Abstract:
In the four months following the trough of the Great Depression in March 1933, industrial production rose 57 percent. We argue that an important channel aiding recovery came through the direct effect of devaluation on farm prices, incomes, and consumption. We call this the farm channel. Using U.S. and British crop price data, we document that devaluation raised prices of traded crops and their close substitutes (other grains). And using state and county auto sales and income data, we document that recovery proceeded much more rapidly in farm areas. Our baseline estimates imply that a one standard deviation increase in the share of a state’s population living on farms is associated with a 20–34 percentage point increase in auto sales growth from winter to fall 1933. This effect is concentrated in states producing traded crops (cotton, tobacco, wheat) or close substitutes, suggesting an important role for devaluation. In annual county data we show that the farm channel is strongest in counties with more indebted farmers. To map these cross-sectional estimates into an aggregate effect, we build an incomplete-markets model that explicitly incorporates both the benefits of the farm channel to farmers (higher farm income) as well as the costs to nonfarmers (higher prices paid for farm goods). The model suggests that by redistributing income to indebted farmers with a high marginal propensity to consume, the farm channel may explain 25-50% of the spring 1933 recovery.