ABSTRACT
In previous two chapters, I examined the effect of college quality on graduates’ earnings, on average and by different groups of students. The analyses suggested that graduating from high-quality colleges had positive and significant effects on earnings although substantial variations existed across different groups of students. These positive effects, by the design of econometric models used, were all evaluated at the mean of the earnings distribution (of the whole sample or of specific sub-samples). Although the finding of a positive average effect of college quality on earnings is convincing and important, it is plausible that the effect of college quality differs across graduates’ final positions in the earnings distribution. Put in a slightly different way, the predictive power of college quality may be different for students at the top of the earnings distribution than it is for students at the bottom. Intuitively, if one ends up in a low-paid job, a degree from a high-quality college might not help much. On the other hand, if one ends up in a high-paid job, a degree from a high-quality college should matter much more.
