There’s been a lot of talk recently about automation of tasks leading to job losses but David Autor, David Dorn, and Gordon Hanson offer evidence (PDF | via Tyler Cowen) that the most important technology displacing American workers has been the container ship:
We analyze in a common framework the differential effects of trade and technology on employment patterns in U.S. local labor markets between 1990 and 2007. Labor markets whose initial industry composition exposes them to rising import competition from China have experienced significant employment reductions particularly in the manufacturing sector and among non-college-educated workers. These employment losses are not limited to manual production jobs but also affect clerical and managerial occupations. Labor markets that are susceptible to computerization due to specialization in routine task-intensive activities have neither experienced an overall decline in employment, nor a differential change in manufacturing employment. However, the occupational structure of employment of these labor markets has polarized within each sector, as employment shifted from routine clerical and production occupations to more highly skilled managerial or professional occupations, as well as to lower skilled manual and service occupations. While the effect of trade competition is growing over time due to accelerating import growth, the effect of technology seems to have shifted from automation of production activities in the manufacturing sector towards computerization of information-processing tasks in the service sector.In a sense it doesn’t really make a difference. In fact, one way I’ve seen the case for free trade in goods sold is with the observation that the economic impact is essentially the same as with any technological improvement. Whether through the use of robots or the use of Chinese labor, it’s become possible for Americans to consume the same quantity of goods while using less American labor thereby freeing up more workers for use in the health care or construction sectors. But there are two problems here. One is that while in the 1990-2007 period we really did see more workers in the health care and construction sectors (and thus bigger houses and more medical services) in the 2008-2013 period we’ve experienced a boom in the unemployed and out of the labor force sectors. That’s a tragic waste of human and material potential and a massive failing of members of congress and the Federal Reserve. The issue looming over the entire period, however, is distributional. With any useful new technology you expand the pie but expand it unevenly. Desktop publishing softare and digital photography have been great for the world, but were terrible for people like my mother who had specialized skills in analog page design. The “technology” of imports from China has, similarly, been good for doctors and nurses but quite bad for manufacturing workers. What you’re supposed to do in response to that kind of dynamic is use progressive taxes to financial broad social insurance programs. And we’ve done some of that during this period—the Bush administration expanded Medicare coverage and the Affordable Care Act will funnel a lot of new money to the working class—but at the moment we’re talking about ways to cut Social Security which seems perverse.
