Why Do Some Firms Make Seemingly Inferior Decisions?

Empirical studies on corporate decisions often employ the treatment model, the Heckman selectivity model, or the matched sample method and find that one choice (treatment) is strictly better than the other. These results lead to a puzzle - why do some firms select the inferior choices? In this paper, I show that methodological issues drive such findings because these methods rule out the possibility that firms with heterogeneous characteristics can make different optimal decisions. The objects of this paper are to discuss how such an optimal decision making process is fundamentally incompatible with the assumptions required by the above methods and to caution the applications of those methods in studying corporate decisions