Debunking the “E-Verify Error Rate”

Everyone agrees that illegal immigrants cross our borders because they want to work here. If we can reduce the lure of illegal employment, we can reduce the pressure on our borders.

That’s exactly what E-Verify does. When an E-Verify employer hires a new worker, the employer gets on line and fills out a short electronic form. As soon as the employer hits “send,” the system checks to make sure that the worker’s name matches his Social Security Number. If the worker is not a U.S. citizen, the system also checks to make sure his work authorization is still valid and shows the employer the picture that should be on the DHS-issued identity card. For most workers, verification is instantaneous.

E-Verify is simple, free, and highly effective in preventing illegal work. It works, and maybe that’s what the interests arrayed against E-Verify don’t like. Whatever the reason, opponents of E-Verify have resorted to charges that just don’t hold up. In this series, I debunk the myths.

The opposition to E-Verify often claims that the program has a high error rate. Some critics claim that the error rate is as high as 4% and will lead to millions of Americans losing their jobs by mistake.

To see how wrong this claim is, we need to look more closely at how E-Verify works. We can draw a precise picture of what happens to a thousand applicants who use E-Verify by using data gathered from October 2006 to March 2007 by Westat, an independent reviewer.


Of the thousand, 942 are instantly verified. Instant verification of legal workers surely can’t be an error.

Fifty-eight are told that they have to do something more to establish that they are lawfully authorized to work. Usually this means they have to go to Social Security to correct the mismatch in name and number. (Typos and similar problems are cured on line, so legal workers usually have a problem only if they changed their names or citizenship status but failed to tell Social Security of the change.)

So five of the thousand must go to Social Security and straighten out their records. For 90% of them, the process takes less than 2 days. Is that an error rate? If so, it’s ten times lower than our critics claim. And, is it really an error to tell workers that their social security credits aren’t being properly recorded? Sooner or later, the worker will want to collect benefits, and they won’t want to face doubts about who earned the credits. (Of course, straightening out Social Security records isn’t fun, but we’re working to reduce the hassle. Just a few weeks ago, we introduced software changes that will automate some of the correction process, reducing the number of legitimate workers who have to go to Social Security offices from five to two or three per thousand.)

That leaves the 53 who walk away. Is that an error rate? There are certainly people who believe it’s an error to keep illegal workers out of the U.S. workforce. But we don’t. It’s our job to enforce the immigration laws.

And common sense suggests that the walkaways are overwhelmingly likely to be illegal workers. It’s just common sense that a legal worker wouldn’t want to walk away from a job he applied for--and has been offered if he straightens out his records. It’s just common sense that a legal worker wouldn’t walk away from the opportunity to correct Social Security records he now knows are wrong – records that will have to be corrected for him to get benefits. And it’s just common sense that about five percent of E-Verify workers would walk away, since a Pew Foundation expert recently estimated that 4.9% of U.S. jobs are held by illegal workers. It’s hard to see the walkaway rate as an error; in fact, that’s the program working as it should.

Stewart Baker
Assistant Secretary for Policy