Working Paper: NBER ID: w22241
Authors: Carolin Haeussler; Henry Sauermann
Abstract: Even though teams have become the dominant mode of knowledge production, little is known regarding how they divide work among their members. Conceptualizing knowledge production as a process involving a number of functional activities, we first develop a conceptual framework to study the division of labor in teams. This framework highlights three complementary perspectives: (1) individual level (the degree to which team members specialize vs. work as generalists), (2) activity level (the degree to which activities are concentrated among few team members vs. distributed among many) and (3) the intersection between the two (e.g., which activities are performed jointly by the same individual). We then employ this framework to explore team-based knowledge production using a newly available type of data – the disclosures of author contributions on scientific papers. Using data from over 12,000 articles, we provide unique descriptive insights into patterns of division of labor, demonstrating the value of the three complementary perspectives. We also apply the framework to uncover differences in the division of labor in teams of different size, working in novel vs. established fields, and on single vs. interdisciplinary projects. Finally, we show how division of labor is related to the quality of teams’ research output. We discuss opportunities for extending and applying our framework as well as implications for scientists and policy makers.
Keywords: division of labor; scientific collaboration; knowledge production
JEL Codes: J24; O31; O32
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| division of labor in scientific teams (M54) | team size (L25) |
| team size (L25) | average number of activities performed by each member (D71) |
| average number of activities performed by each member (D71) | greater specialization (Y80) |
| novelty of a research area (O32) | higher division of labor (L23) |
| higher division of labor (L23) | greater quality of scientific output (L15) |
| share of generalists in a team (C92) | citation impact (A14) |
| share of specialists (I11) | citation impact (A14) |