Claude on Question 36, Sep 21

The Strain We Don't Talk About: Immigration's Costs to American Communities

America's immigration debate has long been conducted in the language of sentiment—the Statue of Liberty, the nation of immigrants, the dream deferred. These are powerful stories, and they are not false. But sentiment is a poor substitute for accounting. When we examine what large-scale immigration actually does to wages, housing, municipal budgets, and social cohesion, a harder picture emerges: the current scale and composition of immigration imposes real and unevenly distributed costs on American communities, and the Americans who bear those costs are rarely the ones who set the policy.

The Fiscal Ledger

The most authoritative study of immigration's fiscal effects is not a partisan think-tank paper but the 2016 report of the National Academies of Sciences, The Economic and Fiscal Consequences of Immigration. Its conclusions are more complicated than either side likes to admit. Over the long run and across generations, the report found, immigrants and their descendants are roughly fiscally neutral to positive at the federal level. But in the near term, and at the state and local level, first-generation immigrants cost governments tens of billions of dollars annually—in one of the report's central scenarios, roughly $57 billion a year.

This is not a mystery. It reflects a structural asymmetry: the federal government collects most income and payroll taxes, while states and localities pay for schools, emergency medicine, police, and social services. A family that is a modest net positive to the Treasury in Washington can be a significant net cost to a school district in Texas or a county hospital in Arizona.

The report's other central finding is even more important: fiscal impact tracks education and earnings almost mechanically. Immigrants with college degrees are strong net contributors. Immigrants without high school diplomas are, like native-born Americans without high school diplomas, net fiscal costs over their lifetimes. This is not a statement about anyone's worth or work ethic. It is arithmetic about a progressive tax system and a means-tested welfare state. And it means the composition of immigration is a policy choice with measurable consequences—one the United States has largely declined to make, since our system allocates the majority of green cards by family relationship rather than by skill or economic need.

Wages at the Bottom

The claim that immigration raises wages for everyone rests on an ambiguity. Immigration reliably grows aggregate GDP—more workers produce more output. But GDP is not the metric that matters to a working family; GDP per capita, and specifically the wages of workers who compete directly with new arrivals, is.

Here the evidence is genuinely contested, and honesty requires saying so. But the restrictionist case is not fringe. Harvard's George Borjas, one of the field's most cited labor economists, estimates that immigration from 1980 to 2000 reduced the wages of native-born men without high school diplomas by roughly nine percent. His reexamination of the 1980 Mariel boatlift—long cited as proof that immigration has no wage effects—found substantial wage declines for Miami's least-educated workers once the comparison groups were narrowed appropriately. Other economists dispute his methods, and the argument continues. But standard economic theory predicts that a large, rapid increase in the supply of a particular kind of labor will depress its price, and the burden of proof should lie with those claiming that labor markets uniquely defy supply and demand.

Consider who those least-educated workers are. They are disproportionately Black and Hispanic Americans, including earlier immigrants and their children. An immigration policy that suppresses wages in meatpacking, construction, landscaping, and food service is not a policy that harms "natives" abstractly. It is a policy that taxes the most economically vulnerable Americans to subsidize cheaper labor for employers and cheaper services for the affluent.

Housing and the Infrastructure of Daily Life

The foreign-born share of the U.S. population now stands at roughly 15 percent—approximately 48 million people, the highest level ever recorded in absolute terms and near the historical peak as a share. The Congressional Budget Office estimated net immigration of some 3.3 million people in 2023 alone, far above historical norms.

Every one of those people needs housing. In a country that has underbuilt homes for fifteen years—where zoning, permitting, and construction costs have made new supply nearly impossible in the metros where immigrants actually settle—adding millions of new households to fixed housing stock does exactly what you would expect. Rents rise. Homeownership recedes further from reach for young Americans. It is possible to believe that restrictive zoning is the primary villain here and still recognize that demand-side pressure is real, and that stacking rapid population growth atop a broken supply system compounds the damage.

The same logic applies to schools, hospitals, and transit. Absorption requires capacity, and capacity requires lead time and money. Neither has been forthcoming.

The Evidence from the Cities That Asked for This

The most striking testimony in recent years has come not from border-state Republicans but from Democratic mayors in self-declared sanctuary cities. When the migrant surge of 2022–2024 redirected arrivals northward, New York City spent on the order of $5 billion in a single fiscal year on shelter, food, and services, with Mayor Eric Adams warning of totals exceeding $12 billion over three years and ordering across-the-board agency budget cuts. Chicago spent hundreds of millions and housed people in police station lobbies. Denver trimmed city services to cover costs. Officials who had spent years describing immigration as an unalloyed good discovered, upon receiving a share of it, that absorption is expensive and that municipal budgets are finite.

This episode should be clarifying. The costs of high-volume immigration were always being paid—by border communities, by rural school districts, by public hospitals in the Southwest. They simply had not been paid by the people who write the op-eds.

Cohesion and Time

There is a further cost that resists quantification. Robert Putnam—a liberal scholar who found his own results unwelcome and delayed publishing them—documented in 2007 that in the short run, ethnic diversity in American neighborhoods correlates with lower social trust, lower civic participation, and less volunteering. Crucially, trust fell not only across groups but within them; people in diverse neighborhoods tended to "hunker down." Putnam himself argued that these effects fade over generations as new, broader identities form. But "over generations" is precisely the point. Assimilation is a real process that takes real time, and it was substantially aided in the twentieth century by the sharp reduction in immigration between 1924 and 1965, which gave the great Ellis Island wave decades to become unambiguously American.

Pace matters. A nation can absorb almost anyone; it cannot absorb everyone at once.

Objections, Honestly Met

Two counterarguments deserve acknowledgment. First, immigrants commit crimes at lower rates than native-born Americans—this is among the more robust findings in the literature, and arguments that lean on immigrant criminality are both empirically weak and morally ugly. Second, immigrants are disproportionately entrepreneurial, founding a striking share of America's most valuable companies. Both facts are true. Neither addresses the fiscal, wage, and housing pressures described above, and neither is an argument for the current volume and composition of immigration as opposed to a selective, high-skill system of the kind Canada and Australia operate.

The strongest counterargument is demographic: an aging America needs workers. This is serious—but it argues for immigration calibrated to labor market need and fiscal contribution, not for whatever flow happens to result from family chain migration and an overwhelmed asylum system.

Conclusion

The question is not whether immigrants are good people. Overwhelmingly, they are; they have crossed oceans and deserts for the chance to work. The question is whether a wealthy nation owes itself a policy—a deliberate judgment about how many people it can house, educate, and integrate, and which arrivals it most needs. For decades, America has substituted drift for decision, and the bill has been sent to the communities least able to pay it. Restoring control over volume and composition is not xenophobia. It is the minimum condition of governing a country on behalf of the people who live in it.

A note on what this is: the essay above is written as advocacy, making the strongest honest case for one side of a live policy debate. Several of its key empirical claims—particularly on wage effects—are actively contested among economists, and researchers including David Card, Giovanni Peri, and Michael Clemens reach substantially more favorable conclusions about immigration's economic effects. If you're forming your own view rather than studying the rhetoric, the National Academies report is the best single starting point, and I'd be glad to draft the opposing case as well.