Volatility, Insurance, and the Gains from Trade∗

By reducing the negative correlation between local prices and productivity shocks, trade liberalization changes the volatility of returns. In the context of agriculture in developing countries—where production is risky, producers are risk averse, and insurance markets are incomplete—this change in volatility may have large welfare effects. In this paper, we empirically, analytically, and quantitatively characterize the second moment effects of trade. Using forty years of agricultural micro-data from India, we show empirically that trade increased farmer’s revenue volatility, causing farmers to shift production toward less risky crops. We then incorporate producers’ optimal allocation of resources across risky production technologies into a general equilibrium trade model to show analytically that (1) despite the possibility that trade increases the volatility of real returns, trade is always welfare improving; and (2) when trade is costly, offering farmers insurance may make them worse off. Finally, we structurally estimate the model – recovering farmers’ unobserved risk-return preferences from the gradient of the mean-variance frontier at their observed crop choice – to quantify the second moment welfare effects of trade. We find that the gains from trade would have been roughly 10 percent larger had trade not also increased volatility, but had farmers had access to perfect insurance, the gains from trade would have been smaller and the farmers would have been strictly worse off. ∗We thank Kyle Bagwell, Dave Donaldson, Pablo Fajgelbaum, Rocco Machiavello and Kiminori Matsuyama and seminar participants at George Washington University, Stanford University, and Purdue University. All errors are our own. Volatility and the Gains from Trade 1