After my previous close encounter with Michael Dobb, the Post’s post-modern fact-checker (“depending on the statistics you use, Mexican economic growth over the last two decades has been either 337 percent, 125 percent, or 83 percent.”), I was prepared to ignore his latest attack on Democratic presidential candidate John Edwards. After all, I have limited experience in a world where any number is as good as any other number. But, here at BTP we maintain a commitment to the real world and real numbers. So, here’s the deal. Mr. Dobb lists as one of the big fibs of 2007 Senator Edwards’ statement that NAFTA cost us a million jobs. By way of response, he cites a study by the Congressional Research Service (CRS) that reports that NAFTA had “little or no impact” on the aggregate number of jobs in the United States. While Dobbs wants to use this discrepancy to imply that Edwards is a big liar, there is actually a very different story at play here. Edwards was referring to the number of jobs lost in the United States due to the fact that the U.S. ran a trade deficit with Mexico of close to $80 billion last year. This trade deficit is approximately equal to 0.6 percent of GDP, which means that, other things equal, GDP in the U.S. would have been 0.6 percent higher if we had balanced trade with Mexico. If we assume that job loss was proportionate to GDP loss, then a 0.6 percent drop in employment would correspond to roughly 840,000 jobs, a figure not hugely different from Senator Edwards 1 million figure. Of course, other things are not equal. There was job growth in other sectors that largely offset the jobs lost from NAFTA, hence the CRS statement that NAFTA had little or no impact on the number of jobs. So is Edwards a big liar because he only talked about the gross job loss from NAFTA rather than its net effect on employment? Well, that seems a real stretch to me. Politicians routinely talk about the jobs they create from some particular project or policy. For example, it is very common for a politician to say that building a highway created X thousand jobs or that exports to country Y created Z thousand new jobs. These claims also refer to gross job creation. If we asked our friends at CRS, they would almost certainly answer that these policies had “little or no impact” on the aggregate number of jobs in the United States. In other words, these jobs gains are mostly just displacement. Workers who would have otherwise have been employed elsewhere are instead employed building a highway or exporting goods to country Y. So Edwards was using a job loss figure in a way that is standard for politicians, and in fact newspapers and policy wonks, to talk about job creation. Should he have pointed out that his figure was a gross number and did not mean a net loss of jobs? I would have liked to see this, but I would also like to see newspapers put budget numbers in context too. Edwards was using numbers in a way that conformed to industry standards. For this, he does not deserve to have the Post call him a liar.
--Dean Baker Addendum: It has been pointed out to me that Edwards said "millions" of jobs, not 1 million jobs. If this is correct, then Edwards did make a serious exaggeration that deserves to be corrected.
