Robots and firms

TY - JOUR

T1 - Robots and firms

AU - Koch, Michael Johannes

AU - Manuylov, Ilya

AU - Smolka, Marcel Buus

PY - 2021/8/1

Y1 - 2021/8/1

N2 - We study the microeconomic implications of robot adoption using a rich panel data set of Spanish manufacturing firms over a 27-year period (1990-2016). We provide causal evidence on two central questions: (1) Which firm characteristics prompt firms to adopt robots? (2) What is the impact of robots on adopting firms relative to non-adopting firms? To address these questions, we look at our data through the lens of recent attempts in the literature to formalise the implications of robot technology. As for the first question, we establish robust evidence for positive selection, i.e., ex ante better performing firms (measured through output and labour productivity) are more likely to adopt robots. On the other hand, conditional on size, ex ante more skill-intensive firms are less likely to do so. As for the second question, we find that robot adoption generates substantial output gains in the vicinity of 20-25% within four years, reduces the labour cost share by 5-7% points, and leads to net job creation at a rate of 10%. These results are robust to controlling for non-random selection into robot adoption through a difference-in-differences approach combined with a propensity score reweighting estimator. To further validate these results, we also offer structural estimates of total factor productivity (TFP) where robot technology enters the (endogenous) productivity process of firms. The results demonstrate a positive causal effect of robots on productivity as well as a complementarity between robots and exporting in boosting productivity.

AB - We study the microeconomic implications of robot adoption using a rich panel data set of Spanish manufacturing firms over a 27-year period (1990-2016). We provide causal evidence on two central questions: (1) Which firm characteristics prompt firms to adopt robots? (2) What is the impact of robots on adopting firms relative to non-adopting firms? To address these questions, we look at our data through the lens of recent attempts in the literature to formalise the implications of robot technology. As for the first question, we establish robust evidence for positive selection, i.e., ex ante better performing firms (measured through output and labour productivity) are more likely to adopt robots. On the other hand, conditional on size, ex ante more skill-intensive firms are less likely to do so. As for the second question, we find that robot adoption generates substantial output gains in the vicinity of 20-25% within four years, reduces the labour cost share by 5-7% points, and leads to net job creation at a rate of 10%. These results are robust to controlling for non-random selection into robot adoption through a difference-in-differences approach combined with a propensity score reweighting estimator. To further validate these results, we also offer structural estimates of total factor productivity (TFP) where robot technology enters the (endogenous) productivity process of firms. The results demonstrate a positive causal effect of robots on productivity as well as a complementarity between robots and exporting in boosting productivity.

KW - EMPLOYMENT

KW - EXPORTS

KW - GROWTH

KW - IMPACT

KW - INNOVATION

KW - JOBS

KW - PRODUCTIVITY

KW - RESEARCH-AND-DEVELOPMENT

KW - TECHNOLOGY

KW - TRADE

UR - http://www.scopus.com/inward/record.url?scp=85111752492&partnerID=8YFLogxK

U2 - 10.1093/ej/ueab009

DO - 10.1093/ej/ueab009

M3 - Journal article

SN - 0013-0133

VL - 131

SP - 2553

EP - 2584

JO - The Economic Journal

JF - The Economic Journal

IS - 638

ER -