The Welfare Costs of Macroeconomic Fluctuations under Incomplete Markets: Evidence from State-Level Consumption Data

Existing estimates of the welfare cost of business cycles suggest that it is quite low and might well be minuscule. Many of these estimates are based on aggregated U.S. consumption data. Arguably, because markets are incomplete and risk-sharing is imperfect, the welfare costs computed with aggregate consumption data are likely underestimates. Yet, incomplete-market models have not yielded significantly greater cost figures. Previous incomplete-market studies, however, have relied on model-generated consumption series that reflect optimal decsions in models calibrated using individual income data. In this paper, we maintain the assumption of incomplete markets but use observed consumption streams instead. Using state-level retail sales figures, we show that the welfare cost of macroeconomic volatility is in fact very substantial. In one half of the U.S. states, the welfare gain from the removal of business cycles can in fact exceed the gain from receiving an extra percentage point of consumption growth in perpetuity. In short, our results indicate that macroeconomic volatility has first-order welfare implications.