No abstract

The ultimate impact of climate change on human systems will depend on the natural resilience of ecosystems on which societies rely as well as on adaptation measures taken by agents, individually and collectively. No sector of the economy is more reliant on climate than agriculture. Evidence from the American settlement process suggests that societies can successfully adapt to new climatic environments. Whether and how much agriculture will manage to adapt to a changing climate remains an open question in the empirical economics literature, however. This article reviews the existing evidence on weather and/or climate impacts on agricultural outcomes from the economics literature, with a focus on methodological questions. Some key econometric issues associated with climate impact measurement are discussed. We also outline important questions that have not been adequately addressed and suggest directions for future research.

China is one of the most heavily polluted nations and is also the largest agricultural producer. There are relatively few studies measuring the effects of pollution on crop yields in China, and most are based on experiments or simulation methods. We use observational data to study the impact of increased air pollution (surface ozone) on rice yields in Southeast China. We examine nonlinearities in the relationship between rice yields and ozone concentrations and find that an additional day with a maximum ozone concentration greater than 120 ppb is associated with a yield loss of 1.12% ± 0.83% relative to a day with maximum ozone concentration less than 60 ppb. We find that increases in mean ozone concentrations, SUM60, and AOT40 during panicle formation are associated with statistically significant yield losses, whereas such increases before and after panicle formation are not. We conclude that heightened surface ozone levels will potentially lead to reductions in rice yields that are large enough to have implications for the global rice market.

Recent work on nonparametric identification of average partial effects (APEs) from panel data require restrictions on individual or time heterogeneity. Identifying assumptions under the "generalized first-differencing" category, such as time homogeneity (Chernozhukov, Fernandez-Val, Hahn, and Newey, 2013), have testable equality restrictions on the distribution of the outcome variable. This paper proposes specification tests based on these restrictions. The bootstrap critical values for the resulting Kolmogorov-Smirnov and Cramer-von-Mises statistics are shown to be asymptotically valid and deliver good finite-sample properties in Monte Carlo simulations. An empirical application illustrates the merits of testing nonparametric identification from an empiricist's perspective.

Abstract-We consider constructing probability forecasts from a parametric binary choice model under a large family of loss functions ("scoring rules"). Scoring rules are weighted averages over the utilities that heterogeneous decision makers derive from a publicly announced forecast (Schervish, 1989). Using analytical and numerical examples, we illustrate how different scoring rules yield asymptotically identical results if the model is correctly specified. Under misspecification, the choice of scoring rule may be inconsequential under restrictive symmetry conditions on the data-generating process. If these conditions are violated, typically the choice of a scoring rule favors some decision makers over others.