| This dissertation consists of three empirical analyses that measure the effects of climate on agriculture, the U.S. economy, and the incidence of conflict around the world.;In the first paper, we measure the economic effects of climate change on U.S. agriculture. Using a panel of over 2100 U.S. counties that includes climate, weather, economic, and geophysical data, we estimate the effects of climate on farmers' profit-maximizing crop choice decisions, conditional on state and year fixed effects. Our estimates show a larger than previously estimated gain of ;Previous research on the economic effects of climate changes focuses, almost exclusively, on agriculture. However, a purely agricultural focus only gives us a partial idea of the effects of climate on the economy. In the second paper, we measure the effects of climate change on economic activity at the county level conditional on industry price indexes, on a county wage index, on population, and on county and year fixed effects. We use data from 1978 to 1997 for over 3,000 counties. We find that a uniform increase in temperature of 5°F (2.78°C) and an 8-percent increase in precipitation decreases county level economic activity by 7.3 percent, with most of this change corresponding to the increase in temperature. We also find that counties in colder areas may benefit from a warmer climate while already warm areas will suffer most of the losses.;Explaining the incidence of conflict is notoriously difficult because of endogeneity problems. And, while the literature has linked climate and conflict, it has not been able to identify the mechanisms that underlie that link. In the third paper, we use different measures of climate as instrumental variables for factors commonly thought to explain conflict, thus addressing both of these problems. We use country-level data from 1969 to 1999 to capture the causes of conflict. Our reduced form estimates indicate that a 10 mm increase in precipitation decreases conflict by 7 percentage points and that a 1°C increase in temperature increases it by 16 percentage points. Estimates from our full model indicate that economic explanations are the main drivers of conflict, but that democracy also plays a role in small grievances. We also find that after we account for climate, weather provides little explanatory power. This calls into question the literature's focus on short-term weather shocks. |
