This project examines the implications of the growth of short term earnings inequality for inequality of lifetime earnings. It has been well documented that earnings inequality grew substantially in the U.S. during the '70s and '80s. This is true in the aggregate, as well as between certain groups of workers (e.g., college vs. high school graduates). Furthermore, inequality also grew even within narrowly defined education/age/race categories. However, most the research on inequality has examined the behavior of repeated cross-sections of earnings distributions over several years. Thus, what has been well documented is that inequality has increased in the short term (that is, when one looks at the earnings distribution at a point in time). Much less is known about what, if anything, has happened to the distribution of present values of lifetime earnings. The observed growth in inequality in the short term may be consistent with little or no growth in inequality over the life cycle, depending on the extent of earnings mobility over time. This issue can only be addressed using panel data on multiple cohorts. The project will use data from the 1969-1996 waves of the Panel Study of Income Dynamics (PSID) to examine how the stochastic processes for earnings of men and income women have shifted over this period. The key point of the study is that one must start with a good model of earnings (or income) mobility in order to accurately assess how changes in the distribution of earnings in the cross-section translate into inequality in the present value of earnings over the life cycle. Yet existing models of earnings mobility that assume normality of shocks to the earnings process do a poor job of fitting observed mobility patterns (specifically, they tend to overestimate the amount of mobility in earnings). In the investigators' prior work they have developed a mixture of normals model that does a much better job of fitting observed mobility patterns. It is a generalization of this type of model that will be used to assess impact of the growth of short-term earnings inequality for inequality of lifetime earnings. The results will have important implications for questions related to poverty. The poverty rate in the U.S. increased between 1973 and 1992. But the poverty rate is defined as income below the poverty line in a single year. Thus, if earnings mobility is substantial, the increase in the poverty rate may not represent an increase in the fraction of the population with economic resources too low to achieve a basic standard of living. This investigation will cast light on this question by assessing whether the fraction of the population with a below poverty line level of permanent income (i.e., the level of annual consumption they can maintain given their life cycle income path) has changed.
