Changes in the Distribution of Earnings Volatility
研究发现美国自1970年以来个人劳动收入波动性的平均上升几乎完全由少数高波动人群驱动，而绝大多数人的波动性并未系统性增加。
Recent research has documented a rise in the volatility of individual labor earnings in the United States since 1970. Existing measures of this trend abstract from within-group latent heterogeneity, effectively estimating an increase in average volatility for observable groups. We decompose this average and find no systematic rise in volatility for the vast majority of individuals. Increasing average volatility has been driven almost entirely by rising earnings volatility of those with the most volatile earnings, identified ex ante by large past earnings changes. We characterize dynamics of the volatility distribution with a nonparametric Bayesian stochastic volatility model from Jensen and Shore (2011).