Estimating the return to college in Britain using regression and propensity score matching

College graduates tend to earn more than non-graduates but it is difficult to ascertain how much of this empirical association between wages and college degree is due to the causal effect of a college degree and how much is due to unobserved factors that influence both wages and education (e.g. ability). In this paper, I use the 1970 British Cohort Study to examine the college premium for people who have a similar ability level by using a restricted sample of people who are all college eligible but some never attend. Compared to using the full sample, restricting the sample to college-eligible reduces the return to college significantly using both regression and propensity score matching (PSM) estimates. The finding suggests the importance of comparing individuals of similar ability levels when estimating the return to college. -- return to college ; regression ; propensity score matching