Learning When to Quit

Résumé

We study a dynamic model of the decision to continue or abandon a research project. Researchers improve their ideas over time, and also learn whether those ideas will be adopted. Projects are abandoned as researchers grow more pessimistic about their chance of acceptance. We estimate the structural parameters of this dynamic optimization problem using a novel data set that contains information on both successful and abandoned projects submitted to the Internet Engineering Task Force (IETF), an organization that creates and maintains standards necessary for the functioning of the internet. Our empirical results show that feedback increases the rate of learning, and we use the model to simulate the costs of researcher over-confidence. (with Bernhard Ganglmair and Emanuele Tarantino).