Endogenizing Temperature Impacts on Economy: Causal Effect of Climate Indicators and Economic Development

Temperature-economic relationships have been studied, while temperature often links as the indicator with economic projections in climate impacts study. Research have supported other climate variables may also contribute to the changes in economy. To verify temperature is a suitable climate indicator on economic development, this paper investigates the existence and direction of Granger causality between 12 surface climatic variables and economic growth. Empirical results for China over the period 1980-2018 suggest, temperature variables exhibit a unidirectional Granger causality to the per capita GDP changes. While unidirectional Granger causality runs from the per capita GDP to other climatic variables such as humid, precipitable water content, and u-wind. The unidirectional Granger causality flow running from temperature to economic development support endogenizing temperature impacts on economy. Thus, we apply a conditional Breitung and Candelon causal effect of temperature on GDP based on Cobb-Douglas-Climate function, by constraining capital investment and labor input. To assess the contribution of temperature impacts on economic changes, a multi-variable OLS of provincial-level panel data is constructed to test the Shapley value decomposition of the temperature variables. Among the three temperature variables, air temperature presents with the highest contribution ratio. Therefore, the linkage of regional temperature changes as indicators and estimators for long term climate impacts on economic development is among the first choice of evaluating climatic variables impacts. This research supports endogenizing temperature impacts on economy. It facilitates a modelling design of endogenizing temperature-economic relationship in projections and IAMs (Integrated Assessment Models).