It might seem that nowadays we're already drowning in too much data and that devoting more energy to interpreting it and less to collecting even more of it would be advisable. On the other hand, large amounts of data can offer the opportunity of looking at old questions in new ways.
Nordhaus describes the construction of a geographically scaled economic data set (G-Econ) that transforms the economic quantity gross regional product (where a region can be a nation, as in gross national product, or a smaller political subdivision) along geophysical dimensions, such as temperature or coastal proximity. Aggregating economic data across multiple sources and scaling output to a cell size of 1° longitude by 1° latitude yields the gross cell product or GCP. The established finding that output per person increases with distance from the equator converts into a decrease in output per area as mean temperature decreases, with a decline of 105 from the maximum at about 10°C to the polar regions. Further analysis reveals that country-specific effects, such as institutional differences, account for only one-third of this variation, with geography contributing to but not explaining all of the rest. -- GJC
Proc. Natl. Acad. Sci. U.S.A. 103, 10.1073/pnas.0509842103 (2006).
