Do climate shocks induce technological change? Firm-level evidence from wildfire exposure and robot adoption

Abstract

Purpose The aim of this paper is to examine whether wildfire exposure is related to firms’ adoption of industrial robots. The study focuses on Chinese manufacturing firms over 2001–2014 and analyzes whether the wildfire–automation relationship varies across adoption margins, financial constraints, population mobility, labor intensity, and time. Design/methodology/approach This study builds an annual firm-city panel by matching firm-level robot import records with city-level satellite wildfire exposure. Robot adoption is measured by imported robot quantity and value. Wildfire exposure is instrumented with the Forest Fire Weather Index. The models include firm and year fixed effects, firm- and city-level controls, and city-level clustered standard errors. Findings The results indicate a positive estimated relationship between wildfire exposure and imported robot adoption. This relationship appears for both robot quantity and robot value. The moderation evidence for financial constraints is mixed, and the joint model does not identify a statistically significant constraint-based difference. The distributed-lag estimates point in the same positive direction, but they should be regarded as suggestive rather than strong evidence of dynamic adjustment. Population mobility and labor intensity show limited moderating roles. The results remain positive across alternative wildfire measures, alternative clustering choices, and a restricted sample excluding firms observed in more than one city. Originality/value This paper extends recent firm-level evidence on wildfire exposure and industrial robot adoption by examining adoption margins, financing frictions, local adjustment conditions, and timing. It contributes to research on climate shocks, technological change, and firm behavior by interpreting the wildfire–automation relationship as a capital-adjustment process rather than only as an average adoption effect.

The authors' abstract, as published at the source. Journal of Economic Studies, 2026 · DOI ↗