Well-meaning if misguided politicians often tout job creation when they promote preferential trade agreements: freer trade will mean higher exports (the benefits of imports are almost never mentioned), and more exports means more production, which means more jobs. That narrow focus is understandable, especially in times of above-acceptable unemployment, when every bill seems to need a jobs angle to sell: witness, for example, Secretary of Agriculture Tom Vilsack’s verbal gymnastics in a statement yesterday, when he referred to the multi-year spending binge that is the Farm Bill as the “Food, Farm and Jobs Bill”. Politicians, not being the most courageous of creatures, don’t usually have the stomach/spine/etc. to make a principled stand in favor of free exchange across borders for its own sake (if you want to read more on the case for free trade, there are plenty of publications available at the Herbert A. Stiefel Center for Trade Policy Studies website, starting with our position statement at the bottom of the homepage).

Unfortunately, the extent of job creation is hard to quantify, and the case may be oversold. Edward Alden at the Council for Foreign Relations yesterday drew our attention to a recent International Trade Administration study that looks at the jobs created by exports and finds, well, that exports create fewer jobs than often is promised, and that the “job intensity” of exports is falling:

In 1993, which turns out to be the first year for which data are available, the report says that each $1 billion of exports supported just over 12,000 jobs… By 2011, however, that same $1 billion in exports supported only 5,000 jobs. About one-quarter of the difference is due to rising prices over the past two decades, but most of the difference is the result of higher productivity in export-intensive sectors which has reduced the need for labor.

That makes perfect sense. Some three-quarters of U.S. exports are in the goods sector, one-third of those directly in manufacturing, and many of the rest in industries that support manufacturing exports. From 1993 to 2011, labor productivity in the manufacturing sector doubled, compared with just a 50 percent increase in overall productivity. In other words, today it takes many fewer workers than in 1993 to produce the same $1 billion in exported products.

The consequence is that even rapidly growing exports have created very few new jobs. Given the average annual productivity improvements over the past two decades, exports need to grow by roughly 5 percent each year just to support the same number of jobs. Even with the extremely strong U.S. export performance over the past two decades, the total number of U.S. jobs supported by exports in 2011 — 9.7 million jobs – is up just 27 percent from the number of jobs in 1993…

The implication of these figures is fairly stark. As the Obama administration has noted often, export jobs are good jobs – employees in export-intensive industries earn some 20 percent more on average than comparable workers in industries that produce goods and/or services only for the domestic market. But there are simply too few of them to make a significant dent in unemployment, or to lift household incomes which have been flat for the past two decades. The largest number of export jobs are in technology-intensive industries such as aerospace, semiconductors, and motor vehicles. [links in original]

Is there anything wrong with making specious claims about jobs in support of something that is in the best interests of the economy? I guess that depends on your tolerance for spurious means in pursuit of worthy ends. For my part, I worry that when jobs claims are discredited, the case for free trade (and hence public support for it) is eroded. So my advice to politicians when it comes to prognostications about jobs created by trade agreements is: don’t go there. Be brave.