Nearly twenty years ago, John J. Miller of the Center for Equal Opportunity and Stephen Moore, then the director of fiscal policy studies at the Cato Institute, published a study responding to the rising demand for immigration law enforcement.

“A National ID System: Big Brother’s Solution to Illegal Immigration” was the name of their Cato Institute policy analysis. They highlighted costs to the liberty of native-born Americans from systems that seek to root out illegal immigrants with identity cards and tracking. I reprised their study in a way and expanded on it seven years ago in “Electronic Employment Eligibility Verification: Franz Kafka’s Solution to Illegal Immigration.”

When I saw Alex Nowrasteh’s research into the results of mandates to use the Department of Homeland Security’s E‑Verify program, I was delighted to see what experience makes available to backers of “internal enforcement” who don’t have our nation’s freedoms in mind. E‑Verify simply does not work. That’s the upshot of our new study, “Checking E‑Verify: The Costs and Consequences of a National Worker Screening Mandate.”

The theory behind checking workers’ data against federal databases when they are hired is that this would exclude illegal immigrants from employment. Doing so would turn off the “jobs magnet” that brings illegal immigrants into the country, and illegal immigration would fall.

But the results from states where E‑Verify has been mandated show a remarkable failure to change the incentives of migrants, as E‑Verify is supposed to do. The wage gains that a Mexican laborer experiences by coming to the United States in the absence of E‑Verify is 253 percent. In Arizona, E‑Verify lowered that wage gain to 240 percent.

The migrant’s calculation is unchanged by E‑Verify: Continue to seek the benefits of working in the United States.

The reasons for E‑Verify’s failure are several. Implementation, for example, has been weak. As noted in a Washington Examiner article on our study, in Mississippi, only 49 percent of new hires were run past E‑Verify despite a universal mandate in that state. Arizona saw 59 percent of new hires put through the system, the highest compliance rate of any state.

Where employers are using E‑Verify, workers can avoid it by using names and Social Security numbers that match up, easily defeating the system so as to get the benefits of working.

For all its wonders, technology is not something policymakers can sprinkle on deep-seated economic and social problems to make them go away. E‑Verify illustrates this. The promise of technology-charged identity and background checks at state or national scale does not survive contact with the real world.

Given its meager immigration-control benefits, the costs of E‑Verify—in dollars and in freedom for employers, workers, and Americans generally—are too high. Here’s hoping that the advocates of “internal enforcement” of immigration law are willing to reap the benefits of experience.

The path forward for immigration law reform is not heavier, more intrusive, and more costly “internal enforcement.” It is creating broader avenues for workers to enter the country and contribute their labor to the American experiment.