Skip navigation
Robots at Work Koichi Kamoshida/Getty Images

The Pessimist's Guide to the Robot Invasion

New research blames robots for slashing U.S. wages and worsening pay inequality.

Industrial robots have had a "large" and negative effect on U.S. employment and wages in local labor markets, according to new research by Massachusetts Institute of Technology's Daron Acemoglu and Boston University's Pascual Restrepo.

One additional robot per thousand workers reduces the employment-to-population ratio by 0.18 percentage points to 0.34 percentage points and slashes wages by 0.25% to 0.5%, based on their analysis. To put that in context, the U.S. saw an increase of about one new industrial robot for every thousand workers between 1993 and 2007, based on the study.

"The employment effects of robots are most pronounced in manufacturing, and in particular, in industries most exposed to robots; in routine manual, blue collar, assembly and related occupations; and for workers with less than college education," the authors write. "Interestingly, and perhaps surprisingly, we do not find positive and offsetting employment gains in any occupation or education groups." Worth noting: the authors estimate that robots may have increased the wage gap between the top 90th and bottom 10% by as much as 1 percentage point between 1990 and 2007. There's also room for much broader robot adoption, which would make all of these effects much bigger.

The unemployment rate has more than halved since its 2009 peak, yet it fails to account for some potential workers: it doesn't capture all of the people who are underemployed or who aren't actively applying to jobs. To examine what's happening in the broader population, Federal Reserve Bank of San Francisco Senior Economist Marianna Kudlyak revisits an index of non-employment in a new analysis.


The alternative measure takes all non-employed people into account, splitting them into different groups that are weighted by their historical likelihood of transitioning into a job. For example, retirees have less than a 2% chance of jumping back in, while non-retirees who want a job have a higher likelihood. The alternative index is close to its 2005-6 levels, so it "tells a story similar to the unemployment rate — that the U.S. labor market has returned to full health," according to the report.

The mortgage-student loan tie

U.S. student loan defaults shot up during and after the Great Recession, and the crashing housing market might have played a significant role in that. That's because plummeting home prices coincide with lower consumer demand and disruptions in the labor market, impacting borrowers' ability to pay, according to research by New York University's Holger Mueller and Constantine Yannelis. The fall in home prices during the downturn accounts for about 24 to 32% of the increase in student loan defaults, they find.

R* on the rise?

Central bankers and policy watchers are fixated on r* — a fantastically nerdy name for the neutral interest rate setting that neither slows nor stimulates the economy. Many researchers say the number has dropped in advanced economies due to aging demographics and depressed productivity growth. If that's the case, it means the Fed and many other central banks have less room to increase borrowing costs, leaving less space to cut during the next recession.

Goldman Sachs economists think the doom and gloom may be overblown. They agree that the neutral rate fell substantially after the crisis, but they view the decline as more of a business-cycle quirk than a permanent situation. Among other factors, they point out that the labor market in advanced economies is outperforming expectations, which suggests that the recovery hasn't been as stagnant as growth numbers might suggest.

 If they're right, it could mean that interest rates will go higher than many expect in this hiking cycle. "We expect the nominal funds rate to reach 3.25 to 3.5% at the end of 2019, significantly above current market pricing," according to their note.

by Jeanna Smialek, Bloomberg

Hide comments


  • Allowed HTML tags: <em> <strong> <blockquote> <br> <p>

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.