E‑Verify is a government program that is intended to stop the hiring of illegal immigrants. Firms that sign up for E‑Verify are supposed to enter the identity information of new hires into the government’s E‑Verify website. E‑Verify then checks that identity information, which is supposed to come from identity documents that the new hire provides, against information in government databases to guarantee that the new hire is eligible to work in the United States. The state of Florida is currently considering whether to mandate E‑Verify for some employers.

If E‑Verify returns a tentative non-confirmation (TNC) then the employer and new employee have a specific amount of time to resolve it by correcting errors or supplying different information. If E‑Verify is not satisfied then it issues a final non-confirmation (FNC) and identifies the new hire as an illegal immigrant. At that point, the employer must fire the new hire. The intent of the program is to make it very difficult or impossible for illegal immigrants to work in the United States.

That’s how E‑Verify is supposed to work. Based on a government answer to a FOIA that Cato filed, we are now able to look at the number of E‑Verify cases per state per year and the number of cases where E‑Verify found that a new hire is ineligible to work. The number of cases is the number of E‑Verify checks run by employers. These new data confirm one major problem with E‑Verify and reveal another: It just doesn’t work.

First, E‑Verify mandates do not result in mandatory use of E‑Verify. As we’ve explained elsewhere, a large percentage of new hires aren’t even run through the program. Arizona mandated E‑Verify for all new hires beginning on January 1, 2008. From then through the second quarter of 2016, the last quarter of information provided by the FOIA, an annual average of just 67 percent of new hires were run through E‑Verify (Table 1). According to Arizona’s mandate, 100 percent of new hires in Arizona were supposed to be run through E‑Verify. If Arizona can’t even force about one-third of all new hires to be run through E‑Verify then the program is in deep trouble.

The new problem revealed by this FOIA is just how few new hires are rejected because the system identifies them as illegal immigrants and, therefore, unauthorized to work. Beginning in 2008 when E‑Verify was mandated in Arizona for all new hires, the share of E‑Verify cases where the system identified the new hire as an illegal immigrant and therefore unauthorized to work peaked at 0.7 percent (Table 2). That was the first year of the mandate. After that, the percent dropped to about 0.1 percent in 2014 and has stayed there.

Of course, illegal immigrants are a much higher share of the population in Arizona than 0.7 percent or 0.1 percent. In other words, many thousands of illegal immigrants are getting jobs and either their employers are not using E‑Verify or E‑Verify is approving them to work. Figure 1 shows that about 6.4 percent of Arizona’s population were illegal immigrants in 2008, which fell to 4 percent in 2016 according to evidence from Pew and the American Community Survey.

Table 2 and Figure 1 likely downplay the ineffectiveness of E‑Verify because illegal immigrant men have labor force participation rates and employment rates above those of legal immigrant and native-born men. It could be that illegal immigrants are much less likely to switch jobs after E‑Verify is mandated, a phenomenon called “job lock.” This is recorded as reduced hires and reduced separations in government economic data. A recent working paper found that job lock among likely illegal immigrants does increase after E‑Verify mandates, decreasing new hires by between 5 and 7.3 percent and separations by between 6.1 and 7.3 percent.

But job lock cannot account for more than a small fraction of the tiny number of illegal immigrants excluded from employment under E‑Verify in Arizona. According to experimental JOLTS data on the state level that measures the amount of new hires and separations, there were about 9.7 million total separations in Arizona from January 1, 2008 through the second quarter of 2016 and about 9.8 million total new hires. Many illegal immigrants undoubtedly left Arizona to avoid E‑Verify, but a substantial number of illegal immigrants switched their jobs in Arizona after E‑Verify’s enactment and the system did not catch them.

Illegal immigrants are getting around E‑Verify in Arizona. Very few of them are being denied jobs by E‑Verify. Even by the standards of immigration restrictionists, it’s becoming an ineffective boondoggle. It’s time we ditch this program and start from scratch.