مشاغل علمی، اختراع ثبت شده، و بهره وری: تجربه صنعت به عنوان سرمایه انسانی علمی و فنی

We examine career patterns within the industrial, academic, and governmental sectors and their relation to the publication and patent productivity of scientists and engineers working at university-based research centers in the United States. We hypothesize that among university scientists, intersectoral changes in jobs throughout the career provide access to new social networks and scientific and technical human capital, which will result in higher productivity. For this study, the curriculum vitae of 1200 research scientists and engineers were collected and coded. In addition, patent data were collected from the U.S. Patent and Trademark Office. The overarching conclusion from our analysis is that the academic scientists’ and engineers’ research careers we studied are quite different than characterized in the research productivity literature that is a decade or more old. The wave of center creation activity that began in the early 1980s and continues today has resulted not only in greater ties between universities and industry, but also markedly different academic careers.

As universities ramp up to increase their technological contributions and their commercial focus, the universities’ cultures inevitably change and, indeed, the very nature of the university as a social institution may undergo change. How do these changes interact with the career patterns and prerequisites of the scientists and engineers who are expected to take the lead in universities’ technology-based commerce? What are the impacts of industrial “opportunities” or “contamination?” The way we approach these questions is to examine the career and attributes of university scientists and engineers who presently engage in high levels of patenting and commercial activities, especially those who have worked in industry as well as in universities. How are they different? The answer to this question may tell us much about the potential and the potential limits of universities’ technology-based commerce. Thus, a central question of our paper is “during the course of scientists’ or engineers’ careers, what are the effects of changes in jobs, especially movement between universities and industry, on productivity, especially patenting?” Specifically, we not only test a model of research productivity, specifically publications and patent counts, considering the effects of diversity of job experiences, but also the interactive impacts of grant awards, early publication, and mentoring opportunities. In contrast to the existing body of literature, many of these variables can be affected by the policies of funding agencies. Our research uses data from a U.S. Department of Energy (DOE) and National Science Foundation (NSF) funded project called the Research Value Mapping Program, which is headquartered at the Georgia Institute of Technology School of Public Policy. The program is studying new social and economic approaches to the valuing of publicly funded research—specifically as it is carried out in research centers. The data for our study include the curriculum vitae (CVs) of 1200 research scientists and engineers supported by DOE, the Department of Defense, and NSF research centers have been collected and coded. In addition to the CV data, patent data were collected from the U.S. Patent and Trademark Office (USPTO) database. Our approach differs from previous approaches in that it examines the pattern of researchers’ careers over time and the effect of job changes and other critical events to the rate of productivity over time. It has its intellectual roots in a scientific and technical human capital (S&T human capital) theory (Bozeman and Corley, 2004 and Bozeman et al., 2001) of knowledge generation that suggests that human and social capital building experiences over time affect the formation and pattern of scientific careers, and these opportunities intersect and act in synergistic ways to affect long-term productivity. In brief, we find considerable intersectoral diversity within the careers scientists and engineers working in science and engineering research centers. And, while higher publication productivity seems to be associated with more “traditional” academic careers, patent productivity seems to be associated with less traditional, more industry oriented careers even though a substantial fraction of those who have worked in industry continue to patent in while in academia. In addition, we argue that there is potential policy benefit in removing artificial boundaries set up in the extant literature on academic and industry careers where careers are treated as if they were consistently academic or industrial in nature. The theory implies that a diversity of job experiences will affect collaborative patterns and the exchange of human capital through the building of a wider variety of network ties and social capital. This assertion is related to Granovetter's (1973) notion that greater benefits accrue to those who are able to tap “weak ties” (e.g., a friend of friend) due to exploitation of human and social capital that is non-redundant with one's one web of human and social capital endowments1.

If there is any single, overarching conclusion from our analysis it is that academic scientists’ and engineers’ research careers are quite different than characterized in the research productivity literature that is a decade or more old. This is particularly the case for the researchers we studied, all of whom are affiliated with university researcher centers. The wave of center creation activity that began in the early 1980s and continues today has resulted in markedly different academic careers and greater ties between universities and industry. One of the individuals most responsible for launching the wave of university researcher centers in the U.S. is Erich Bloch, former National Science Foundation Director and Presidential Science Advisor. In a recent interview (Bozeman and Boardman, 2004a and Bozeman and Boardman, 2004b), Bloch described the motivations behind the creation of the NSF Engineering Research Centers program, a program that has served as a template for the more than 300 present-day NSF university centers as well as centers developed by other government agencies and in other nations: The idea (of multidisciplinary collaborative research) needed to penetrate the whole engineering community. I thought 12 (engineering research centers) would send a message. When I looked at NSF it was the paymaster for principal investigators. I thought this was dead wrong. The idea was that the NSF should be concerned about competitiveness and look for new avenues to tie together industry and government and universities. At least within the domain of university research centers, there does, indeed, seem to be considerable ties and these are reflected not only in more passive ways but also through changes in careers. It is fair to say that there is now a revolving door between industry and university research jobs. Our data provide some preliminary evidence of the rate at which the door revolves. On average, nearly one in six of our respondents’ total jobs positions were industry jobs and one in eight of their career years were spent in industry jobs. Approximately half had one or more jobs in industry. One-quarter had worked for government. And, nearly half began their careers in non-academic jobs—for 33 percent the first job was in industry and for 15 percent the first job was government. An average of 24 percent of the grants awarded to researchers in the RVM dataset came from industry. While our data are representative of university centers, not all of academe, it is still surprising to us that academic and industry careers have melded to such a large extent. One finding that is particularly at odds with the conventional ivory tower model is 20 percent of our respondents took their first academic job 5 or more years into their career. In the past, the conventional wisdom was that a first job in industry all but foreclosed an academic career. Perhaps what has changed, however, is the nature of “an academic career,” with increasing emphasis on industry relations, applications and technology development in addition to the conventional formula of publish-for-tenure. If, indeed, academic careers have changed fundamentally, in university centers and perhaps elsewhere, it is doubly important to understand that the formula for publication productivity may be quite different from the one for patenting productivity. Not surprisingly, those scientists who have spent a substantial percentage of their career in industry jobs have more funding from industry and a higher rate of patent productivity. This may be due, at least in part, to the trust and social capital required for commercial activity (Daellenbach and Davenport, 2004). In contrast, a higher proportion of the career spent in industry is negatively related to publication productivity. Just as important, and especially policy relevant, the source of grants affects the type of productivity. Those with government grants have higher publications productivity but government grants are actually negative for patenting. Similarly, those with industry grants have higher patenting rates but lower publications rates. The fact that basic grant and contract sources do not complement one another may be a significant policy issue. Some public policies promote movement back and forth between industry and universities and others promote active collaboration or personnel exchange, such as the NSF Grant Opportunities for Academic Liaison with Industry (GOALI) program. From a policy perspective, it is often assumed that the job exchanges and transitions are useful. But most of the literature on research productivity either assumes or demonstrates negative effects on productivity. Moreover, many institutional barriers remain to university scientists’ working effectively with industry (Hall et al., 2001). But as policymakers seek to encourage flexible scientific careers and university–industry ties, it is crucial to sort out the effects of collaborative work and career transitions on productivity. Our results suggest that the impacts of intersectoral career transitions are complex. As discussed above, job transformations lead to publications productivity boosts not only when the transformation is from industry-to-academia but also from academic-to-industry. Why? It is easy to see why moving from an industry position to an academic one would increase publications, but why does a move from academia, generally publications-dominated, to industry has the same positive effect? We advanced many possibilities, including various methodological artifacts such as selection effects and time lag effects. Nevertheless, this is worth further explanation and, moreover, seems to support the nature of accumulating S&T human capital in multiple settings. Regarding S&T human capital and public policy, perhaps our most disturbing finding is the negative impact of postdoctoral positions on productivity. This is worrisome not only because of the proliferation of postdoctoral positions, including serial postdocs, but also because some public policies actively promote postdocs through traineeships and other programs. Our findings show that those with no postdoctoral job had a patent rate more than five times those who had postdoctoral jobs. Regarding publications, those with no postdoctoral job averaged 3.6 publications per year compared to 3.2 publications per year for those with one postdoc and 2.7 for those with more than one postdoctoral job. It is not easy to sort out either the causality or the policy implications of these findings. Among the many possibilities: (1) postdoctoral positions really have become exploitative and are chiefly a low wage means of advancing senior scientists’ agendas; (2) one postdoc position may be useful but the point of diminishing return occurs before a second one; (3) the most talented researchers are the ones who immediately receive and take offers from permanent or tenure track positions; (4) there may be a correlation between having a postdocs and discrimination against women and foreign nationals; (5) the general finding may mask very important field effects (though a level of aggregation below that we examined here). The basic point is that the negative relation in the RVM dataset of both publications and patenting productivity to having a postdoctoral position is worrisome, especially so because the percentage of persons entering postdocs continues to increase. For example, in 1973 only 27 percent of the people earning biomedical Ph.D.'s went into postdoctoral positions, but by 1995 the proportion had risen to 63 percent (Monastersky, 2004). Now that in some fields most researchers have a postdoctoral research position in their careers it seems imperative to know more about why that experience seems to dampen productivity. Finally, if government policy chooses to focus on the commercial productivity of academic institutions, as seems to be the policy trend, then there is reason to believe that the solution may be human capital in nature. The hiring of researchers with industrial job experience and, perhaps, visiting positions and exchanges with industry may be a productive means of boosting the commercially relevant innovation of universities. The GOALI program, to encourage academic researchers to take sabbaticals in industrial research jobs and vice versa, and similar programs that can be found at Department of Energy and the Department of Commerce agencies, warrant evaluation in relation to the career diversity findings we present. The question implicit in the title to our paper is “What are the S&T human capital implications of academic careers that include industry working experience?” Our results suggest important effects, sometimes profound ones. The revolutions of the revolving career door warrant close watching.