Knowledge Specialization in Ph.D. Student Groups
Researchers have argued that specialization within groups yields productivity gains. We evaluate this statement with a focus on groups of Ph.D. students. Using an established technique in computer science called Latent Dirichlet Allocation, we construct a novel measure of the dispersion of Ph.D. students' research interests based on their dissertation abstracts. We then relate this measure to Ph.D. group publications. For our study, we use a rich dataset on groups of Ph.D. students who studied at a major Swiss University, during the 1993-2008 period. We find robust evidence that within-group knowledge specialization is associated with a larger number of publications. However, when specialization increases beyond a critical level, it hinders the group's publication output. We interpret these results as an indication that gains, in the amount of research output, can be achieved if Ph.D. students specialize according to their comparative advantages. However, beyond a certain level, knowledge specialization has a detrimental impact on research output, due to increasing communication costs and an increased likelihood of conflict insurgence.