It looks erratic to us but may one day seem inevitable to historians that Donald Trump should have fought his anti-establishment campaign for the Republican nomination on the terrain of immigration policy. Already it is hard to recall that it was the establishment, not Trump, that insisted the battle be fought there. The candidate, at his announcement speech, spent a few minutes on immigration but then moved on to China, ISIS, Obamacare, the national debt, the Second Amendment, his desire to be a kind of National Cheerleader, and his own net worth. Trump’s skepticism about mass immigration won the attention of his primary opponents and the journalists who covered him because it seemed crazy—almost pitiable.

For a generation, mass immigration has held a place of honor in each party’s political theology. It fits Democrats’ anti-racism and Republicans’ supply-side economics. There is a bipartisan magic about open borders. When the needs of immigration conflict with those of democracy, it is democracy that gets pushed aside. Federal and state authorities have left unenforced, and even flouted, the laws that govern employment, deportation, access to public services, and voting rights for non-citizens. In a 1994 referendum, five million Californians sought to deny welfare benefits to illegal immigrants, giving the state’s Proposition 187 a 17-point landslide at the polls. But District Court Judge Mariana Pfaelzer decided they were wrong. And that was that.

Immigration is thus a synecdoche for the way society has evolved over the last half-century. The economy boomed as we tapped resources that our forefathers’ technological and—so we liked to think—moral backwardness prevented them from tapping. Since 2008 it has become clear that what had looked like a boom was actually a bubble made of $45 trillion in government, household, and business debt. The benefits of immigration are obvious to anyone who has ever eaten sushi, left a rumpled hotel room and found it spotless on returning a few minutes later, or golfed on three or four different well-groomed courses in the same small city. The costs of immigration, by contrast, are discussed only within a Losers’ Corner of poisonous internet comment threads and drive-time radio shows. What is more, the payoffs—that sushi, those golf courses—came immediately. The liabilities were mostly “off-balance-sheet,” and have yet to be settled. The welfare state’s responsibility for the swelling ranks of the aging poor is barely nodded at in the budget. The adaptation of the U.S. Constitution to fit immigration, rather than vice versa, is a huge cost, too. The adaptations required by mass immigration are so large that it can be judged a success only if the compensating economic benefits are vast. The best recent economic research indicates that they are not.

* * *

Among academic economists, George Borjas, a professor at Harvard’s John F. Kennedy School of Government, has a reputation as a debunker of pro-immigration myths and narratives. This is not out of any a priori hostility to immigration. Having left Cuba as a child in 1962 after the Castro government confiscated his family’s clothing factory, he is himself a beneficiary of American openness.

But four decades in academic life have convinced Borjas that most of those who claim to study immigration—in academia, journalism, and politics—are better thought of as advocates for it. University of Maryland economist Julian Simon once warned him that “anti-immigration people” were quoting Borjas’s conclusions, and urged him to intervene to put a stop to it. When Borjas was studying whether there was a tendency of immigrant labor pools to deteriorate in quality over time, a Rockefeller Foundation grant officer advised him not to “open that can of worms.” Often it is advocacy that brings economists to the subject of immigration in the first place, not vice-versa. The studies most widely circulated in academia, journalism and government are marked, Borjas believes, by “arbitrary conceptual assumptions, questionable data manipulations, and a tendency to overlook inconvenient facts.”

* * *

His new book, We Wanted Workers, shows that much of what we think we know about immigration is, therefore, dubious or false. There is no shortage of life-science doctorates in the United States—despite years of dire warnings from businesses and universities that would benefit from more foreign biologists. Immigrants are, on average, seven years older (44) than the native-born (37), despite the rhetoric that takes migration as a synonym for rejuvenation. And despite the assurance of the Department of Homeland Security (DHS) that there were 11.4 million illegal immigrants in the country in 2012, that number is probably bogus, Borjas shows. The DHS arrives at it by using census surveys of foreign-born, subtracting the number of legal immigrants who have spoken to interviewers, and assuming the count has missed 10% of the illegals. That assumption, in turn, is based on the suggestion of one junior researcher regarding the counting of Mexicans in Los Angeles County in 2000. Programs based on that 10% estimate have wound up spectacularly oversubscribed. In 2015, when the California Department of Motor Vehicles offered driver’s licenses to illegal immigrants, they showed up at twice the numbers expected.

Borjas is not a historian of immigration. He is uncomfortable discussing the subject in its fascinating non-quantitative aspects, such as, say, whether immigration can drive a culture off the rails. But careful social science always offers insight on such matters. Borjas was first drawn to immigration economics, he says, by a curiosity about how immigration pools change over time and, specifically, whether they decline in quality. This is a question that many Americans were asking in the 1960s about Cuban immigrants like Borjas, and that Trump has raised about Mexicans today. (“They’re not sending their best,” he alleged last year.)

* * *

There are actually plenty of reasons, in theory, to expect such a decline. A large migration might lower the “entry wage,” thereby decreasing the incentives for higher-skilled workers to come. One of the mistakes immigration advocates often make is to assume that incentives are constant over time. So, for instance, the motley and polyglot wave of mostly Mediterranean and Eastern European immigrants who arrived at the turn of the 20th century was a success at practically every stage. It accounted for three quarters of Henry Ford’s workforce by the eve of World War I. It would later form the backbone of a middle class so well assimilated that, by the middle of the 20th century, it would be mocked by folk singers and student radicals as bland, “lily-white,” and conformist. But it would be foolish to expect similar results in a world in which the University of California recently warned faculty in a memo that to call America a melting pot is to commit a “micro-aggression” against minorities.

Because incentives change, today’s Mexican newcomers are learning English less quickly than their predecessors did a couple of decades ago. Borjas blames the growth of ethnic enclaves. Economic assimilation has probably slowed down, too, but this slowdown can be concealed if one is selective about one’s data. So the Wall Street Journal reassures its readers that the “children of immigration” are doing better economically than the generation of their parents. This is not necessarily true, Borjas shows. The entry wage for immigrants, which was 11% lower than that of natives a generation ago, has fallen more recently to 28% below. The Journal matches recent immigrants against the children of immigrants, who came decades ago. It thus gives a huge and wholly unwarranted boost to the case that everything is hunky-dory when it comes to newcomers fitting in. If you compare today’s kitchen staff at Mi Taco Sabroso to today’s children of oil executives who fled the Ayatollah Khomeini’s revolution in 1979, you will leave the false impression that first-generation rags lead reliably to second-generation riches.

Borjas’s skepticism about the standard immigration narrative is the more damning because, on almost all social science matters, he seems not to have a contrarian bone in his body. Economists, like all knowledge specialists, are often prisoners of the research agenda taken up by their most gifted contemporaries. Borjas has no bone to pick with that agenda, which is often obsessively focused on uncovering bigotry and prejudice. Thus he quotes a study by economist Stephen Trejo to the effect that relative youth, bad English, and miseducation explain three quarters of the wage gap between Mexicans and U.S. whites but only a third of the black-white gap—“leading,” Borjas writes, “to the conclusion that much of that gap reflects the pernicious effects of racial discrimination.” Leading how? Not via any evidence Borjas cites. “In the long run,” he writes elsewhere, “immigration may be fiscally beneficial because the unfunded liabilities in Social Security and Medicare are unsustainable and will require either a substantial increase in taxes or a substantial cut in benefits.” But those requirements do not make it wise or even advisable to add more unfunded liabilities in the form of immigrant retirement costs.

* * *

Borjas’s criticism of the standard immigration narrative thus carries no political agenda at all. It is confined to the ways that that narrative fails to hold up on its own terms. If he has arrived at conclusions more pessimistic than those of his colleagues, he has done so not by challenging their ideology but by correcting their errors. Several are laid out in We Wanted Workers. Let us examine three:

1. Immigrants are more welfare-dependent than the most frequently quoted statistics indicate, and far more welfare-dependent than the population at large.

If one looks at data from the Census Bureau’s Survey of Income and Program Participation, one finds that 46% of households headed by an immigrant resort to welfare in some form, versus 27% of households headed by an American. Supporters of mass immigration, however, from community organizers to the Wall Street Journal, prefer to use a different and more easily manipulable data set from the Census Bureau’s Current Population Survey, and to arrange it by individuals rather than households. This sounds more…individualistic. It also gives the impression that rates of welfare dependency among newcomers and natives are more roughly comparable. But it is a trick, Borjas shows. If an undocumented single mother from Mexico, say, bears two children after arriving in the U.S. and winds up on welfare, the system shows an increase of one immigrant and two natives. The welfare system is propped up by native households, each of which pays, by Borjas’s estimate, about $470 per year to cover losses from immigration.

2. Competition from immigrants dramatically reduces the wages of the workers whose qualifications most resemble theirs.

This is the sort of common-sense conclusion that you need not ever spend a day in economics class to understand. Yet for three decades economists have clung doggedly to the doctrine that immigrants can offer efficiencies to an economy without lowering the wage in the industries where they work. This is nonsense on the conceptual level: the lowered wages are the efficiencies.

Borjas focuses on a celebrated 1990 study of the Mariel boatlift, by Princeton labor economist David Card. In the course of a few weeks in 1980, 125,000 people—a variety of dissidents, criminals, and ambitious youngsters—were allowed to flee Cuba, and wound up in Florida. Despite what was considered, in the days before China joined the world economy, a massive labor shock, Card found no evidence that the newcomers depressed general wages around Miami. Politicians have thus quoted his study ever since. Barack Obama dredged it up in 2014. It is useful. But it is, Borjas shows, wrong. Card had looked at the workers in metropolitan Miami as a whole. The Miamians with whom the Marielitos competed most directly—high-school dropouts—were effectively “tucked away” inside this larger group. But once poorer workers were isolated, it was easy to show that their weekly wage fell between 1979 and 1985. It fell, in fact, by an astonishing $100 a week. Incantations about diversity do not abolish the laws of supply and demand. After a 2006 raid on a chicken plant in Stillmore, Georgia, which rousted out illegal workers, the plant had to hire locals, and did so at significantly higher wages.

Whether immigrants help or hurt a sector of the economy has to do with whether they enter it as “complements” or competitors. Today’s immigrants are complements for rich people, who tend not to act as their own valets, chefs, gardeners, or maids. Others do those jobs. If the cost of them gets cheaper, rich people’s lives get better, and the number of people who can live like rich people may increase. The lives of the natives who used to perform those tasks get worse. The rule of thumb is that a 10% increase in the workforce of a given sector will result in a 3% fall in wages.

3. The primary effect of immigrants on the country receiving them is a massive regressive redistribution of income and wealth among natives.

This redistributive effect is, for Borjas, “the key insight I have gleaned from decades of research on the economics of immigration.” The main thing about immigration is not wealth creation. It is not entrepreneurship. It is not diversity. It is redistribution from the poor to the rich. That this should be so jarring and implausible-sounding to contemporary sensibilities shows how censored the discussion of immigration economics has been—for this has always been one of the basic consensus conclusions of most economic models of migration.

* * *

Borjas wants us to know that, whenever economists predict vast economic benefits from open immigration, they are using equations devised by individual economists, often with an agenda in mind. In the very simplest model of an open economy, with the world divided into a relatively poor southern hemisphere and a relatively rich northern one, and with no frictions or moving costs, we would find that most of the world does indeed get richer from free movement. But let us consider why this is so. Under such a model the vast majority of the Global South’s workers—billions of them—will move north. That so many people would want to move is not wholly implausible. Almost a third of those born on the island of Puerto Rico—who are American citizens—have moved to the mainland United States. In 2015 the U.S. “diversity lottery” for immigrant applicants drew 15 million applications for 50,000 spots.

According to Borjas, the world gets richer because

the earnings of the North’s native workforce will drop by almost 40 percent, while Southern workers will more than double their earnings. One last redistributive impact is worth mentioning—and again it is one of those nuisance statistics that is swept under the rug: the income of capitalists worldwide will increase by almost 60 percent.

This “basic” model, out of which almost all of our projections of the benefits from immigration come, contains wildly optimistic assumptions. It assumes that the massive inrush of foreign peoples will do nothing to alter the infrastructure, constitutions, or associations that gave the advanced economies their competitive edge. If such things do change—and crowded hospitals, affirmative action, and bilingual education programs are all evidence that they do—then the “gains” from immigration rapidly evaporate. They can even turn into losses.

* * * 

In 1995, Borjas tried to estimate the actual effects of immigration on the United States, and published his conclusions in the Journal of Economic Perspectives. He found that GDP rose by $2.1 trillion, but virtually all of those gains—98%—went to immigrants themselves. Let us not lose sight that these gains constitute a great addition to the world’s happiness. If we were judging open immigration not as an economic policy but as an international aid program, we might consider it a success.

But we aren’t. When economists talk about “gains” from immigration to the receiving country, they are talking about that remaining 2%—about $50 billion. This “surplus” disguises an extraordinary transfer of income and wealth. Native capitalists gain $566 billion. Native workers lose $516 billion.

The growth effects of immigration are small—“at best a wash” is the way Borjas puts it towards the end of the book. But the redistributive effects—from poor to rich, from workers to financiers—are staggeringly large. The role of immigration in producing inequality appears comparable to that of more commonly cited villains: technology, world trade, tax cuts.

As always Borjas ducks the political implications of such questions. For the purposes of exposition, that is helpful. Whether rising inequality is worth worrying about is a matter of political taste; whether it is worth fighting is a matter of ideology. Some people will see increasing inequality as betokening a dangerous plutocratization. Other views are possible. Mass immigration picked up momentum in the mid-1970s, which was both the most socioeconomically equal decade in the history of the United States and a time, like our own, when the country appeared to have lost much of its competitiveness. A free marketer might say there was a wage bubble occasioned by a tight market in proles. The key economic problem of Western democracies—Britain above all, but America as well—appeared to be the power of trade unions. Immigration has always been the most tried and true way of breaking them.

* * *

People can differ about whether immigration is at all times a good policy, whether it served a necessary function in the 1970s that it no longer does, or whether it was a mistake from the get-go. But there is one disturbing non-economic element of the question that we cannot pass over. It used to be said, when oil prices were rising, that, in order to maintain a lifestyle it couldn’t afford, the United States was “importing inflation.” In a similar way, with immigration we may be importing oligarchy.

Surely when we talk about the American competitive advantage, the country’s democratic institutions are the heart of the matter. If, as George Borjas has shown, immigration suppresses the wages of workers, and transfers much of their wealth to elites, then liberalized immigration is a policy that cannot be carried out without simultaneous injuries to democracy. For why would native workers favor a system that makes them poorer? Perhaps they have somehow been hoodwinked out of an accurate assessment of the effects of the system. Perhaps they have lost their purchase on democracy itself. Either way, in exchange for a nickel here and a nickel there, we appear to have created a political problem of considerable gravity.