A educational worker in manufacturing earns more

Support for manufacturing is back on the agenda in America, thanks in part to concerns about long-term growth prospects and the "challenge" posed by China. Politicians as different as Barack Obama and Rich Santorum have expressed a desire to do more to help manufacturers. This all makes most economists uncomfortable, not least because most pro-manufacturing argumentation in recent decades has tended toward a thinly-veiled mercantilism. I, too, am sceptical about the importance of government support for machinery cone crusher manufacturing, or indeed of the broader merit of manufacturing activity, relative to other economic pursuits. Why does manufacturing matter?
Laura D'Andrea Tyson recently wrote a piece for Economix examining precisely that question, part of a new effort to make manufacturing support respectable among dismal scientists. I found it less than convincing. Manufacturing matters, she began, because America needs to close its current account deficit. Ok, but exports of services and resources combined are now larger than exports of manufactured goods. Are we sure it isn't easier, or more important, to scale up the former rather than the latter? Neither is it clear that manufacturing trade balances, as presently computed, accurately account for the actual American content in traded goods. (Very little of the value added in iPhones "imported" from China is accounted for by Chinese activity.)
Moreover, trade balances are essentially a macro issue. Even if one agrees that it's critical to increase net exports, it isn't clear that micro reforms are nearly as important as macro capital flows. The trade-weighted dollar is 30% weaker than it was a decade ago, and further declines are likely as (or if) the crisis-period flight to safe assets reverses. That, and not the merits of manufacturing, is the biggest trade concern.
Mrs Tyson goes on to argue that manufacturing matters because its jobs are high-productivity, and high value-added positions that tend to pay well. Perhaps, but this may well confuse cause and effect. High-wage manufacturing jobs tend to leave America when workers aren't sufficiently productive. There is a selection effect at work, in other words. Skill- and capital-intensity also matter, of course, but that's true across sectors. The Brookings Institution's Metropolitan Policy Programme hosted a symposium this morning on why, and which, manufacturing matters, in which a similar point was repeatedly made—wages are higher in manufacturing. One speaker noted that controlling for age and education a manufacturing worker such as a ball mill technician earns around 7% more than a non-manufacturing peer. But what if one controls for location? Is it more important to move people from non-manufacturing positions to manufacturing positions, or to move them from low-wage metropolitan areas to high-wage ones?