Temperature Effects on Productivity and Factor Reallocation: Evidence from a Half Million Chinese Manufacturing Plants

Abstract

Understanding the relationship between temperature and economic growth is critical to the design of optimal climate policies. A large body of literature has estimated a negative relationship between these factors using aggregated data. However, the micro-mechanism behind this relationship remains unknown; thus, its usefulness in shaping adaptation policies is limited. By applying detailed firm-level production data derived from nearly two million observations of the Chinese manufacturing sector in the period of 1998-2007, this paper documents the relationship between daily temperature and four components in a standard Cobb-Douglas production function: output, total factor productivity (TFP), labor, and capital inputs. We detect an inverted U-shaped relationship between daily temperature and TFP; by contrast, the effects of temperature on labor and capital inputs are limited. Moreover, the response function between daily temperature and output is almost identical to that between temperature and TFP, thereby suggesting that the reduction in TFP in response to high temperatures is the primary driver behind output losses. In addition, temperature affects both labor and capital productivity. A medium-run climate prediction indicates that climate change will reduce TFP by 4.18%, and result in output losses of 5.71%. This loss corresponds to CNY 208.32 billion (USD 32.57 billion) in 2013 values. Given that TFP is invariant to the intensity of use of labor and capital inputs and reflects both labor and capital productivity, the Chinese manufacturing industry is unlikely to avoid climate damages simply by implementing factor allocation. Thus, new innovations that expand the technology frontier for all inputs should be developed to offset weather-driven TFP losses if other adaptation strategies are infeasible.