We measure the welfare gain from removing aggregate consumption fluctuations in a model where each individual faces incomplete consumption insurance. We show that, because this welfare gain is a convex function of the overall consumption risk—aggregate plus idiosyncratic—each individual faces, to gauge the magnitude of the gain, it is important to match individuals' overall risk prior to any policy. In an economy calibrated to match individuals' overall risk, even removing 10 percent of aggregate fluctuations can result in a large welfare gain. Further, large gains do not necessarily depend on the countercyclical nature of idiosyncratic risk. (JEL E21, E32)
On measuring the welfare cost of business cycles
