President Trump’s administration rolled out its most significant change yet to immigration policy Monday: the “public charge” rule. This regulation will ban legal immigrants if government bureaucrats believe that they might use welfare. Officials claim the rule will protect taxpayers and make immigrants self-sufficient, but the rule isn’t designed to meet these goals. It’s just designed to keep immigrants out.

In the late 19th century, when Congress first passed the law on which the regulation relies, a “public charge” meant a ward of the state — someone whom the government had the primary responsibility to care for. These people lived in “almshouses” and were entirely dependent on the government for their subsistence. Congress rightly decided it wanted only immigrants who could support themselves or be supported by family or private charity.

As almshouses disappeared and the modern welfare state emerged, implementation of the public-charge standard fluctuated arbitrarily depending on the administration. Finally, in 1999, the Immigration and Naturalization Service introduced a clear definition of a public charge: anyone who receives most of their cash income from the government. This standard accounted for both the receipt of benefits but also how much the immigrants were supporting themselves.

But the Trump administration rejects this historic understanding. Its rule will ban immigrants if they might — again, in the future — use any welfare at all (including noncash programs such as Medicaid) for more than 12 months in any 36-month period. Using four different programs for three months would count as 12 months of use, making anyone who falls on hard times for even the briefest period a “public charge.”

The real problem with the public charge’s new definition is not that it’s harsh; immigrants can certainly survive without welfare. The problem is that it’s economically misguided. No one who cares about public finances would design a rule that entirely ignores the degree to which the immigrants support themselves. Mere use — projected by a bureaucrat in any amount — would trigger a public-charge denial, even if it was a tiny fraction of the person’s income.

For example, a government adjudicator could estimate that that an immigrant would receive 95 percent of their income from private sources and just 5 percent from the government, yet that (projected) 5 percent would require the applicant be denied. Long gone are the days when public charges were people almost entirely dependent on the state. Now, this administration wants us to believe that people almost entirely independent from the state are public charges.

There’s an irony in how the rule defines “dependency,” too. At one point, it defines a dependent of an immigrant as someone who receives more than 50 percent of their financial support from that immigrant. But then, in its next breath, the rule labels a dependent of the government as anyone who receives anything — even less than 5 percent of their income — from the government.

It’s contradictory, but if the goal is just to reduce legal immigration, the contradiction makes perfect sense.

The problem with the rule runs deeper. The entire methodology of how a bureaucrat will predict someone’s future likelihood to use benefits is designed to set immigrants up to fail. It sets up a checklist of factors, including age, education, family size, English language ability, income, assets, etc. The government plans to count the negative factors and “weigh” them against the positive ones.

But empirically, it’s just not true that a person with two “negative” factors is twice as likely to use government benefits than someone with just one factor in isolation — or any more likely at all. For example, the government plans to “negatively weight” dropping out of high school and not speaking English. But someone who both dropped out of high school and doesn’t speak English is no likelier to use benefits than someone who just dropped out of high school.

This wouldn’t be a big deal if these factors in isolation were good predictors of welfare use. But according to data that the government cites in the rule, they aren’t. Nearly 70 percent of noncitizens who don’t speak English didn’t use any welfare at all in 2013. This was virtually the same percentage as for high school dropouts.

In other words, these factors aren’t good at predicting welfare use, and few immigrants — just 1 in 5, according to government data — used any welfare in 2013. Still, the government plans to use these factors anyway. Adding this inaccurate and misleading methodology to an already hopelessly flawed definition will produce disaster for legal immigrants. When the rule takes effect in October, legal immigrants will receive a blitz of denials.

Barring noncitizens from welfare makes sense. But banning immigrants from the country based on an inaccurate methodology won’t help public finances because most immigrants, according to the National Academy of Sciences, will pay more in taxes than they receive in benefits over their lives.

Far from helping to “protect taxpayers” and requiring self-sufficiency, the public-charge rule will harm the economy by turning away hard-working immigrants who are contributing to the United States. Unfortunately, the effect of the misguided policy is — for the president — a feature, not a bug