Last October, President Trump signed a proclamation that bars legal immigrants from entering the United States unless they can prove they have health insurance coverage or the means to pay for it within 30 days of arriving in the U.S. Since it is unlikely that prospective immigrants can secure health insurance before arriving in the U.S., implementing this rule will require consular officers to estimate “reasonably foreseeable” medical costs and ensure immigrants have the resources to pay them.

The administration’s rationale for the rule is a belief that immigrants are likely to be a burden on the U.S. health care system, although there is little evidence to support this argument. In general, immigrants use fewer health resources and services than native-born Americans. Recent immigrants — about 5 percent of the population — are responsible for less than 1 percent of public medical spending; recent immigrants with public insurance had health expenditures one-sixth that of publicly insured native-born Americans. Using a straightforward interpretation of “reasonably foreseeable medical costs,” we simulated the potential effects of a rule requiring people to have resources equal to the expected costs of their future medical care. We found that the vast majority of those affected would not incur the predicted level of costs. At the same time, the rule would reduce less than half the burden on the medical system incurred by immigrants with high medical costs.

It would be very challenging to substantially reduce uninsured medical costs without disadvantaging many potential migrants who are not likely to be a burden on the system. Across the population, a very small minority of people account for a majority of expenditures. It is difficult to predict who will fall into the costly tail of that distribution.

In our simulation, we assumed that the “reasonably foreseeable” costs of a person with a given health condition could be approximated as the actual average costs of people diagnosed with that condition. We used data from the Medical Expenditure Panel Survey from 2009 to 2015 and examined spending among people with common clinical conditions and those with total annual health spending of more than $20,000. The conditions included acute diagnoses like hip fractures, as well as chronic conditions such as HIV infection, diabetes, and cancer. We found that within each health condition, spending was concentrated among a small group of people. Those at the 75th percentile of the spending distribution spent about 10 times as much as those at the 25th percentile of the distribution.

Most people required to have resources equal to the “reasonably foreseeable costs” of an illness would not actually spend this amount. Across conditions diagnosed, between 54 percent and 83 percent of people would not spend the average “reasonably foreseeable” cost during their second year after diagnosis. That is, the “reasonably foreseeable cost” standard would have a high false positive rate, requiring many people to have more resources on hand than would be needed.

While the great majority of people affected would be forced to have excessive resources under this standard, it would still fail to eliminate the majority of potentially uncompensated care costs. Across all conditions, about 80 percent of all spending is incurred among the minority of people (18%–46%) whose spending exceeds the “reasonably foreseeable cost” threshold. After subtracting the payments this group would make with the “reasonably foreseeable” amounts on hand, about 52 percent of all spending on these conditions would nonetheless remain uncompensated.

Our simulations suggest that a “reasonably foreseeable cost” standard based on expected costs would falsely identify many people who would not incur high costs while failing to address the majority of the uncompensated-cost problem. Because of differences in diagnostic and coding practices across countries, the task of matching patients to reasonably foreseeable costs would be much more challenging in an immigrating population. And because immigrants, on average, use health care services more frugally than do native-born Americans, our estimates of spending likely overstate their health care costs. Requiring individuals to accumulate resources against the risk of a potential medical condition is not a realistic way to address health costs. Providing legal immigrants with access to health insurance, for example, by giving them equitable access to the Affordable Care Act marketplaces, is a much better way to address potential uncompensated care costs than creating a costly system of medical examination and risk prediction. This policy is liable to needlessly exclude immigrants who could be more than healthy enough to contribute meaningfully to this country in exchange for a very modest reduction in uncompensated health care expenditures.