The Senate outline requires that all undocumented immigrants who came to the U.S. as adults register with the government. At that point, they'll have to pay back taxes and fines, pass a background check, learn English and a bit of U.S. civics, and demonstrate a work history and current employment. Only at that point will they get the right to apply for a green card, at which point they'll have to queue up behind the millions already waiting. Once they get the green card, it will take about five years before they can become full-fledged U.S. citizens.
Judged by a standard of basic fairness, all of this might seem like common sense. We don't like tax dodgers. Green card holders on the way to becoming citizens are already required to learn English. Waiting your turn doesn't sound like a unreasonable imposition.
But common sense has its limits. The danger is that by making the path to citizenship too arduous, we could discourage immigrants from coming out of the shadows. Take the issue of fines and back taxes. In 2007, most immigrant households earned less than $40,000 a year, according to the Migration Policy Institute (MPI). The 2007 Senate reform bill would have charged $5,000 just to initially register with the government, and another $5,000 for a green card. As the MPI chart below shows, that would mean a quarter of household income for three in five families.
Some past reform legislation has included smaller penalties. In 2010, New Jersey Democratic Senator Bob Menendez, one of the forces behind the new bipartisan plan, suggested a $500 registration fee. But add the price of English classes and back taxes, and the costs mount. Survey findings suggest that many immigrants who already have green cards don't go on to naturalization because of the costs involved. Despite our impulse to penalize immigrants who have broken the law, it's clear that each fine we place before undocumented workers is one more reason for many of them remain undocumented.
Fine, you might say. Their loss. But it's our loss, too. Limiting the number of undocumented workers is important so that they don't keep undercutting the wages of other low-skill legal immigrants, or even low-skill American workers. And while we can try to deal with that issue by making it more difficult for employers to hire undocumented labor. But as of now, our system for doing so is full of holes.
(3): E-VERIFY BECOMES A QUAGMIRE
The bipartisan Senate plan proposes to "create an effective employment verification system which prevents identity theft and ends the hiring of future unauthorized workers." But unless they're truly planning to build something from scratch -- in which case it could be a good long while before we see any useful software -- chances are the final product will be something based on E-Verify, the electronic employee ID system that was first piloted in 1997.
Today, 20 states require at least some employers to run their hires through E-Verify, which checks whether workers are legally in the country using data from the Social Security Administration and Department of Homeland Security. The federal government, meanwhile, checks all new workers against it. Two nasty problems in the system have revealed themselves.
First, it's expensive. A study by Bloomberg Government found that found that a national E-Verify law could cost small businesses $2.6 billion. That's on top of the hundreds-of-millions of government dollars needed to operate it each year.
Second, it's not effective. A December 2009 analysis commissioned by the government found E-Verify caught less than half of all undocumented workers because it was unable to detect identity fraud. In other words, a day laborer with a stolen Social Security number could slip right past Uncle Sam's watch. It gets worse: Overall, the system had an error rate of 4.1 percent. About one in six of those mistakes impacted legal workers who, for whatever reason, showed up as undocumented. In the end, the report found that about 0.5 percent of all workers are incorrectly flagged, and unable to correct their records. That may sound small, but as the Center for American Progress has noted, it means roughly 770,000 Americans could lose their jobs due to system error.
Those are the big, obvious issues with E-Verify, and it's possible they can be rectified. But if they're not, its easy to imagine the debacle that would follow. Consider what would happen if Washington makes a hash of fixing the identity fraud problems before rolling it out. At that point, instead of paying the full cost of legalizing, many undocumented workers could rationally decide that it would be more profitable to buy someone's Social Security number to evade the feds -- or just continue using one they've already purchased (as many as three quarters of illegal immigrants currently pay social security and medicare taxes using stolen numbers). If the problem is fixed, but the price of registering with the government is too high, E-Verify might encourage others to work entirely under the table -- off the books and off the tax rolls. In the meantime, hundreds of thousands of law abiding workers inappropriately blacklisted by the system might be faced with the bureaucratic nightmare of convincing Washington that they really should be allowed to have a job.
I could go on with the parade of pitfalls (note: I haven't even touched the guest-worker or high-skill-immigration plans.) But instead, I'll reiterate the takeaway: The framework for the plan isn't bad. But it's how Congress fleshes it out that counts. A system that's too punitive for aspiring citizens, or doesn't fix the glitches in employment verification, or doesn't create crystal-clear goals for the border could turn out to be as bad as the problems it was designed to fix. And while the spirit of compromise has been carrying this issue so far, some concessions for the sake of a deal might not be worth it.