Key points are not available for this paper at this time.
Using U.S. panel data on adult males, I compare the "profile heterogeneity model" of earnings dynamics, in which the earnings/experience profile varies across individuals, to a competing model in which earnings "has a unit root." The latter specification enjoys increasing popularity among researchers. My analysis questions this favor, suggesting the profile heterogeneity model provides a more consistent representation of the data. I also provide new estimates of the variation in earnings growth rates. Previous evidence is from relatively unrepresentative samples. Individuals one standard deviation above the mean enjoy a 20%-30% earnings advantage in just 10 years.
Michael Baker (Tue,) studied this question.