Trade, Technology, and the Labor Market
A $1,000 rise in import exposure per worker lowers the employment rate of non-college workers by an estimated 1.21 percentage points.
Trade and technology have quite different effects on U.S. businesses, according to new research by David Autor, David Dorn, and Gordon Hanson. In Untangling Trade and Technology: Evidence from Local Labor Markets (NBER Working Paper No. 18938), they find that local labor markets exposed to rising Chinese import competition see significant declines in jobs, whereas those susceptible to computerization see a polarization of occupations, but no net job loss.
The authors also observe that job loss accelerated during the 2000s in labor markets hit by foreign competition, while the impact of computerization decelerated, at least in manufacturing. "Our analysis reveals a surprising degree of divergence between the labor market consequences of these two phenomena -- both across industrial, occupational, geographic and demographic groups, and over time as the trajectory of these forces has evolved," they conclude. "Trade competition leads to sharp declines in local manufacturing employment, with corresponding growth in local unemployment and non-employment, particularly among workers without college education. In contrast, exposure to technological change has largely neutral effects on overall employment, yet leads to substantial polarization of occupational composition within sectors."
To approximate local labor markets, the authors examine 722 commuting zones (CZs), covering the entire mainland United States, from 1990 to 2007. They determine how susceptible each of these CZs was to technological change, based on Census data on occupational patterns, and to Chinese competition, measured by the local market's industry mix in 1980. The impact of technology -- or what the authors call "task-replacing technical change" through computerization -- is spread throughout the United States. In contrast, those CZs affected by trade competition from China are geographically concentrated, leading to quite a bit of geographical disparity.
For example, the Providence, Rhode Island CZ -- a manufacturing hub -- experienced an increase of imported Chinese goods of $2,330 per worker between 1991 and 2000. Between 2000 and 2007, the value of Chinese goods that might otherwise have been made locally rose to $3,490 per worker. By contrast, the New Orleans CZ had few industries directly competing with China and saw only small increases in import exposure: $170 and $490 per worker respectively in those two time periods.
Because the CZs are so disparate in terms of industry specialization, this regional approach makes it easier to identify the labor-market effects of trade shocks, which hit young and old, male and female, with roughly equal force. But the least educated are hit the hardest. A $1,000 rise in import exposure per worker lowers the employment rate of non-college workers by an estimated 1.21 percentage points; the impact on college workers is less than half that, 0.53 percentage points. These effects don't show up as much as a rise in the unemployment rate as they do in a decline in labor-force participation.
By contrast, computerization hits women harder than men. A CZ at the 75th percentile of exposure to computerization typically sees a decline of 1.8 percentage points in the female employment-to-population ratio over a decade relative to a CZ at the 25th. "The effects of exposure to routinization also appear larger for older versus younger workers, though this difference is less precisely estimated," the authors write. This study also looks at what kind of workers are affected by these changes. For trade shocks, the impact is largest for routine-task intensive occupations, such as repetitive production and office clerical jobs. It is also significant for manual-task-intensive jobs, like vehicle driving, cleaning, and security. The effect is evident, but not significant, for abstract-task-intensive occupations, such as managerial and technical workers. For computerization, the only significant negative impact is on routine task-intensive occupations; it is as large as the impact of trade shocks for this group. "But these losses are largely offset by employment growth in abstract and manual-task-intensive occupations," the authors write.
How can computerization have had such small, insignificant effects on manufacturing employment with all the labor-saving improvements of computer-aided manufacturing? Extending their study back to the 1980s, the authors find that computerization did have substantial impacts on job task composition in the 1980s and 1990s but not since then. In contrast, the effect of computerization in non-manufacturing industries has accelerated in that period, almost quadrupling from the 1980s to the 2000s. "Concurrent with the rapid growth of U.S. imports from China, the effect of trade competition on the manufacturing sector has become stronger over time, while the effect of technological change on employment composition in the manufacturing sector has subsided," the authors conclude. "Conversely, the impact of technology on the non-manufacturing sector is growing as technological change seems to be shifting from automation of production in manufacturing to computerization of information processing in knowledge-intensive industries."