In this paper we set out a simple model of optimal schooling investments that
emphasizes the interaction between schooling choices and income determination; and
estimate it using a fresh sample of data on over 700 identical twins. According to the model,
equally able individuals from the same family should attain the same observed schooling
levels, apart from random errors of optimization or measurement. A variety of direct and
indirect tests provides no evidence against this hypothesis.
We estimate an average return to schooling of 10% for genetically identical
individuals, but estimated returns are slightly higher for less able individuals. Unlike the
results in Ashenfelter and Krueger (1994), which were based on a much smaller sample, we
estimate that schooling is positively correlated with ability level, so that simple cross-section
estimates are slightly upward biased. Taken together these empirical results imply that
more able individuals attain more schooling because they face lower marginal costs of
schooling, not because they obtain higher marginal beneﬁts. The results stand in sharp
contrast to recent claims that genetic factors predetermine education and income, and that
such differences are not amenable to alteration by public or private choices.
This paper uses a new survey to contrast the wages of genetically
identical twins with different schooling levels. Multiple measurements of
schooling levels were also collected to assess the effect of reporting
error on the estimated economic returns to schooling. The data indicate
that omitted ability variables do not bias the estimated return to
schooling upward, but that measurement error does bias it downward.
Adjustment for measurement error indicates that an additional year of
schooling increases wages by 12-l6t, a higher estimate of the economic
returns to schooling than has been previously found.
Description of files: TWINS90.DAT This is an ASCII file that contains a public use file with selected variables from the 1990 twins data Ashenfelter and Krueger used. The variables are separated by columns. The input statement in PUBLIC.SAS lists the proper order of the variables; this can easily be adapted if you want to use a program other than SAS. PUBLIC.SAS This is a mainframe SAS program that reads in the data from TWINS90.DAT. This program labels the variables. It also runs regressions from selected Tables in Ashenfelter and Krueger (1993). PUBLIC.LOG This contains the SAS listing file of PUBLIC SAS. TWIN2.SAS This is a SAS program that created the data in TWINS90.DAT.