Card is best known amongst intellectuals for his minimum wage work, but he also has been central in estimating the returns to higher education, using superior methods. In particular, he has induced many economists to downgrade the import of the signaling model of education. Here is one excerpt from his Econometrica paper, appropriately entitled “Estimating the Return to Schooling: Progress on Some Persistent Econometric Problems:
A review of studies that have used compulsory schooling laws, differences in the accessibility of schools, and similar features as instrumental variables for completed education, reveals that the resulting estimates of the return to schooling are typically as
big or bigger than the corresponding ordinary least squares estimates. One interpretation of this finding is that marginal returns to education among the low-education subgroups typically affected by supply-side innovations tend to be relatively high, reflecting their high marginal costs of schooling, rather than low ability that limits their return to education.
The empirical problem arises of course because intrinsic talent and degree of schooling are highly correlated, so the investigator needs some recourse to superior identification. How can you tell if apparent returns to schooling simply reflect a higher talented cohort in the first place? So you might for instance look for an exogenous change to compulsory schooling laws that affects some children but not others (a few of those have come in the Nordic countries). That likely will be uncorrelated with child talent, and so it will help you separate out the true causal return to additional schooling, because you can measure whether the kids with that extra year end up earning more, controlling for other relevant variables of course. And see Alex’s discussion of the Angrist and Card paper on similar questions.
See also Card’s survey of this entire field, written for Handbook of Labor Economics. One impressive feature of these pieces is they show how many disparate methods of measurement all point toward a broadly common conclusion. Whether or not you agree, these papers have been extremely influential, and they are one reason why Claudia Goldin, in my recent CWT with her, asserted that very little of higher education was about the signaling premium.