express two divergent perspectives. First, they insist defensively on their status as “federal law enforcement agents,” who are not only armed (a factor in determining status and pay) but just as professional and critical as colleagues—such as FBI or DEA agents—who enjoy much greater recognition and prestige than they. Second, border patrol agents—virtually to a man, or a woman—at some point volunteer, usually while sitting in a van and talking late at night on patrol, that if they were in the migrants’ shoes, they, too, would be attempting to enter the United States illegally in order improve things for themselves and their families!

Ah, but there’s the rub! Border and other immigration enforcement personnel lack status in the federal law enforcement hierarchy—and suffer from burnout and attrition more than their fellow federal agents, say, at the DEA—precisely because the laws they enforce are not considered, even by themselves, as commensurate with those enforced by other federal agencies. As legal scholars put it, presence in the United States without appropriate permission and documentation is not a malum in se, but rather a malum prohibitum—not an inherent moral wrong or crime, but a statutory violation.

However much elected officials and nervous bureaucrats lean on the sharp dichotomy between documented and undocumented immigrants to deal with this most intractable of issues—and however much large numbers of Americans consequently embrace this distinction—the fact is that problems associated with the undocumented are also much in evidence among poorly educated, unskilled legal immigrants and their families. Nevertheless, the complexity and volatility of immigration policy and politics have resulted in our elites needing to simplify the framing of the issue to something like: legal immigrants are good; illegal immigrants are bad. And while this distinction has itself come under scrutiny and criticism from immigrant advocates who defend the undocumented and thereby attempt to blur the distinction (“There are no illegal human beings!”), the bright line drawn between legal and illegal immigration, however inadequate and misleading, remains useful to the participants in this contentious debate.

My point is that for more than half a century, the United States has been conducting an experiment in relatively open borders. Once again, this has been nothing like the extreme open-borders policy that our authors pose as the null hypothesis. After all, there are millions of would-be immigrants patiently waiting in line for years overseas to obtain visas and join their families already here. Then, too, tens of thousands of immigrants get deported every year. Nevertheless, the resulting chaos and lack of attention to promoting the settlement and assimilation of migrants—legal as well as illegal—approximates the kind of experiment in open borders more than our authors acknowledge. And the outcome is obviously unacceptable to many Americans whose anxieties and concerns may be poorly or offensively articulated, but are not without foundation.

Finally, it is critical to consider the political context in which these migration dynamics have been playing out. While the authors address the impacts of immigrants on our political as well as economic institutions, their understanding of the former is limited and overlooks important changes since the last wave of immigrants ended during the 1920s. An ephemeral but nevertheless revealing example is the rhetoric from globalization-obsessed elites who have been fueling discourse about “open borders” and “a borderless globe” since the fall of the Berlin Wall. Here the desired—and hardly unsuccessful—goal is typically to draw attention away from “illegal immigrants” and toward a borderless world in which “no one is illegal.”

A more pertinent example of critical institutional change overlooked by our authors is the disappearance of the state and local party organizations that skillfully wielded ethnic appeals and patronage to draw immigrants and their children into politics. Today, those institutions have been replaced by media-driven, racially framed claims that in the wake of the civil rights movement imitate the political posture of African Americans. The upshot is that Hispanics, as well as other newcomers, have been conditioned to mobilize themselves not as newly arrived immigrants, but as long-suffering, racially oppressed minorities who merit remedies similar to those afforded, however begrudgingly, to black Americans. One outcome is the oft-repeated mantra that America is becoming “a majority-minority nation,” which has hardly been welcome news among many already marginalized white Americans. All the more ironic, then, is the evidence that most immigrants, Hispanics in particular, conform to the broad pattern of upward mobility and assimilation displayed by previous migrants from Europe. Indeed, half of Hispanics consistently identify themselves as racially white, and they intermarry with white Americans in ever increasing numbers.

These critical changes in our political institutions escape the attention of Nowrasteh and Powell. To be sure, their primary focus is on immigration’s impact on economic institutions. But given their broader concerns to promote dramatically less restrictive immigration policies, it behooves them to consider not only such specific developments in our politics, but more generally how political institutions shape and influence how economic outcomes are defined and interpreted. For as David Hume noted, men fight for their interests, but their interests are a matter of opinion.


[1] See Ruben G. Rumbaut and Walter A. Ewing, “The Myth of Immigrant Criminality and the Paradox of Assimilation: Incarceration Rates among Native and Foreign-Born Men,” Immigration Policy Center (Washington, DC: Spring 2007). Also Noah Pickus and Peter Skerry, “Good Neighbors and Good Citizens: Beyond the Legal-Illegal Immigration Debate,” in Carol M. Swain, ed., Debating Immigration, Second Edition (NY: Cambridge University Press, 2018), 194-216. And Peter Skerry, “Immigration and Social Disorder,” in Norton Garfinkle and Daniel Yankelovich, eds., Uniting America: Restoring the Vital Center to American Democracy (New Haven: Yale University Press, 2006), 124-138.

The Conversation

Fractionalization, Trust, and Populist Backlash

I thank Ilya Somin and Eric Kaufmann for their thoughtful response essays (this post was written before Peter Skerry’s response was published). Both of them raise a number of interesting points. Not surprisingly, I find myself mostly in agreement with the points raised by Ilya, so this response expands on three issues—fractionalization, trust, and right-wing populist backlash—that Eric raised, though I will reference some of Ilya’s points in the process.

Immigration and Fractionalization[1]

I agree with Eric Kaufmann that unrestricted immigration will result in more ethnic and linguistic diversity in destination countries. However, I question whether this would lead to worse outcomes. He cites the well-known papers of Easterly and Levine (1997) and Alesina et al. (1999) that show that increased ethnolinguistic fractionalization is associated with lower spending on publicly funded goods and services and lower economic growth. However, Banting and Kymlicka (2006) point out that most of the evidence on fractionalization (including these two papers) comes from sub-Saharan Africa and the United States. In Sub-Saharan Africa, the fractionalization was mostly created by international interventions that imposed national borders on populations who otherwise would not have chosen political unifications with each other. In the United States, much of the fractionalization comes from African Americans whose ancestors were brought to the country as slaves. Immigration, on the other hand, increases fractionalization when people choose to go to a particular destination country. If immigrants, at least in part, select destination countries based on their institutions of governance, then there is less reason to believe that immigration-driven fractionalization would have the same effects as the fractionalization in the cited studies.

Furthermore, even if immigration-induced fractionalization did reduce government spending on publicly funded goods and services, one could argue that the result would be a beneficial increase in economic freedom. Alesina and Glaeser (2004) argue that fractionalization and ethnic heterogeneity are the main reasons that the United States has a smaller welfare state than most western European countries. The clear implication is that if immigration leads to greater heterogeneity it should shrink welfare states. Similarly, Razin et al. (2002) propose a median voter model that relies on relative income positions, rather than ethnic fractionalization, to predict that native-born taxpayers will shift their preferences away from high-tax, high-benefits welfare policy more than immigrants, who join the pro-tax, pro-benefits coalition at the bottom of the income distribution. They study 11 European countries from 1974 to 1992 and find that a larger share of low-education immigrants in the population leads to smaller social transfers and lower rates of taxation on labor. The overwhelming bulk of the scholarly literature measuring the consequences of economic freedom that finds that increases in economic freedom are beneficial, not harmful (Hall and Lawson 2013).

However, ethnic fragmentation may impact governance institutions other than welfare state spending. Easterly and Levine (1997) find a negative relationship across countries between ethnic diversity and the shares of government-provided goods such as schooling, electricity, roads, and telephones. Similarly, Alesina et al. (1999) find a negative correlation in U.S. cities, metropolitan areas, and counties between ethnic fragmentation and shares of spending on government provided goods and services such as trash pick-up, roads, sewers, and education. These findings also could be interpreted as support for the view that government will be smaller, and thus economic freedom higher, when there is greater fractionalization.

If ethnic fractionalization were to harmfully impact economic freedom it would have to come through undermining the security of property rights and the rule of law. But securing property rights and the rule of law, unlike most of the publicly funded goods and services studied in the fractionalization literature, are largely non-rival public goods rather than scarce rival publicly funded goods that a fractionalized population has to agree how to allocate. It would be a mistake to assume that fractionalization induced lower public spending on rival goods necessarily translates into less ability to protect property rights and maintain the rule of law.

Immigration and Trust

Kaufmann raises an important point that greater diversity is associated with lower levels of generalized trust. Perhaps immigration induced diversity lowers economic growth by lowering trust. I didn’t address this point in my opening essay but Alex Nowrasteh and I examine immigration’s impact on trust at length in chapter 8 of Wretched Refuse. We find little reason to believe that immigration will harm either economic growth or institutions of economic freedom through decreased trust.

First, there is much evidence that increased diversity diminishes trust, but the effects are small and confined mostly to racial differences between blacks and whites, not ethnic differences caused by immigration (Alesina and La Ferra 2002). Furthermore, Robert Putnam’s work that Kaufmann references finds that the impact of ethnic diversity on trust is fairly small.

Second, generalized trust is more important for economic activity when formal institutions are weak and unreliable. If people think that the courts will not enforce their contracts and protect their property, then generalized trust can partially substitute as a mechanism that enables people to realize gains from exchange. However, the flip side is that generalized trust is much less important if people trust courts and the government to enforce their contracts and protect their property. Generalized trust has the greatest positive effect on income when institutional strength is low, while the positive effect of trust shrinks when institutions are stronger (Ahlerup et al 2009). Since people tend to immigrate to places that have stronger formal institutions, any immigration induced decrease in generalized trust becomes less important.

Furthermore, even if immigrants have lower levels of generalized trust, they have higher trust in formal institutions. Alesina and La Ferrara (2002) show that individual Americans living in more racial fragmented communities have a lower propensity to trust other people but do not have lower levels of trust toward formal institutions. In Wretched Refuse Alex and I show that while immigrants have lower levels of generalized trust than native-born Americans, they have higher levels of trust in our courts and legal system, the three branches of the national government, and in business.

In Wretched Refuse we summarized our overall assessment of the relevance of trust as:

1) The overall evidence of a causal link between social trust and growth is not well established. 2) Whether trust promotes good institutions that are important for economic growth or whether good institutions promote trust is not clear. 3) To some extent trust and good institutions likely substitute for each other in promoting the realization of gains from trade that lead to higher productivity and growth. To the extent that destination countries have good institutions, this makes the question of whether immigrants create lower levels of generalized trust less important. Thus, although much studied, in our opinion the social trust literature and its link to immigration does not undermine the standard case for free immigration.

Immigration and Right-Wing Populism

Does immigration lead the native-born population to embrace right-wing populist platforms that could undermine economic freedom and growth? Ilya Somin offered some thoughts on the problem of nativist backlash in his essay. He argued three points, 1) because nativists overestimate the number of immigrants that they are unlikely to be pacified by reductions in immigration; 2) that there are better “keyhole” ways to address nativist concerns than restricting immigration; and 3) that its morally perverse to restrict immigration to prevent nativist backlash. I’m largely in agreement with all three points, so I will confine myself to raising a couple of more points in response to Eric Kaufmann’s concern that immigration will fuel right-wing populism that rejects globalization. My points will focus specifically on economic freedom.

Kaufmann cites the Dennison (2020) study that found a significant correlation between changes in immigration levels, the salience of immigration for voters, and populist right voting from 2005 to 2016.[2] What is the impact of that shift in voting? Did it reduce economic freedom? Dennison studied 11 European countries that had an average economic freedom score of 7.8 (out of 10) in 2005. The economic freedom of these countries was essentially unchanged at an average 7.77 in 2016. The average is not hiding any significant changes. No country’s score in Dennison’s data changed by more than 0.21 points. In short, even if immigration increased populist voting, that voting did nothing to decrease countries’ economic freedom. This could be because immigration induced populism does little to change economic policies or it could be because some populist policies increase economic freedoms while other populist policies decrease economic freedoms and these changes roughly offset each other.

What about populist right-wing policies that specifically reject globalization? What are these policies likely to be? The three most important are 1) increased immigration restrictions 2) increased tariffs, quotas, and barriers to international trade, and 3) a more isolationist foreign policy. On the first of these, it would be perverse to impose immigration restrictions now to prevent right-wing populists from doing it later. As to the second, the world is much closer to free trade in goods, services, and capital mobility, than it is in free trade in labor. So, even if the price paid for freer immigration is less free trade on these other margins, the world comes out ahead because the economic gains from greater free trade in labor will almost certainly dwarf the economic losses caused by decreased freedom to trade goods and capital. Finally, since war is the health of the state, an isolationist foreign policy is more likely to promote economic freedom by limiting the size and scope of the state than it is to undermine economic freedom.

I agree that right-wing populist reactions to greater immigration could be problematic but efforts at controlling that backlash should not include quantitative restrictions on immigration.


Ahlerup, P., Olsson, O., and Yanagizawa-Drott, D. (2009). Social capital vs. institutions in the growth process. European Journal of Political Economy, 25(1), 1-14.

Alesina, A. F., Baqir, R. and Easterly, W. (1999). Public goods and ethnic divisions. Quarterly Journal of Economics, 114: 1243-84.

Alesina, A. F. and Glaeser, E. L. (2004). Fighting poverty in the US and Europe. New York: Oxford University Press.

Alesina, A., La Ferrara, S. (2002). Who trusts others? Journal of Public Economics, 85, 207-34.

Banting, K. and Kymlicka, W. (2006). Introduction: Multiculturalism and the welfare state: Setting the context. In Multiculturalism and the welfare state, edited by Banting, K. and Kymlicka, W. New York: Oxford University Press, pp. 1-45.

Clark, J. R., Robert Lawson, Alex Nowrasteh, Benjamin Powell, and Ryan Murphy. (2015) “Does Immigration Impact Institutions.” Public Choice 163: 321-335.

Dennison, J. (2020). “How Issue Salience Explains the Rise of the Populist Right in Western Europe.” International Journal of Public Opinion Research 32(3): 397-420.

Easterly, W. and Levine, R. (1997). Africa’s growth tragedy: Policies and ethnic divisions. Quarterly Journal of Economics, 112: 1203-1250.

Hall, J, and Lawson, R. (2013). Economic freedom of the world: An accounting of the literature. Contemporary Economic Policy, 32: 1-19.

Razin, A., Sadka, E. and Swagel, P. (2002). Tax burden and migration: A political economy theory and evidence. Journal of Public Economics, 85: 167-90.


[1] Parts of this section draw on my prior work in Clark et al. (2015).

[2] Kaufmann’s essay incorrectly cites the paper as Dennison and Geddes (2019) in the text but lists the correct paper in the references. The correct citation is Dennison, J. (2020). “How Issue Salience Explains the Rise of the Populist Right in Western Europe.” International Journal of Public Opinion Research 32(3): 397-420.

What We Can Learn from the United States’ Mostly Closed Borders

Thanks to Peter Skerry for also providing a thoughtful response to my lead essay and Wretched Refuse more generally. Peter raises a number of important issues that are not directly addressed in Wretched Refuse or my lead essay. Wretched Refuse is explicitly an empirical examination of the new economic case for immigration restrictions, not an examination of all issues that could stem from increased immigration. My lead essay focuses on one aspect of that examination—immigration’s impact on economic freedom.

However, my focus on how immigrants impact formal and informal institutions that impact productivity in destination countries does implicitly and indirectly address many of the issues raised by Peter. If, as we argue in Wretched Refuse, immigration does not harm institutions responsible for high productivity, then the overall economic gains from greatly liberalized immigration are massive. So massive, in fact, it’s hard to not find ways for public policy to redirect some of the massive gains to address other issues—fiscal, distributional, infrastructure adjustments, and so on. So-called “keyhole” solutions in Ilya’s response essay. If that is done, my work with Alex is not just a utilitarian perspective. Elsewhere, the case has been made that if the forecast gains from liberalized immigration are anywhere near accurate, then every major ethical framework recommends unrestricted immigration (Caplan and Naik 2015; Caplan and Weinersmith 2019; Powell and Brennan 2021).

Accordingly, my response here focuses on the aspect of Peter’s essay that is most directly relevant for assessing how immigration could impact the institutions responsible for our prosperity. How has recent immigration impacted economic freedom in the United States?

First, however, we need to be clear, that contrary to his title “What We Can Learn from Our Relatively Open Borders,” the United States has nothing close to an “open border,” if by that we mean an essentially unrestricted quantity of immigrants entering. Peter oscillates in his essay about how “open” the U.S. border is, from stating plainly that “we clearly do not” have a de facto open border to later saying that the United States “has been conducting an experiment in relatively open borders” in recent decades. So, let’s clear up how restricted immigration to the United States actually is. The fact that somewhere between 11 and 13 million foreign-born people reside in the United States illegally does not mean that the border is relatively open. It isn’t hard to see the evidence that the border is mostly closed. Otherwise, why would so many people be on waitlists (sometimes decades long) to legally migrate to the United States? They would “simply” come illegally. Of course, there is nothing simple about coming illegally, otherwise why would coyote fees for smuggling people across the border be so high? The high smuggling fees are an indication of how closed the border is.

More scientifically, we can turn to the wage differentials between the United States and immigrants’ origin countries. It’s hard to find any labor market on the planet where there are real wage differentials much greater than 1.5 across geographic regions where works are free to relocate. Clemens and coauthors’ (2019) study of real wage differentials between the United States and 42 poorer immigrant origin countries found wage differentials for identical workers (controlling for gender, age, education, etc.) of 3.95 between the United States and the median origin country. The differences ranged from a high of 16.4 to a low of 1.7 among the 42 countries. If immigration to the United States was relatively open, these wage differentials would not exist.

That said, even with a mostly closed border we can still learn something about the impact of immigration on U.S. institutions related to our productivity. In my lead essay, I highlighted some of what I think we can learn from when the United States had largely unrestricted immigration until 1921, its most closed borders from 1921 to 1965, and the intermediate degree of closedness from 1965 to the present. Peter wants to focus more on what we can learn from this most recent period.

The number of foreign-born people in the United States has increased substantially over the last 30 to 40 years, and the distribution of those people varies tremendously across U.S. states. Most immigrants live in California, Florida, New Jersey, New York, or Texas. Between 1980 and 2010, the average share of a state’s population that was foreign born was 7.6 percent, but the standard deviation was almost as large, at 6.5 percentage points. A few papers have taken advantage of this variation to examine how recent immigration has impacted U.S. state-level economic freedom.

Padilla and Cachanosky (2018) look at 10-year panels from 1980 through 2010 and employ pooled OLS, fixed-effect OLS, and system GMM methods to analyze how initial shares of foreign-born adults or naturalized citizens impacted state-level economic freedom a decade later. They control for initial levels of freedom, so that they are essentially using initial stocks of immigrants to examine how they impact the change in economic freedom. In follow up papers Padilla and Cachanosky (2019) follow a similar methodology but then separate the foreign born into four different quartiles based on the economic freedom in their origin country to see if immigrants from less free origins might have a different impact than those from more free origin countries. Similarly, their third follow up paper (Padilla et al. 2019) divides the foreign born by educational attainment to see if less educated immigrants are detrimental to economic freedom. Tuszynski and Stansel (2019) is closely related to the Padilla and Cachanosky studies but attempts to better deal with endogeneity by using two decade-lagged stocks of immigrants. Yao et al (2019) take a different approach and use a synthetic control methodology (like we did for Israel and Jordan) to see the impact the Immigration Reform and Control Act of 1986 (IRCA) had on state level economic freedom. After reviewing each of these studies in much greater detail in Chapter 5 of Wretched Refuse, Alex and I summarize their overall findings:

Taken as a whole, the group of papers studying the effect of immigration on state-level economic freedom is not as positive as our findings for how immigrants affect country-level economic freedom. However, the US state-level research does not find any compelling systematic evidence that greater immigration erodes state-level economic freedom either. Results are often insignificant or weakly significant at the 10 percent level and, even when they are statistically significant, their magnitude is small. The statistically significant changes reviewed here would rarely result in any state changing its economic freedom ranking among the other states if it received a one standard deviation increase in its immigrant share. Taken together with the study of IRCA, the overall results in this section suggest essentially no meaningful relationship between immigration and US state-level economic freedom (Nowrasteh and Powell 2020: 94).

This finding is largely consistent with what we should expect from surveys on the policy opinions of the foreign-born in the United States and their political participation rates. Alex’s coauthored work with Sam Wilson found that naturalized immigrants (those eligible to vote) and descendants of immigrants had policy views not statistically different from the native-born on government spending on welfare programs, Social Security benefit levels, government spending on the environment, and government programs to reduce income differences. The opinion differences on these issues shown by non-citizen immigrants was larger but still not statistically significant (Nowrasteh and Wilson 2017).

In his recent book, Bryan Caplan (Caplan and Weinersmith 2019) also examines the policy opinions of immigrants using the General Social Survey and finds that immigrants and the native born are equally hostile to taxes on the poor and middle class and, if anything, immigrants are more hostile to taxes on the rich than the native-born. In contrast to these more general results, he finds that immigrants without a high school degree are more in favor of greater government regulation of the economy and spending. However, there are good reasons why these policy views might not influence actual policy as measured by economic freedom. First, many of these low-skilled immigrants are not eligible to vote. Second, immigrant voter turnout rates (48 percent of those eligible) are lower than native-born turnout rates (72 percent), and voter turnout rates among low-education immigrants eligible to vote are even lower (27 percent).

To sum up, although the United States does not currently have anything close to unrestricted mass immigration, I do think we can learn something from how recent immigrants have impacted U.S. institutions of economic freedom. So far, what we’ve learned is that they don’t have much impact on state-level economic freedom.


Caplan, B. and Naik, V. (2015). A Radical Case for Open Borders. In B. Powell, eds., The Economics of Immigration: Market-Based Approaches, Social Science, and Public Policy. New York, NY: Oxford University Press.

Caplan, B. and Weinersmith, Z. (2019). Open Borders: The Science and Ethics of Immigration. New York, NY: First Second.

Clemens, M., Montenegro, C. and Pritchett, L. (2019), The place premium: Bounding the price equivalent of migration barriers. The Review of Economics and Statistics, 101 (2): 201-213.

Nowrasteh, A. and Powell B. (2020). Wretched Refuse: The Political Economy of Immigration and Institutions. Cambridge: Cambridge University Press.

Nowrasteh, A. and Wilson, S. (2017). Immigrants Assimilate into the Political Mainstream. Economic Development Bulletin No. 27. Washington DC: Cato Institute.

Padilla, A. and Cachanosky, N. (2018). The Grecian horse: Does immigration lead to the deterioration of American institutions? Public Choice, 174(3-4), 351-405.

Padilla, A. and Cachanosky, N. (2019). The Grecian Horse II: Do Immigrants Import Their Home Country’s Institutions Into Their Host Countries? The Case of the American States.” SSRN.https://ssrn.com/abstract=3316415.

Padilla, A. and Cachanosky, N, and Beck, J. (2019). Immigration and Economic Freedom: Does Education Matter?. Journal of Private Enterprise. Forthcoming.

Powell, B. and Brennan, J. (Forthocming) The Ethics of Doing Business with Illegal Immigrants. Independent Review.

Tuszynski, M, and Stansel, D. (2019). Examining the Relationship between Immigration and State Institutions: Does Region of Origin Matter? Working Paper.

Yao, L., Bolen, B. and Williamson, C. (2019). The Effect of Immigration on US State-level Institutions: Evidence from the Immigration Reform and Control Act. Working Paper.

Immigration Restriction and Liberal Democracy

I appreciate Benjamin Powell’s considered response to my piece on the risks of ethnic fractionalization, and a subsequent response to Peter Skerry’s important essay. I think Powell may be correct in the narrower claim that economic freedom is not affected by immigration. Indeed, most of the studies cited, including their own (“no meaningful relationship between immigration and U.S. state-level economic freedom”), seem to report as much. But this raises two issues. First, this is a very different claim from one that argues that immigration will increase economic freedom. Second, the ad hoc support for “keyhole” policies and the rationale that restricting immigration to prevent backlash is “morally perverse” are not properly justified in empirical or normative-theoretical terms.

Before proceeding, I would note in response to Powell’s reply that the work of Alesina et al. (1999) explicitly noted that their findings held even when controlling for African-American share, so this objection to the diversity-solidarity argument is not, in my view, convincing. In Europe, studies have generally replicated the diversity-trust findings, and here the groups in question are entirely immigrant in provenance rather than a native minority, such as African-American. The cross-sectional relationship between ethnic fractionalization and economic outcomes was also, in my view, not persuasively addressed. It may be the case that immigrant diversity going forward will not affect social solidarity and the political economy of growth, but we need to know why. Given the clear inverse relationship in the cross-sectional data, the onus in my view is on the free movement side to explain why the same relationship will not obtain in the future. While the causal mechanism from ethnopolitics to political sclerosis to lower growth is more indirect, the aggregate data is powerful. Perhaps a positive effect on economic freedom and per capita income is more demonstrable when the scope conditions are limited. That is, based on only admitting a certain kind of immigrant (high skilled?) or that the economic impact is only positive for a certain kind of host society, such as Dubai (where citizenship rights are limited to a restricted group, and there is only a minimal welfare state for noncitizens) could be demonstrated.

But let me return to my main argument as set out in the introduction. Powell and Somin’s normative argument for open borders, unlike Powell’s empirical claim that immigration does not reduce economic freedom, seems to rest on narrower ground. The claim that all ethical frameworks support free movement is not correct. Numerous ethical approaches, such as liberal nationalism (Miller 2016) or communitarianism, support the case for restriction. For Michael Walzer (1983), there can be no communities without boundaries, these restrictions are necessary to protect communal solidarity. For Chris Wellman (2008), nations are free associations that have the right to associate or not associate with outsiders (i.e. accept immigrants) the way they have the right to enter and leave supranational organizations or trade deals. The relations between a country and the rest of the world with regard to admitting entrants are associational, similar to those governing relations between the Sons of Italy and the rest of American society. There is no right to migrate because there is no global political community and social contract akin to that between a citizen and their national government.

Restriction of immigration is not a violation of individual liberty or equal treatment as understood in classical liberal theory, which was nationally bounded (cosmopolitan theory takes a different view, of course). By contrast, free movement rules that override the will of the demos are undemocratic unless the national democracy supports these frameworks. Attempting to implement supranational policies by stealth, or in a fashion that bypasses democratic assent, is in my view unethical. The populism that is ascendant in our time is, in important part, a response to what many see as the insufficiently consultative and transparent character of such processes. Populism brings neglected views to the democratic conversation. However, it presents a risk in that it opens the door to an indiscriminate anti-elitism and can stoke anti-minority sentiment and a politics of unreason. This means it is important not to aggravate it unless elites are on solid normative ground and have first attempted to bring public opinion with them. That is, “moral leadership” is only appropriate where populism’s claims are normatively repugnant, as with George Wallace’s segregationist third party movement which would deny equal rights to African-American citizens.

Such is not the case with immigration restriction, which is linked to the desire of many citizens to slow the rate of cultural change to their national “imagined community” (Anderson 1983). Immigration attitudes have very little to do with material considerations and are largely linked to psychological, cultural, and ideological dispositions (Hainmueller and Hopkins 2014). Attitudes are also only modestly shaped by local experiences. This is why “keyhole” approaches, such as moving funds for public services to high-immigration locales or redistributing income to poorer natives, are unlikely to make a difference to public sentiment.

As I have pointed out elsewhere (Kaufmann 2019a), attachment to a community—even a specific version of community—cannot be conflated with hostility to outsiders (i.e. racism, xenophobia). As decades of research has shown (Brewer 1999), these orientations are generally not correlated except in times of conflict, such as civil war, yet advocates of open immigration persistently elide these distinct motivations to label them racist. Whether this stems from a position of ignorance or dishonesty is unclear—I am inclined to give people the benefit of the doubt, so assume it is the former. Yet I am also aware that the overuse of the charge of racism in public discussions of immigration prior to 2015 helped silence mainstream politicians, opening up a market for populist parties like the Sweden Democrats or leaders like Donald Trump who were willing to violate taboos. Is this really who we want leading the conversation over immigration levels?

From a deontological point of view, restricting immigration to a level compatible with the democratic wishes of the population is the right policy. From a consequentialist standpoint, doing so to preserve social harmony and reduce political polarization, so long as rights are respected, is the ethical position. Should assimilation proceed as it has in the past, conservative and order-seeking people (Stenner 2005) are likely to accept higher immigration once again. In survey experiments, mention of assimilation and retention of a country’s character reduces opposition to immigration while extolling the benefits of diversity, as Stenner predicts, arouses resistance (Kaufmann 2019b; Ekins and Kemp 2020). The comparatively high immigrant-origin diversity of western countries today—excepting the American experience of 1840-56 and 1890-1924 when similar populist pressures emerged—is mobilizing a significant number of citizens to vote for reduced numbers. There is nothing perverse here, just the workings of a democracy with the right to freely associate with citizens of other states, and which is fully in accord with the principles of liberalism.


Alesina, A., et al. (1999). “Public Goods and Ethnic Divisions.” Quarterly Journal of Economics 114(November): 1243–1284.

Anderson, B. (1983). Imagined Communities: Reflections on the Origin and Spread of Nationalism. London, Verso.

Brewer, M. B. (1999). “The psychology of prejudice: Ingroup love and outgroup hate?” Journal of social issues 55(3): 429-444.

Ekins, E. and D. Kemp. (2020). “What Increases Public Support for Immigration? Results from a New Experiment,” Cato Institute, December 29.

Hainmueller, J. and D. J. Hopkins (2014). “Public Attitudes Toward Immigration.” Annual Review of Political Science 17: 225-249.

Kaufmann, E. P. (2019a). Whiteshift : populism, immigration, and the future of white majorities. New York, Abrams Press.

Kaufmann, E. (2019b). “Can narratives of white identity reduce opposition to immigration and support for Hard Brexit? A survey experiment.” Political Studies 67(1): 31-46.

Miller, D. (2016). Strangers in our midst, Harvard University Press.

Stenner, K. (2005). The authoritarian dynamic, Cambridge University Press.

Walzer, M. (1983). Spheres of Justice: A Defence of Pluralism and Equality. Oxford, Blackwell.

Wellman, C. H. (2008). “Immigration and freedom of association.” Ethics 119(1): 109-141.

Immigration Advocates Must Consider Virtue, Not Just Economics

Benjamin Powell and Alex Nowrasteh entitle their book Wretched Refuse? The Political Economy of Immigration and Institutions. Yet at no point do they define or explain what they mean by “institutions,” a topic which social scientists have long debated. Nor do our authors acknowledge that despite invoking the label “political economy,” their analysis relies on the same tools and methods as standard neoclassical economics. And while they address broad concerns about the relationship between immigration—past, present, and future—and topics such as corruption, terrorism, and cultural change in America, the fundamental thrust of their argument is, as the title of their second chapter suggests, “An Economic Case for Free Immigration.”

After reading Powell’s responses to Somin, Kaufmann, and myself, my question is: what exactly are he and his coauthor purporting to demonstrate? Is theirs an analysis of the broad impacts of unrestricted immigration on the political economy, as their subtitle and much of their book suggest? Or is it a narrower and more straightforward examination of the economic implications of a “free immigration” regime?

I begin with Powell and Nowrasteh’s understanding of “institutions.” Unfortunately, as just noted, they fail to elaborate or explain what they mean by this term. I am left to assume that as with most economists, an institution here is any on-going set of relationships or rules (whether implicit or explicit) by which individuals opt to come together to achieve specific ends.

What our political economist authors fail to acknowledge, however, is that for many other social scientists, perhaps especially political scientists, institutions are distinct from mere organizations.

Indeed, both the eminent sociologist Philip Selznick and the distinguished political scientist Hugh Heclo define an organization as merely a means or tool of accomplishing a given task or set of tasks; whereas an institution entails a seriousness of intention and purpose toward goals and objectives whose attainment points toward higher aspirations, even moral imperatives. Of course, such a perspective implies a non-positivistic social science; but isn’t that what the authors presumably aspire to when they invoke the label of political economy? Yet that term implies a much broader lens through which to assess the impacts of immigration than the single criterion of economic freedom, which is the only one the authors systematically explore.

Powell and Nowrasteh measure economic freedom across five domains: size of government; protection of private property rights; sound money; freedom to trade internationally; and regulation of businesses, labor markets, and credit markets. By examining this composite variable and its relationship to migrant influxes, they report substantial increases in economic freedom associated with unrestricted migration levels.

Powell and Nowrasteh also spend considerable time and effort exploring the non-economic impacts of mass migration. But unfortunately, they fail to reconcile the one set of findings with the other. For example, they allude to Thomas Jefferson’s concerns about the negative impacts of “the principles of the governments” that immigrants leave behind or of the “unbounded licentiousness” these newcomers might subsequently indulge in. But surely, Jefferson regarded these traits primarily as threats to republican virtue, not economic growth. How are such concerns about immigrants today evaluated and reconciled with the authors’ findings about increased economic freedom? The authors never say.

Similarly, Powell and Nowrasteh offer as instructive examples of free immigration the effects of exogenous migration shocks on Israel and Jordan. Indeed, they cite very positive impacts on economic freedom in these two cases—without adequately addressing the fundamental difference between these two cases and anything one might possibly infer about opening up U.S. borders to unfettered immigration (or anything even approximating such a policy). I raised this point in my initial remarks, but I am still waiting for a response.

In the same vein, Powell and Nowrasteh raise the intriguing issue of the impact of large numbers of socialist immigrants during the period of unrestricted immigration to the United States that ended during the 1920s. Their evidence, based on records of the Socialist Labor Party (which was founded in Germany in 1876 and then continued here), indicates that in one cohort of such immigrant workers, three-quarters failed to remain in the party once in the United States. Intriguing, yes. But does this indicate, as the authors suggest, that the overall impact of immigrant socialists on America has been—counter-intuitively—to advance economic freedom and slow the growth of government?

Such a claim would not impress either critics or admirers of Walter Reuther, the son of German immigrant socialists, who went on to lead the United Auto Workers, the Congress of Industrial Organizations, and the AFL-CIO, as well as play a significant role in the civil rights movement. Nor would it impress historically literate civil libertarians, who could justifiably point to how America’s openness to violent anarchist and syndicalist immigrants contributed to the post-World War I reign of terror that led to the creation of the FBI and launched the dubious career of J. Edgar Hoover. Overall, the broad historical claims made by these authors are only suggestively, not convincingly supported by their statistical analyses.

This is undoubtedly why Powell and Nowrasteh also present persuasive evidence that in the past unrestricted immigration resulted in an ethnically, linguistically, and religiously diverse mix of newcomers whose heterogeneity retarded their coming together for collective action. Here again, the authors appear to assume—but do not specifically support—the notion that any such collective effort necessarily threatens economic freedom.

This leads to my final point. If the United States were to take Powell and Nowrasteh’s advice and move toward unrestricted immigration, why should Americans assume the resulting mix of new arrivals would lead to an overall beneficial outcome? Relying on elaborate statistical analyses and models, our authors argue that terrorist risks would be minimal. Even if I were persuaded by such analyses, I doubt the American public would be. But more broadly, why should we assume that such an open immigration regime would have benign results only? Even if such a policy resulted in significantly enhanced economic growth as well as increased economic freedom, it might also have negative social and cultural impacts that under a “free immigration” regime could be neither predicted nor controlled. These are the kinds of questions inevitably raised by the authors’ hypotheticals, but that they fail to address—certainly in any convincing manner.