Fig. 2 previews my main approach and results. Panel A of Fig. 2 displays unadjusted average earnings by quarter,
both before and following welfare entry. The black solid line indicates individuals who attend college following entry
into Colorado Works but do not earn a credential following entry (drop-outs), the thick dashed line represents the earnings
trajectories of degree recipients, and the gray line represents the earnings of women who do not attend college prior
to or following welfare entry. I exclude women who were already enrolled in college at entry (described in the fourth column
of Table 1). Thin dotted lines represent 95% confidence intervals.