The bureaucratization of science
Introduction
“Secretaries post off papers from the laboratory at an average rate of one every ten days. However, far from being reports of what has been produced in the factory, members take these papers to be the product of their unusual factory.” (Latour and Woolgar, 1979:47).
Science is increasingly becoming a team activity (Wuchty et al., 2007). While this trend began decades ago (Price, 1963, Swatez, 1966), the sizes of contemporary research teams in many fields are beginning to approach that of medium-sized firms (Biagioli, 2003, Birnholtz, 2006, Milojević, 2014, Pavlidis et al., 2014, Salonius, 2008). Rather than a focus on an individual's lab bench, scientific work increasingly takes place in a setting that more closely resembles a small “factory” or “quasi-firm”, run by a “small businessperson” lab director (Etzkowitz, 1983, Hackett, 1990, Latour and Woolgar, 1979, Shrum et al., 2007). This growth in the size of scientific work teams raises the question of the impact of size on the organization of scientific work (Carayol and Matt, 2004, Chompalov et al., 2001, Swatez, 1966). We extend prior work on the organization of science by examining the internal organization of scientific projects, in particular how the structuring of research projects varies by size and environmental context, building on the classic sociology of organization structures (Blau, 1970, Child, 1973, Meyer, 1972, Pugh et al., 1968).
We argue larger research teams are associated with more bureaucratic structuring of the team: greater division of labor, standardization, hierarchy and decentralization. Furthermore, project scope and team interdependence also affect bureaucratic structuring. Finally, the size–structure relation is contingent on the level of interdependence in the research team.
In addition to developing the sociology and economics of science, this work also tests the utility of organization theory for explaining the structures of self-organizing groups of professionals, and by examining the effects of size at modest group sizes (with the bulk of the projects having on the order of 5–10 people), to see how sensitive these size–structure relationships are across even a modest size range.
Two key insights drive this discussion. First, a scientific project is not a point mass, but consists of a group of members organized along a variety of dimension (Barley and Bechky, 1994, Carayol and Matt, 2004, Chompalov et al., 2001). And, this internal structure may be critical to performance (Andrews, 1976, Carayol and Matt, 2004, Cummings et al., 2013, Fox and Mohapatra, 2007, Hollingsworth, 2009). Secondly, science is not science. Fields differ significantly in their structure and dependencies (Collins, 1975, Fuchs, 1992, Hargens, 1975, Whitley, 1984). Therefore, we examine the internal structure of scientific projects, and the environmental contexts in which these structures operate.
In the following sections, we discuss the changing nature of scientific work, use organization theory to generate hypotheses about the structural implications of these changes, test these hypotheses using recently collected data from a broad sample of research projects across scientific fields, and then conclude with a discussion of the implications of these findings for the sociology and economics of science: in particular, training, careers, and the reward structure in science.
Section snippets
The growth of organized science
While science being conducting in organizations (such as universities, government labs, and industry labs) is not a new phenomenon (Blau, 1994, Pelz and Andrews, 1976), we are observing a fundamental change in the organization of individual research projects. While traditionally science is seen as an individual endeavor (Hagstrom, 1964, Shrum et al., 2007), increasingly scientific projects are group activities, and the groups are growing larger (Adams et al., 2005, Wuchty et al., 2007). While
Size, interdisciplinarity, technology, and the bureaucratization of scientific work
This work on the relation between size and structure begins with Weber’s classic analysis of the characteristics of bureaucratic organization (in contrast to paternalistic or collegial organization), which emphasizes the importance of division of labor, formalization and standardization, hierarchy and decentralization, as well as specialized competence and internal careers, among other aspects of the ideal-type bureaucracy (Weber, 1978). Weber (1978) notes that bureaucratization is associated
Data and methods
To test these questions requires information from a large sample of projects spanning fields and institution types. We will use data from a survey of scientists in the US. The population of interest is scientific projects in the fields of science covered by the Web of Science. Here, field is defined by the field of the journal where the paper is published, as defined by the ISI classification (classified into 22 fields covering all ISI science and social science journals, see Appendix A). This
Results
We begin with a discussion of descriptive statistics (see Table 1). Average project size is 7 members (including an average of almost three non-authors). The mean of interdisciplinarity is 2 fields. About three-quarters of projects have internal division of labor, while 24% and 31% report some external division of labor and existence of a non-author technician, respectively (cf. Leahey and Reikowsky, 2008). Thus, division of labor, including use of technicians, is quite common in contemporary
Implications
Our results suggest that the increasingly common large research groups in science have more bureaucratized work organization. In the following sections, we discuss the potential implications of these changes for scientific training, careers, and authorship.
Conclusions
Scientific work is increasingly “organized”. This provides an opportunity for incorporating organization theory into the study of science (cf. Chompalov et al., 2001, Collins, 1975, Fuchs, 1992). We find that university research groups commonly share the features of bureaucratic structuring, including division of labor, standardization and formalization, hierarchy, and decentralization. In addition, size is a key driver of this bureaucratic structuring, consistent with predictions based on
Acknowledgments
We thank Ed Hackett and Rogers Hollingsworth for encouraging this project with valuable advice. We also thank the participants in conferences and seminars at the American Sociological Association Annual Meeting, New York; Politecnico di Milano, Department of Business Engineering; Georgia Tech School of Public Policy; University of Bordeaux, IV; Conference on the Organisation, Economics and Policy of Scientific Research, Collegio Carlo Alberto, Torino; Korean Advanced Institute of Science and
References (115)
- et al.
Scientific teams and institutional collaborations: evidence from U.S. universities, 1981–1999
Res. Policy
(2005) - et al.
Does research organization influence academic production? Laboratory level evidence from a large European university
Res. Policy
(2004) - et al.
Founders’ human capital and the growth of new technology-based firms: a competence-based view
Res. Policy
(2005) - et al.
Credit where credit is due? The impact of project contributions and social factors on authorship and inventorship
Res. Policy
(2013) - et al.
Creativity in scientific teams: unpacking novelty and impact
Res. Policy
(2015) - et al.
Organizational design of university laboratories: task allocation and lab performance in Japanese bioscience laboratories
Res. Policy
(2015) - et al.
Organizational alienation: a comparative analysis
Am. Sociol. Rev.
(1966) - et al.
Rescuing US biomedical research from its systemic flaws
Proc. Natl. Acad. Sci.
(2014) Creative process
- et al.
Organizational size, structure, and efficiency: a test of a Blau–Hage model
Am. J. Econ. Sociol.
(1982)