Trends in Earnings Volatility among US Men.

• Author: Moffitt, Robert A.

Economists have been concerned about the volatility of earnings and income for decades because it creates uncertainty for families and individuals and makes it more difficult for them to plan future consumption. Volatility may discourage investments financed by loans that have to be paid off by a future income stream. Education and training, which can be very costly at the college level, may be forestalled due to uncertainty about future earnings payoffs.

Sociologists and economists alike have shown that family income volatility has harmful effects on children and their development, especially younger children. Economists recognize that neither private insurance markets nor government programs like unemployment insurance can adequately protect individuals against most earnings risk. Consequently, those with sufficient income to forgo current consumption often attempt to self-insure by engaging in precautionary saving, but this is rarely enough to smooth future consumption in the face of significant volatility.

Whether earnings volatility has risen over time in the United States is an important question for economics and for government policy. One wellknown development that may have led to such an increase is deindustrialization, which has reduced the number of stable, long-term blue-collar jobs and replaced them with jobs in the service sector, retailing, and other industries that often have high rates of turnover and unstable earnings. Some growing industries, like high tech, have several dominant firms and many smaller firms with high failure rates and intense competition, leading to unstable employment and earnings profiles for many individuals in those industries. But against these well-recognized forces has been a reduction in volatility at the macroeconomic level, commonly called the "Great Moderation." That term was used to describe the reduction in macro-level volatility that began in the 1980s and ran through 2007. (1)

I have studied trends in the earnings volatility of US men for many years, beginning with my 1994 paper in the Brookings Papers on Economic Activity, coauthored with Peter Gottschalk. (2) The focus of this initial work was on men because their jobs are more concentrated in manufacturing and other industries hit particularly hard by deindustrialization. We took data from a well-known household survey, the Panel Study of Income Dynamics (PSID), which is the longest-running panel household survey in the world. It started in 1968 and is ongoing. We used the data to track White men's earnings from 1970 to 1987--focusing on White men because of small sample sizes for other groups--and used very simple techniques to see if their earnings had become more unstable. We found that earnings volatility, measured as the standard deviation of the change in earnings from one survey wave to the next, rose dramatically over that period, particularly among less-educated workers. Volatility rose by 50 percent for high school graduates and 96 percent for high school dropouts. The disproportionate increase among those less educated suggested a potential role for deindustrialization. We also noted that the increase in volatility was partly responsible for the growth in cross-sectional earnings inequality because higher levels of earnings volatility and "transitory shocks" in a given year increase the dispersion of earnings.

Work since the Gottschalk-Moffitt Study

Gottschalk and I continued to update our work periodically with data from...