Climate Change and Measures of Economic Growth:Solving the Spatial Mismatch Problem

Many studies have been undertaken to quantify the economic costs of climate change. However, while climate data is measured at a grid cell level, economic data are measured at the national level. In order to form correct damage estimates researchers must reconcile the two. Aggregating climate data up to the national level has been the more common approach but results are sensitive to how the averaging is done and the averaging process itself can bias the results. An alternative approach has been to project economic data down to the grid cell level. Nordhaus (2006) developed the G-Econ database to do this, but while it provides considerable spatial detail it provides only four quinquennial observations per cell from 1990 to 2005. We develop herein a model to predict within-grid cell economic activity using national, regional and local economic activity. The latter is measured using a unique dataset showing annual fight volumes at hundreds of urban and rural airports worldwide from 1976 to 2010. We show that the model has a high level of explanatory power and can be used in an iterative algorithm to infill and extrapolate the G-Econ data base to provide annual observations for grid cells in approximately 150 countries over the 1976 to 2010 interval. We supplement this with satellite nightlight data which provides even more spatial detail but over a shorter time frame.