Explaining Heterogeneity in the Organization of Scientific Work

• 成果类型: Article
• 署名作者: Rahmandad, Hazhir; Vakili, Keyvan
• 署名单位: Massachusetts Institute of Technology (MIT); University of London; London Business School
• 刊物名称: ORGANIZATION SCIENCE
• ISSN/ISSBN: 1047-7039
• DOI: 10.1287/orsc.2019.1303
• 发表日期: 2019
• 页码: 1125-1145
• 关键词: scientific teams; collaboration; ACADEMIC RESEARCH; science policy; research relevance; research funding; university industry relationship; performance heterogeneity

摘要

Prior studies of academic science have largely focused on researchers in life sciences or engineering. However, while academic researchers often work under similar institutions, norms, and incentives, they vary greatly in how they organize their research efforts across different scientific domains. This heterogeneity, in turn, has important implications for innovation policy, the relationship between industry and academia, the scientific labor market, and the perceived deficit in the relevance of social sciences and humanities research. To understand this heterogeneity, we model scientists as publication-maximizing agents, identifying two distinct organizational patterns that are optimal under different parameters. When the net productivity of research staff (e.g., PhD students and postdocs) is positive, the funded research model with an entrepreneurial scientist and a large team dominates. When the costs of research staff exceed their productivity benefits, the hands-on research approach is optimal. The model implies significant heterogeneity across the two modes of organizing in research funding, supply of scientific workforce, team size, publication output, and stratification patterns over time. Exploratory empirical analysis finds consistent patterns of time allocation and publication in a prior survey of faculty in U.S. universities. Using data from an original survey, we also find causal effects consistent with the model's prediction on how negative shocks to research staff-due to visa or health problems, for example-differentially impact research output under the two modes of organization.