What an Immigrant Costs

The two charts are not equally solid, and the difference matters. The first is a census. The Department of Homeland Security counts green cards, and the count is the count. The second is a model. The United States does not track what any individual pays or receives, so the National Academies built an estimate by assigning shares of the federal budget to demographic groups. It is the most careful version of that exercise anyone has done, and it is still an assignment rather than a measurement.

Which is why the scenario selector is the important control on this page. The two options differ on one question: when another person joins the country, does the cost of national defense, interest on the federal debt, and other pure public goods go up? If you say yes and divide those costs per head, a first-generation immigrant appears to receive $9,767 in federal benefits against $7,117 in federal taxes. If you say no, because an aircraft carrier costs the same whether it protects 330 million people or 331 million, the same immigrant appears to receive $6,154 against the same $7,117 in taxes, and becomes a net federal contributor. That is a swing of $3,613 per person per year, produced entirely by an accounting convention. The person, the taxes, and the services are identical in both columns.

Two findings survive either choice. First, native-born citizens consume more federal resources per head than first-generation immigrants, in every scenario the study modelled. The reason is demographic rather than moral: Social Security and Medicare dominate federal outlays, immigrants arrive disproportionately as working-age adults, and a large share of them will have paid in for years before drawing anything out. Second, the second generation, the children of immigrants, pay more federal tax per head than any other group in the study, including the third-plus generation. Whatever the first generation does or does not cost, its children are the strongest net contributors in the table.

The scarcinality reading. A recurring claim of the framework is that a number presented as a physical fact often turns out to be a claim about ordering. This is that pattern in an unusually pure form. The public argument about whether immigrants cost or contribute is conducted as if it were an empirical dispute waiting on better data. It is not. Both sides can cite this same study, honestly, and arrive at opposite conclusions, because the sign of the answer is set before any data is consulted, by a decision about how to allocate the cost of things no individual consumes separately. The scarcity being argued over is not a scarcity of federal resources. It is a scarcity of agreement about who is charged for the commons.

And the admissions chart underneath it makes the same point from the other side. The spike to 1.8 million in 1991 was not a surge of arrivals; it was the 1986 legalisation moving roughly 2.7 million people who were already living and working in the country from one legal category to another. Nothing physical changed in those years. Their consumption of federal resources did not change. Only their classification did, and with it, which side of this ledger they appear on.

Sources: Department of Homeland Security, Office of Homeland Security Statistics, Yearbook of Immigration Statistics Table 1, persons obtaining lawful permanent resident status by fiscal year, extracted from the published series; fiscal years 2024 and 2025 from the annual flow reports. Per-capita fiscal figures from the National Academies of Sciences, Engineering, and Medicine, The Economic and Fiscal Consequences of Immigration, 2016, Table 8-2, reporting 2013 dollars for first, second, and third-plus generations each including dependents, populations of 55.5, 23.3, and 237.3 million respectively. State and local figures are excluded here because the question asked was federal; at the state and local level the first generation runs a larger deficit, driven mainly by the cost of schooling children.