Cross-Sectoral Variation in the Volatility of Plant-Level Idiosyncratic Shocks

Abstract

We estimate the volatility of plant-level idiosyncratic shocks in the U.S. manufacturing sector. Our measure of volatility is the variation in Revenue Total Factor Productivity which is not explained by either industry- or economy-wide factors, or by establishments' characteristics. Consistent with previous studies, we find that idiosyncratic shocks are much larger than aggregate random disturbances, accounting for about 80% of the overall uncertainty faced by plants. The extent of cross-sectoral variation in the volatility of shocks is remarkable. Plants in the most volatile sector are subject to about six times as much idiosyncratic uncertainty as plants in the least volatile. We provide evidence suggesting that idiosyncratic risk is higher in industries where the extent of creative destruction is likely to be greater.