respectively, when conditioning on entry into a two-year institution. From the hazard
estimates in Figures 2 and 3, one can show that the estimated probability of obtaining a
four-year degree, conditional on first entering a four-year institution, declines from 0.43
to 0.34 (or by 21 percent) when an individual delays college enrollment by one year after
graduating high school. Conditional on first entering a two-year institution the
probability of attaining a four-year degree drops from 0.24 to 0.17 (or by 29 percent )
when an individual delays college enrollment by one year.
The estimates in Table 3 also show that a higher Armed Forces Qualifying Test
(AFQT) score significantly increases the probability of college enrollment and the chances
of graduation while significantly reduces the risk of dropout. Female high school
graduates have significantly higher college enrollment hazards than males. African
American and Hispanic high school graduates are significantly more likely to enroll in
college and significantly less likely to drop out than their white counterparts. However,
the estimated graduation hazard rates are significantly lower for Hispanics. Others (see
Cameron and Heckman (2000) and National Center for Education Statistics(2001)) have
also found that once educational achievement is controlled for using test scores like
AFQT, minorities are more likely attend college than whites.12 Kane and Spizman (1994)
12 This does not mean that the actual enrollment rate of minorities is higher than that of white youths.
Since black and Hispanic youths have lower levels of AFQT scores, parental income and parental
education, overall minorities are less likely to attend college than white students.
18