Do Industrial Robots Contribute to Unemployment and Lower Wages?

Over the past few years, rapid advances in technology have posed greater threats to jobs, especially those vulnerable to automation. Recently, McKinsey published a report analyzing the technical feasibility of automation in several occupations and concluded that service, manufacturing, and construction industries are all at risk of automation. Yet, the probability of automation varies considerably within industries and across sectors. Indeed, the feasibility of automation, though important, is not the only factor that plays an important role in whether or not machines can replace human labor. In a new study, Daron Acemoglu and Pascual Restrepo build a model to explain the aggregated effect of robots on employment and wages in the U.S. labor market.

In their study, the authors focus specifically on the effects of industrial robots. Industrial robots are those that do not require a human operator and can also be programmed to perform routine and manual tasks such as assembling, packaging, and painting. Their analysis begins by building on a theoretical model that considers variations in the adoption of industrial robots across industries and also incorporates labor market trading. From the model, the authors identify two effects: a positive effect caused by the increase in productivity from the adoption of robots and a negative one as a result of workers being displaced.

In order to empirically evaluate their theoretical results, the authors apply their framework to the U.S. labor market. Their data consist of the stock of robots by industry, country, and year from the International Federation of Robotics (IFR), which researchers used to build a measure of robot exposure; employment data by country and industry in 1990 from the EU KLEMS dataset; and census data from the America Community Survey. The data from these sources were used to construct measures of employment by industries including: manufacturing, construction, education, research and development, and other non-manufacturing industries.

Based on the available data the authors used commuting zones as a proxy for the U.S. labor market. Additionally, they control for trade patterns by constructing measures of exposure to imports from China and Mexico, as well as another measure for exports from Germany, Japan, and Korea because they have a higher adoption rate of robots. Finally, they build a measure of offshoring using the shares of intermediate inputs that are imported by each industry.

The results show a significant and negative effect of robots on employment and wages. Analyzing the results by commuting zone, they find that, on average, one additional robot per thousand workers now reduces aggregate employment to population ratio and wages by 0.34 percentage points and 0.5 percent, respectively. In addition, they show that one more robot per a thousand workers on average reduces employment by 6.2 workers and decreases annual wages by $200 in affected commuting zones. When analyzing the effects of industrial robots in the aggregated economy they find that one more robot per thousand workers reduces aggregate employment by 5.6 workers and wages by 0.25 percent. Finally, the authors argue that the effect of robots on the labor market is different from the effect of imports from China and Mexico, and from the other measures for which they controlled.

This research provides evidence of a negative impact on employment and wages due to the adoption of industrial robots in the U.S. labor market. Most importantly, it brings policymakers and researchers one step closer to understanding how the accelerated pace of automation can have aggregate impacts on the economy. Moving forward, challenges will arise in discovering the best and most effective way to reconcile the benefits and costs of these modern technological advances.

Article source: Acemoglu, D., & Restrepo, P. “Robots and Jobs: Evidence from US Labor Markets.” National Bureau of Economic Research, Working Paper No. 23285. (2017).

Featured photo: cc/(PhonlamaiPhoto, photo ID: 612375398, from iStock by Getty Images)