Abstract

Abstract: The economic impact of global warming varies across firms because of differences in climate, technology, and adaptive capacity. Aggregate estimates of the average effect of warming are thus insufficient to model climate change vulnerability in developing countries. In this study, I measure the distributional effect of climate change in Brazilian agriculture by estimating the quantile and interquantile regressions of land value on climate, using agricultural census data for 490,000 commercial farms. The effect of a 1°C rise in average temperature on land values ranges from -5% for the most productive farmers located in the colder South region to -34% for the least productive farmers located in the warmer North region. The impact is most severe in the extreme 0.01 quantile of the land value distribution. The productivity inequality between farms in the extremes of the distribution of land values may double with marginal warming.