Community-Based Production of Open Source Software: What Do We Know about the Developers Who Participate?
Paul A. David
Stanford University - Department of Economics; University of Oxford - All Souls College; UNU-MERIT (Maastricht)
Joseph S. Shapiro
Yale University, Department of Economics; National Bureau of Economic Research (NBER)
September 23, 2008
This paper seeks to close an empirical gap regarding the motivations, personal attributes and behavioral patterns among free/libre and open source (FLOSS) developers, especially those involved in community-based production, and considers the bearing of its findings on the existing literature and the future directions for research. Respondents to an extensive web-survey's (FLOSS-US 2003) questions about their reasons for beginning to work FLOSS are classified according to their distinct motivational profiles by hierarchical cluster analysis. Over half of them also are matched to projects of known membership sizes, revealing that although some members from each of the clusters are present in the small, medium and large ranges of the distribution of project sizes, the mixing fractions for the large and the very small project ranges are statistically different. Among developers who changed projects, there is a discernable flow from the bottom toward the very small towards to large projects, some of which is motivated by individuals seeking to improve their programming skills. It is found that the profile of early motivation, along with other individual attributes, significantly affects individual developers' selections of projects from different regions of the size range.
Number of Pages in PDF File: 71
Keywords: Open-source software, FLOSS projects, community-based peer production, population heterogeneity, micro-motives, motivational profiles, web-cast surveys, hierarchical cluster analysis
JEL Classification: J22, L17, L23, L25, L39, P13
Date posted: October 19, 2008 ; Last revised: September 29, 2013