The demand for H-1B visas to work in the U.S. is so high that last month, for the third time, the federal government announced it would use a lottery to distribute them.
To economists, a lottery is evidence of a problem: Something is preventing supply and demand from coming into balance. In this case, that “something” is the annual cap of 65,000 on H-1B visas, which are awarded to foreign workers with theoretical or technical expertise in specialized fields, including scientists, engineers, and programmers. And, oddly, fashion models. (Another 20,000 visas are available to workers with a master’s degree or higher.)
The latest quarterly issue of the American Economic Review, one of the top U.S. economics journals, contains an article arguing that H-1B visas benefit not only those who receive them, but also the companies that hire the visa holders. The journal was published online for subscribers this week. The article’s title: “Why Do Programmers Earn More in Houston than Hyderabad? Evidence from Randomized Processing of US Visas.”
Indian programmers who won H-1B visas the last two times lotteries were conducted, in 2007 and 2008, earned five times as much as those who got stuck back home—around $65,000 a year vs. around $12,000 a year—according to the article. Why didn’t U.S. employers try to save money by having the programming done in India, where it’s cheaper? Because there are important, if hard-to-measure, advantages of having programmers in the U.S. Such advantages perhaps include “the productivity effects of face-to-face interaction,” wrote the article’s author, Michael Clemens, a senior fellow at the Center for Global Development in Washington.
I talked with Clemens Tuesday by phone while he was on the Amtrak from D.C. to New York. Clemens said his paper demonstrates that employers suffer when they don’t manage to get the foreign workers they want because of the cap on H-1Bs. His paper found that many of those who were turned down by the U.S. went to other countries, where they made good livings helping those countries’ info-tech sectors.
“These workers massively contribute to the productivity of American firms,” he said “They generate American jobs. They foster innovation. We are arbitrarily sending them to other countries. We are gifting them to countries with which we compete for talent on a global scale.”
The H-1B lottery, while evidence of a dysfunctional market, is a gift to economists because it provides the randomization that is crucial to a statistically valid experiment. The applicants whose names are picked are just like the ones whose names aren’t picked. So if the winners who come to the U.S. make more money than the lottery losers who stayed home in India, it’s proof positive that something about being in the U.S. raised their pay. (By law, H-1B visa holders must be paid the same as Americans doing the same work.)
Taking advantage of the natural experiment made possible by the lottery, Clemens found that Indian programmers are far more valuable to U.S. employers when they are in the U.S. than when they are in India. To him, that’s evidence that the cap is inhibiting economically useful transactions. That contrasts with the view of H-1B critics such as University of California at Davis computer scientist Norman Matloff, who argues that “the H-1B work visa is fundamentally about cheap, de facto indentured labor.”
Said Clemens: “This is being done for purely political reasons. The H-1B cap was set by a political process, not an economic one. And that’s really unfortunate.”