Following the Code: Spillovers and Knowledge Transfer (with Peter Naftaliev and Uriel Stettner)

Abstract:
It is believed that there are significant knowledge spillovers in Open Source Software (OSS). If such spillovers exist, it is likely they occur via two channels: In the first channel, programmers take knowledge know-how, and experience gained from one OSS project they work on and employ it in another OSS project they work on. In the second, programmers reuse software code by taking code from one OSS project and employ it in another OSS project. In previous work, we found knowledge spillovers via the first channel. Focusing on the second channel, in this paper we develop a methodology to measure software reuse at the micro-micro level in a large OSS network. We then examine whether there are (spillover) benefits from software reuse. Finally, we examine which factors explain software reuse. Key findings involving software include the following:

• Controlling for other factors that explain success, projects that reuse code from a greater number of projects have higher success.

• Controlling for other factors that explain whether a project reuses code from other projects, younger projects are more likely to reuse code than older projects, while older projects are more likely to have their code reused.