Network exploration and exploitation: Professional network churn and scientific production

The production of scientific knowledge is an inherently social process making professional networks important for producing science outcomes. Although prior work has demonstrated the connection between collaboration and productivity, most research that examines scientist networks begins from the perspective that structure predicts productivity. Institutional approaches to explaining productivity are useful, but generally ignore the role of individual agency or strategic network behavior. Our study utilizes the dynamic perspective of network churn to assess how professional network composition and structure change overtime via processes of network exploration and exploitation. Using two waves of survey data from a national sample of academic scientists and engineers across six disciplines in the United States, we investigate how network churn affects the quantity and quality of scientific production. Our results suggest that while network exploration generally improves production quality, it can hurt quantity. Network exploitation tends to have the opposite effect, resulting in short term gains but potentially limiting the innovativeness of future research. By recognizing the tradeoffs associated with alternative networking strategies, policy makers in universities and other research organizations can begin focusing on interventions that more effectively target scientists’ strategic network behavior.