Story from the trenches: When state laws slam into federal rules

This dispatch from the mean streets of Phoenix:

One of our clients hired a woman and, complying with Legal Arizona Workers law and the E-Verify rules, ran her through the database within three days of hire.

She had recently gotten married and used her married name in the application. But she hadn’t changed her information with the Social Security Administration. Not surprisingly, E-Verify returned a tentative non-confirmation. Our system provided the forms needed and she headed over to the local SSA office.

They changed her information in the system. Told her they’d send her a new SSN. And told her it was OK in 24 hours to resubmit her case.

The business did. And she came up with another tentative non-confirmation. A call to the SSA. “No, you’re in the system. Give it a day or so.”

Two days later, another submission attempt. Another tentative non-confirmation. Federal rules state that she can’t be fired during the allowed 10-day resolution period. But now those 10 days are up, and despite SSA assurances, she’s termed out. If the company keeps her on, they’ve essentially committed the same sin in the State of Arizona’s eyes as if they knowingly hired an illegal employee and kept them on despite the non-resolution of a tentative non-confirmation.

This is what happens when the state law — a non-positive in E-Verify equals a presumptive illegal worker or a hell of a lot of paperwork explaining why the worker’s legal — crashes into both federal rules and bureaucracy.

The final resolution: We called the E-Verify folks and found that the system hadn’t released the initial report. So we killed out the report, the company re-ran the employee, she came back “SSA employment authorized” and she’s working again.

Arizona statistics and what they mean

We’re big Mark Twain fans, and when he says there are “Lies, damn lies and statistics,” we tend to agree.

But here’s the cold hard facts about how the first quarter returns came out for our clients:

Percentage of forms run resulting in “tentative non-confirmation” returns: 8

Percentage of “tentative non-confirmation” returns that were uncontested: 9 percent

Percentage of “tentative non-confirmation” returns that were contested and resolved positively: 94 percent

So what does it mean?

The most salient stat is the 94 percent positive resolution rate on tentative non-confirmations. It tells us that the E-Verify database still has its error problems (roughly an 8 percent error rate, based on USCIS data released in December 2007; full evaluation is here) but is doing its job pretty well.

You’re looking at a 425 million record database, and with that error rate, you’re looking at 34 million mistakes. The most common error flags that we see is name mismatch, which seems to be centered on the middle name. For instance, “Mary Elaine Smith” is listed on the Social Security Card and “Mary Smith” was entered into the E-Verify database. The result was a tentative non-confirmation.

We terminated the query and re-entered the exact same information but included the middle name and got an “SSA Employment Authorized” back. Go figure.