New Data on Outsourcing

The BLS released new estimates this week of the number of jobs lost in the US to outsourcing. The BLS data lends support to an argument that I have made here before: outsourcing has not been a major source of the weak US labor market. Note that the BLS data is not the first of its kind; a study by the IIE that I wrote about a few months ago found basically the same thing. To be fair, however, the BLS data is not perfect, and may indeed understate outsourcing job losses in a couple of ways. This NYTimes article points out a few possible criticisms of the BLS figures.

Officials acknowledged that the numbers clearly undercount the total number of jobs lost offshore. For one thing, the new data covers layoffs only at companies employing at least 50 workers where at least 50 filed for unemployment insurance and the layoffs lasted more than 30 days. Even more important, the report does not account for jobs created by American companies overseas that did not involve a direct layoff in the United States.

First of all, the last sentence is meaningless; if the job created overseas did not involve a layoff in the US, then why should we worry about it? The other point has more merit. But I’m skeptical that one could find a substantial impact on the US job market by looking at layoffs of fewer than 50 people. If a few hundred firms outsourced some portion of their business last year, yet laid off only a 10 or 20 people at a time, then this would plausibly add maybe 5, 10, or perhaps even 20 thousand people to the ranks of the unemployed, but such numbers are still a tiny fraction of the millions of jobs lost during the Bush presidency.

Another criticism is that the survey is self-reported, so some people argue that many firms that have reduced US employment due to outsourcing simply did not report it accurately. This article on CNN/Money gives an example:

Thea Lee, the chief international economist for the AFL-CIO, said she believes the BLS report greatly underestimates the amount of jobs lost to overseas outsourcing. “This is a survey that is anonymous and self reported — if they’re given a choice between vague company reorganization and outsourcing, they can chose to put reorganization and there will be no follow up,” she said. “Another 7.2 percent refused to answer question. Those who have looked more deeply into the extent of outsourcing have found far greater use” of it.

I’m also skeptical of this argument, however; why would firms lie on the survey? True, firms may not want to publicly admit to outsourcing jobs overseas, but it’s common knowledge that survey results by the BLS are always kept completely anonymous. I suppose that it’s possible that firms didn’t believe in the anonymity of the survey and therefore lied in their responses, but I need some more convincing rather than a simple assertion to that effect.

The critics of the BLS’s outsourcing job-loss numbers also point to other outsourcing surveys (such as the famous one by Forrester Research, or another by Goldman Sachs) that show that US firms hire hundreds of thousands of workers overseas every year. The problem is that it’s wrong to assume that every worker being hired overseas causes a US worker to lose a job. It’s perfectly possible – and indeed, exactly what we would expect – that US firms would be hiring workers overseas while not laying off people at home. So both the outsourcing survey data that shows lots of outsourcing happening, and the BLS data that shows few job losses as a result, can simultaneously be true. In fact, they are both perfectly consistent with economic theory. And when you have data and theory in agreeing with each other, Occam’s Razor (and good economic practice) suggests that the theory and data are probably both correct.

My conclusion is that there is a lot of evidence that outsourcing has a tiny impact on the US job market, and that its effects are swamped by other things happening to the economy. If you worry about the health of the labor market in this country (and I still do, despite the recent moderately strong job growth), then worry about the Bush administration’s poor fiscal management, tax policies, and other policies (such as my pet theory, competition policy) that are overwhelmingly to blame for the fact that the US economy is still millions of jobs short of what it could otherwise have.

Kash