The data for this study come from the 1979 National Longitudinal Survey of
Youth (NLSY79), an ongoing study of 12,686 young individuals who were aged 14 to 21
as of January 1, 1979. The NLSY79 includes information on college attendance, type of
college (two-year versus four-year college) as well as personal characteristics such as age,
sex, race, number of siblings, Armed Forces Qualifying Test (AFQT) score, and
information on the education levels and occupations of parents.
We apply some restrictions to the NLSY79 data set to construct the sample for
our analysis. The sample in this study covers only those who were less than eighteen
years old as of January 1, 1979 and excludes the military sub-sample. We also drop any
observations that contain inconsistent responses over the survey years. Data on average
in-state tuition levels for public four-year institutions for 1986-87 was merged to the
individual NLSY79 data based on state of residence in 1979. This tuition data is from
approximately the midpoint of the observation period.6 Although individuals may attend
out of state universities or private universities, the in-state tuition level could be thought
of as the lower bound on the cost of college enrollment.
Since the accuracy of the schooling information is critical for our analysis, it is
essential to have an accurate dating of educational events so that waiting duration to first
6Source of state tuition data is State Comparisons of Education Statistics: 1969-70 to 1996-97, National
Center for Education Statistics.
13