Trade uncertainty, automation, and labor markets – U.S.
Hamid Firooz, Sylvain Leduc, and Zheng Liu in this paper, develop a small open economy model featuring labor search frictions (Diamond-Mortensen-Pisarides framework), endogenous automation decisions, and heterogeneous worker skills. The model incorporates international trade in intermediate inputs with time-varying iceberg trade costs that have stochastic volatility to capture trade uncertainty from geopolitical, climate, and trade policy risks. For the empirical analysis, the authors use industry-level data from the U.S. covering 14 two-digit ISIC industries from 1997-2022. Key data sources include industrial robots data from the International Federation of Robotics, the Trade Policy Uncertainty (TPU) index TPU index constructed by Caldara et al. (2020), which is based on the frequency of articles in several major U.S. newspapers that discuss economic policy uncertainty and contain one or more phrases related to trade policy. The empirical strategy employs panel regressions with industry and time fixed effects, and two-stage least squares estimation to identify the automation transmission channel.
Key findings
- Trade uncertainty creates incentives for reshoring but does not necessarily bring jobs back to the home country when firms have access to labor-substituting automation technologies. The study finds that heightened trade uncertainty reduces imports while boosting automation investment, particularly in industries more exposed to offshoring.
- Increased automation raises labor productivity and facilitates reshoring, but also displaces unskilled jobs. The threat of automation weakens the bargaining power of unskilled workers in wage negotiations, depressing their wages while raising skilled wages, resulting in higher skill premiums and wage inequality.
- Trade uncertainty operates through three key channels: (1) an expenditure-switching effect that redirects demand toward domestic producers, (2) a precautionary savings channel that lowers interest rates and stimulates automation, and (3) an option value channel that may delay automation but is dominated by the other effects under their calibration.
Reshoring, Automation, and Labor Markets under Trade Uncertainty
Authors: Hamid Firooz, Sylvain Leduc, Zheng Liu
From: North Carolina State University, Federal Reserve Bank of San Francisco
Trade uncertainty, automation, and labor markets – Globally
Marius Faber, Gleb Kozliakov, and Dalia Marin in this paper analyze data for 18 developed countries, 17 developing countries, and 19 industries over the period 2000–2014, exploiting variation in pre-existing trade linkages to measure each developed country–industry pair’s exposure to uncertainty shocks in developing countries. Uncertainty is captured by the World Uncertainty Index, which counts occurrences of “uncertain” and its variants in Economist Intelligence Unit reports. The authors employ a shift-share (Bartik) design to generate plausibly exogenous variation in developing-country uncertainty, interacting this exposure with an industry-level robot replaceability index to assess how automation capabilities condition reshoring responses.
Findings
- Higher uncertainty in developing countries induces reshoring of production inputs to developed countries—but only in industries with high robot replaceability. A one-standard-deviation increase in exposure to uncertainty raises the ratio of domestic to foreign inputs by approximately seven percent.
- This reshoring effect stems entirely from reduced imports rather than increased use of third-party domestic suppliers, implying that firms bring production in-house when faced with uncertainty.
- The reshoring response more than triples in the post-Global Financial Crisis subperiod, reflecting heightened risk aversion, advances in automation, and a low-interest-rate environment.
Global Value Chains in a World of Uncertainty and Automation
Authors: Marius Faber, Gleb Kozliakov, Dalia Marin
From: Swiss National Bank & University of Basel, TU Munich & UC Davis
