Skynet is not coming for our jobs

THE post-2008 economic recovery has been anemic by most measures, with job growth in particular making weak gains.  However, job growth in the U.S. has slowly declined over the past three decades, specifically since 2000 (the last year annual job growth in the U.S. hit 2%).

Among the myriad of explanations for why job growth has slowed, the rise of technology and the demand for higher skills is among the more popular (framed by a few as some ‘the robots are coming for our jobs‘ apocalypse).

Economists such as David Autor at MIT posit that technological changes are displacing workers both in terms of increased productivity, and the skills required or workers to use that technology.  As a result, job gains are either at the high end of the skill level spectrum or the low end (the so-called, job polarization), as evidenced by this graphic:


To counter this, workers need to increase their skill and education in order to obtain these high-skilled jobs.

This theory has been challenged.  For starters, massive investments were made in technology during the 1990s, which saw large job and wage gains across the skill level spectrum–something we should not expect if a rise in technology were inimical job and wage growth.

Moreover, if we break down the above graph, we see that job polarization has not been a problem since the end of the 1990s:


Now, a recent paper by Autor himself further dampens the technological change argument.  In that paper, Autor and his co-authors examine the impact of Chinese import competition on the labor market from 1999 – 2011.  They find that 2 to 2.4 million jobs were lost as a result of that competition.

Obviously import competition is not the only cause for stagnant job growth.  A paper published last year by the Brookings Institute examined the decline of the labor share of income in the U.S. (a metric different than employment, but indicative of a softening labor market).  They found that the labor share of income remained relatively constant prior to the 1980s, before rapidly declining:

The authors cited four reasons for this decline:

  • The headline measure overstates the decline of the labor share by one-third because of the way self-employment income is estimated;
  • The declining labor share is not merely a recent phenomenon — although the most recent decline has been dominated by changes in the trade and manufacturing sectors, even prior to 1980, there have been substantial, though offsetting, movements in labor shares within industries;
  • There is limited evidence that the share is dropping due to the substitution of capital for (unskilled) labor to exploit technical change embodied in new capital goods; and
  • The decline in the percent of unionized workers in the workplace is not, in fact, a major cause of the decline of the labor share.

Not only was the technological change argument questioned, but that trade (import competition) is a driver in the decline in labor share of income.  Again, trade is not the only driver of the decline in either job growth or labor share or income, nor is the role of technology muted.

However, the recent Brookings paper together with the more recent Autor work (as well as work by others) continue to point to U.S. policymakers as having a bigger hand in the decline in job growth (as well as wage stagnation) than is often argued.



John Haskell

About John Haskell

John graduated from the University of Southern Maine with a degree in Political Science, and from the University of Maine School of Law. He has worked in both the public and private sectors, and currently, works with a small business services company in the Mid-Coast area.