Immigration is a large and complex subject. My recent story in The Atlantictried to do justice to that complexity. But that attempt of course came with a cost: The article was long and often highly detailed. It became easily possible to lose track of the piece’s argument.
So before replying to some of the article’s notable critics, let me recapitulate the argument as briefly as I can:
1) The economic benefits to Americans of America’s present immigration policies are very small. Almost all of those benefits are captured by the most affluent Americans.
2) The fiscal costs of present immigration policies are high. The argument that immigrants strengthen Social Security and Medicare is false. America’s bias in favor of low-skilled immigrants, legal and illegal, means that immigrants pay relatively little in taxes, while requiring a lot by way of services.
3) The most important costs and benefits of immigration are neither economic nor fiscal, but social and political.
4) Among those social and political costs: the radicalization of politics that we see in so many developed countries. Brexit, the rise of the French National Front, the triumph of authoritarian populists in Italy and other European Union countries, and America’s own election of Donald Trump—all were fueled by many causes, but high levels of immigration provided the spark.
5) The pressures of immigration are not going away on their own. Over the coming decades, more and more people in the global South will be able to afford to make their way—legally, illegally, or as asylum seekers—to the global North. Unless properly managed, this impending movement of people may destabilize democratic governments on an even greater scale than anything we have yet seen.
6) Many of the benefits of immigration can be gained, and many of the harms minimized, by more strategic immigration policies, including stricter enforcement in the workplace to protect legal workers; faster removal of unfounded asylum claimants; more emphasis on skilled rather than unskilled immigrants; and lower total numbers of immigrants.
Different critics have focused on different aspects of the article.
In Slate, Jordan Weissman disputed the first of these points, citing a Wharton economic model that restricting immigration would reduce GDP per capita by 0.3 percent by 2040. How much of a loss are we talking about in real money? You can do the math yourself on the Wharton site. The Wharton model projects GDP per capita in 2040 as $68,900 under current policy. A 0.3 percent loss would cut GDP per capita in 2040 to … $68,693.30. As I said: small.
In TheWashington Post, Radley Balko disputed the second point. “Undocumented immigrants contribute more to the economy than they take out, and are less reliant on social welfare than native-born Americans,” he wrote. This objection is accurate so far as it goes, but misleading. People illegally present in the United States are of course ineligible for most forms of social welfare. Their families, however, may be eligible. As I report in my article, altogether about half of immigrant-headed households use means-tested social services in any given year. The average immigrant to the United States costs his or her state and local government about $1,600 more in services each year than he or she pays in taxes.
On Twitter, Noah Smith of Bloomberg Opinion disputed the fourth point. He notes that the United States suffered from a great deal of social and political turmoil from 1967 to 1972, when the immigrant share of the population was historically low. It is, he insists, therefore unhistorical to believe that immigration has much to do with the rise of authoritarian politics since 2010, when immigrant shares of the population have been high.
The U.S. (and other democracies) were wracked by political violence in the years when the Baby Boomers reached their late teens. That indeed happened. But it’s not the only thing that can happen. Smith seems to be arguing that if one demographic change at one time causes trouble, then you need not worry about the possible ill effects of other demographic changes at other times. If you’ve caught the flu, you need never worry about breaking your leg.
Yet Trump voters, Brexit voters, National Front voters, Alternative for Germany voters, Italian Lega voters all tell us that immigration is their concern. It seems bold to disbelieve them.
More generally: The revolutionary turmoil of the 1960s and the reactionary political mobilization of our times are different events occurring at different times, calling for different explanations. Why is that hard to understand?
The core argument of Jordan Weissman’s reply in Slate—and this is a point stressed by other critics as well—is that Americans have expressed increasingly positive attitudes to immigration since Donald Trump became president. If I’m worried that immigration is pushing Americans toward authoritarian politics, maybe the concern has dispelled itself?
There’s obviously some truth here. Americans rightly recoil from Trump’s rhetoric, and that has propelled many to a new sympathy for immigrants and immigration. Which is a good thing. But you have to ask: What are we measuring? Feelings about immigration? Or about Trump?
Within days of his nomination by Trump, Neil Gorsuch had become the second-least-popular Supreme Court nominee in almost 30 years. Only Harriet Miers scored worse. In her first weeks as speaker of the House this term, Nancy Pelosi polled as the most popular speaker in 25 years.
Trump is unpopular. Things endorsed by him become unpopular by association. Things attacked by him become popular on the same principle.
I read the recent immigration numbers as a similar kind of Trump effect. This effect will last only as long as his presidency. The inherent strains of large-scale immigration are a more enduring global fact. One is weather; the other is climate.
In The Federalist, John Daniel Davidson accused me of trying to centrally plan the American labor market.
Market forces are at play. U.S. employers have a demand for cheap labor and plenty of foreign-born laborers are willing to meet it. As conservative as it might sound to say that we need more high-skilled immigrants and fewer low-skilled immigrants, it’s not really conservative at all. Federal bureaucrats are going to decide how many foreign-born workers U.S. industries need, and what their respective skill-sets should be? Usually, conservatives take the view that trying to centrally plan the U.S. economy is a utopian fantasy, doomed to fail. Why would a centrally-planned immigration system be any different?
But under present law, it is not employers who select the country’s legal immigrants. In seven cases out of 10, it is those immigrants’ previously settled relatives. By shifting away from family reunification as the basis of legal immigration, and prioritizing skills, we would actually increase the role of employers in the process.
It is true that illegal immigration is almost exclusively driven by employers’ desires to pay less for labor. But if that wish will drive U.S. policy, we will have to allow employers to violate many laws: not only immigration laws, but laws against wage theft, against child labor, against unsafe working conditions.
On Twitter, Yoram Hazony complained at length about my use of the word fascist in this sentence: “If liberals insist that only fascists will enforce borders, then voters will hire fascists to do the job liberals refuse to do.”
Hazony has written a much-acclaimed book defending nationalist politics. He wonders why nationalists should not simply be respectfully described as “conservative.”Others have objected to the imputation that liberals won’t defend borders, pointing to the costly border-security measures approved by Democrats in the current Congress and earlier.
To this double-barreled criticism, I offer two answers. First, It’s not helpful or accurate to apply the same label to Marine Le Pen, Nigel Farage, and Donald Trump as was once applied to Ronald Reagan, Margaret Thatcher, and Helmut Kohl. The nationalist political formations of the 2010s make different appeals to voters than the conservative parties of the 1980s did. They win support from different voters.
The rise of these nationalist parties is forcing a rearrangement of the political grammar of the developed world. The conservative parties of the 1980s defended markets and were skeptical of economic redistribution. The nationalist movements of the 2010s are skeptical of markets and defend economic redistribution, provided that the redistribution benefits people of the correct ethnic stock and cultural outlook.
The old conservatism and the new nationalism may share some ideological DNA, but to describe them as the same thing is salesmanship, not political analysis.
Second, the reaction to my article should confirm, I believe, that for many highly politically engaged people on the left, the enforcement of borders is indeed an inherently racist and fascist project. One Democratic candidate for president, Beto O’Rourke, has called for dismantling all physical barriers on the U.S.-Mexico border—explicitly because barriers make it too difficult for would-be border-crossers to enter the United States. The Democratic leadership in Congress has been more responsible, but if you think my language describes an imaginary state of affairs, just scroll through my Twitter timeline for proof of how real it is.
Many critics have singled out one passage in the article for special scorn. I wrote:
Heavy immigration has enabled the powerful—and the policy makers who disproportionately heed the powerful—to pay less attention to the disarray in so many segments of the U.S. population. Because the country imports so many workers, employers do not miss the labor of the millions of men consigned to long-term incarceration. Without the immigrant workers less prone to abuse drugs than the native-born, American elites might have noticed the opioid epidemic before it killed more Americans than died in the Vietnam, Korean, and Iraq Wars and the 9/11 attacks combined. The demand for universal health coverage might gain political force if so many of the uninsured were not noncitizens and nonvoters. None of this is immigrants’ fault, obviously. It is more true that America’s tendency to plutocracy explains immigration policies than that immigration policies explain the tendency to plutocracy. Managing immigration better is only one element of restoring equity to American life. But it is an essential element, without which it is hard to imagine how any other element can be achieved.
Critics have interpreted this passage as blaming immigrants for the opioid epidemic and other U.S. social problems. If on reading that passage in full, you think I’ve done this—well, then I’ve certainly failed to write clearly enough. One of the major goals of the article was precisely to avoid the language of blame in favor of a language of consequences.