GC079-09
The Long-term Effect of Temperature and its Variability on Economic Activity - New Evidence from Spatial Econometric Analysis of High-resolution Nightlights Data

Friday, 11 December 2020: 19:25
Virtual
Manuel Linsenmeier, London School of Economics, Geography and Environment, London, United Kingdom; Grantham Research Institute on Climate Change and the Environment, London, United Kingdom
Abstract:
While there has been substantial progress in understanding the effects of annual weather fluctuations on economies, the long-run influence of climate has been less intensely studied. This is especially true for climate variability. In this paper, we study the causal effect of temperature and its seasonal and interannual variability on economic development. To this aim, we proxy economic activity by high-resolution satellite data of lights at night and combine it with reanalysis weather data as well as several additional geographic datasets. For identification of causal effects we use the recently developed econometric framework of spatial first-differences. We embed our empirical analysis in a simple economic framework to illustrate the effect of temperature variability on economic activity for different assumptions about the effect of temperature levels. Our results indicate a non-linear effect of annual mean temperature on economic activity with an optimal temperature around 17 degrees Celsius. In agreement with our theoretical framework we find a significantly negative effect of seasonal variability on economic activity at all temperature levels. The effect of interannual variability is positive at low temperature levels and negative at high temperatures. Our results resolve some of the discrepancies between previous studies. Specifically, our results suggest that the effect of temperature on economic activity is linear around the global optimum but of higher order at low and high temperatures. Our results show little signs of adaptation to either seasonal temperature fluctuations or average annual mean temperature levels. Overall, our results draw attention to the effect of climate variability, especially seasonal variability, which is projected to change by climate models but has so far been overlooked in assessments of the economic impacts of climate change.