Results reported in this paper indicate that the returns-to-schooling
profile exhibits a peculiar departure from log-linearity. In
particular, for white males, the return to the 15th year of schooling
(the third year of college) seems to be much smaller than that predicted
by log-linearity. That is, people with 15 years of schooling do not
appear to earn any more than those with 14 years of schooling. I
discuss several theories and evidence that may explain the peculiarity.
Amongst them, measurement error in schooling seems the most plausible.

This study uses direct information on years of schooling and degree attainment taken from a special
Current Population Survey (CPS) to test for sheepskin effects. I find significant earnings gains of
approximately 9, 11, and 21 percent for achieving a high school diploma, an associate’s degree, and
a bachelor’s degree, respectively. I also use the data to impute a traditional measure of years of
schooling (corresponding to the education information on the CPS prior to 1992) from the new
CPS question on degrees. The imputation is found to be relatively accurate and to lead to similar
estimates of the return to schooling.