The Evolution of Labor Earnings Risk in the U.S. Economy∗

A large body of literature has documented an increase in wage inequality over the 1970’s and 1980’s (see, for example, Levy and Murnane, 1992). This increase in wage inequality has occurred both within and between education-experience groups. While the former unfolds during the 1970’s, the latter is experienced over the 1980’s. Recent studies show that these trends are weakened in the 1990’s. The analysis of Juhn, Murphy, and Pierce (1993) links the increase in wage inequality to technological progress. As technology advances the demand for ability grows at a faster pace than its supply causing skill prices to increase and the wage gap between skilled and unskilled workers to widen. To arrive to this conclusion they map quantiles of the distribution of the residuals in log wage equations to quantiles of the distribution of ability. This is possible provided the following four conditions hold: 1) there is one, and only one, type of unobserved ability; 2) the distribution of this ability across individuals is invariant over time; 3) there are no unantecipated shocks in earnings so that the economy is described by a deterministic environment; 4) wages are measured without error. Regarding the first condition, Heckman and Rubinstein (2001), Carneiro and Heckman (2003), and Heckman, Stixrud, and Urzua (2006) have shown that abilities are multiple in nature. These abilities may be combined in different quantities to produce the same outcome. For example, an individual with low ∗This research was supported by NIH R01-HD043411. Cunha is grateful to sponsorship from CAPES 1430/99-8 and the Claudio Haddad Dissertation Fund at the University of Chicago. †University of Chicago, Department of Economics. ‡University of Chicago, Department of Economics and the American Bar Foundation.