Is selection on observables informative about selection on unobservables? The correct answer is probably “No.”

This is a thesis on whether selection on observables is informative about selection on unobservables. By creating my own adaption to the paper of Black et al. (2015), I create a method by which one can investigate whether selection on observables is informative about selection on unobservables. Although this is pioneering research, I am able to provide some suggestive evidence that selection on observables is not informative about selection on unobservables. As a ‘by-product’ I found some evidence that it is sometimes better not to control for observables than to control for these variables.