Next, I test whether my results are robust to different
parametrization of f(sit, cit, τ); results presented in Table 4.
Jacobson et al. (2005) estimate models that allow for both a
discrete impact of any college enrollment and gains that are
proportional to the number of credits earned. Their results
suggest that “just showing up” in college increases women’s
earnings by 5.5% , which could be attributed to gains from
networking or nonacademic services provided by community
colleges. In contrast, when I estimate models that allow for a discrete impact of community college attendance, I
find evidence of a statistically significant negative impact on
earnings equal to $181 per quarter and a negative but statistically
insignificant impact on the probability of employment
(Table 4, columns 1 and 4).