The earnings of 200 million U.S. workers change each year for various reasons. Some of these changes are anticipated while others are more unexpected. Although many of these changes may be due to pleasant surprises—such as receiving salary raises and promotions—others involve disappointments—such as falling into unemployment. Arguably, some of these factors have rather short-lived effects on an individual’s earnings, whereas others may have permanent effects. Many labor economists have been interested in these various shocks to earnings. How big are the more permanent shocks to earnings? How large are they relative to those that are temporary in nature? What are the sources of these shocks? In this blog post, we exploit a novel data set that enables us to explore the properties of earnings shocks: their magnitudes as well as their origins.