Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
I develop a method for measuring a binary treatment's effects when more than one control group is available but each control group is imperfect. To study effects of the minimum wage, for example, I compare border counties where the minimum wage increases to the set of neighboring counties, the set of neighbor‐of‐neighboring counties, and so forth. The key innovation is to model the ratio of the biases present in these comparisons. Crucially, the model contains a testable implication. The method can address spillovers and, in another application, is used to measure treatment effects away from the discontinuity in a regression discontinuity design.