NBER WORKING PAPERS SERIES
ESTIMATING THE PAYOFF TO SCHOOLING USING
THE VIETNAM-ERA DRAFT LOTTERY
Joshua D. Angrist
Alan B. Krueger
Working Paper No. 4067
NATIONAL BUREAU OF ECONOMIC RESEARCH
1050 Massachusetts Avenue
Cambridge, MA 02138
May 1992
We are grateful to Kevin McCormick, Ronald Tucker, and Greg Weyland for
creating a special Current Population Survey extract for us. We are also
grateful to Dean Hyslop for excellent research assistance and to Michael
Boozer for helpful comments. The data used in this paper will be made
available to other researchers. This research was supported by the National
Science Foundation, Grant No. SES-9012149. This paper is part of NBER’s
research program in Labor Studies. Any opinions expressed are those of the
authors and not those of the National Bureau of Economic Research.NBER Working Paper #4067
May 1992
ESTIMATING THE PAYOFF TO SCHOOLING USING
THE VIETNAM-ERA DRAFT LOTTERY
ABSTRACT
Between 1970 and 1973 priority for military service was randomly
assigned to draft-age men in a series of lotteries. Many men who were at risk
of being drafted managed to avoid military service by enrolling in school and
obtaining an educational deferment. This paper uses the draft lottery as a
natural experiment to estimate the return to education and the veteran premium.
Estimates are based on special extracts of the Current Population Survey for
1979 and 1981-85. The results suggest that an extra year of schooling acquired
in response to the lottery is associated with 6.6 percent higher weekly earnings.
This figure is about 10 percent higher than the OLS estimate of the return to
education in this sample, which suggests there is omitted-variable bias in
conventional estimates of the return to education. Our findings are robust to
a variety of assumptions about the effect of veteran status on earnings.
Joshua D. Angrist Alan B. Krueger
Economics Department Economics Department
Hebrew University Princeton University
Jerusalem 91905 Princeton, NJ 08544
Israel and NBER
and NBERFor many years economists have sought to estimate the monetary
retum to education without bias from omitted variables that are correlated with
educational attainment and with earnings capacity (see Griliches (1977) for a
survey of this literature). A positive bias would arise, for example, if
individuals with higher earnings capacity obtain more schooling. A variety of
econometric techniques have been applied to try to overcome this problem, but
the magnitude and importance of omitted-variable bias in the estimated return
to education is still an unanswered question. Willis (1986; p. 589), for
example, concludes that, "Given the complexity of the issues and the non-
representative character of the data sets that have been employed in the
literature on ability bias, it is difficult to reach any firm conclusion about the
magnitude or even the direction of the bias..."
Concern over potential ability bias in the return to education has been
heightened in recent years by evidence that the return to education has
increased in the 1980's. Of course, if individuals’ schooling levels were
randomly assigned, the payoff to education could be consistently measured
simply by comparing the earnings of more and less educated individuals. In
the absence of random assignment, ability bias can be overcome by using
instrumental variables that are correlated with education but have no other
effect on earnings. This paper reports our analysis of a natural experiment
that generates such instruments. The natural experiment stems from the
Vietnam-era draft lottery, which was used to determine priority for induction
into the U.S. military between 1970 and 1973.
As a consequence of the draft lottery, the risk of induction into the
armed services was randomly assigned on the basis of individuals' birthdays.
1 See, for example, Blackburn and Neumark (1991). For convenience, we
follow the practice of the previous literature and label omitted variables in
wage equations that are correlated with education as ability. We note that
other omitted factors, such as family wealth, might influence educational
achievement and earnings.