Did victories in certification contests affect the survival of organizations in the American automobile industry during 1895-1912? A replication study

Research summary: We successfully replicate the highly influential study: The social construction of reputation...,(Rao, 1994) which reports that cumulative victories in certification contests are negatively associated with firm failure. The replication is robust to the inclusion of additional controls. As in the original, tests of whether the theory is most powerful under higher uncertainty are not supported. Further, placing second, third, or merely participating in races also negatively predicts firm failure, and there is insufficient information in the data to tease out the importance of these predictors versus race victories. We discuss the assumptions under which the evidence can be interpreted as supportive of a more general argument of loose coupling, where affiliation with certification contests reduces firm failure. Managerial summary: We successfully replicate a study that related victories in races to the survival of early automobile firms. This result was interpreted as evidence that rank-order certification contests legitimized firms and led to survival. We then demonstrate that there is insufficient information to tease out the relative importance of victories, as opposed to placing second, third, or merely participating in predicting survival. Our result is consistent with an argument that affiliation with certification contests, not only winning them, increases a firm's chances of survival. It is also consistent with an argument that firms with better quality automobiles won races and survived. An implication of our work is that there is insufficient evidence to determine if firms in new industries should expend finite resources on participation in certification contests or improvement of product quality.