Association for Public Policy Analysis & Management

In 2014 over $52 billion was invested in U.S. doctoral-granting research institutions in the fields of science and engineering (S&E) with federal research and development (R&D) expenditures accounting for over 60 percent (Lanahan et al., 2016). Despite the high level of R&D expenditures, our understanding of the efficacy of these investments remains limited (Jaffe, 2002) with previous research focused on outcomes of senior scholars (Jacob & Lefgren, 2011a; Goldfarb, 2008; Arora & Gambardella, 2005; Azoulay et al., 2011) and teams (Conti & Liu, 2015; Jones et al., 2008; Wuchty et al., 2007). However, in response to a series of recent discussions in Nature lamenting the limitations of the current S&E graduate programs (Gould et al., 2015; Callier & Polka, 2015; Woolston, 2015; Nature, 2015), we focus on R&D funding for graduate students – an often overlooked, yet integral population of the S&E innovative workforce.

More specifically, we consider the role of gender for research production. Numerous studies have found gender discrepancies in research productivity due to institutional biases and variation in work-life demands (e.g. Pezzoni et al., 2016; Moss-Racusin et al., 2012; Buffington et al., 2016). Moreover, as Moss-Racusin et al. (2012) cogently notes, “the formative pre-doctoral years are a critical window, because students’ experiences at this juncture shape both their beliefs about their own abilities and subsequent persistence in science” (pg. 16475).

We draw upon a sample of high quality S&E graduate students that either are award recipients or honorable mentions of the prestigious U.S. National Science Foundation (NSF) Graduate Research Fellowship Program (GRFP). This accounts for approximately the top 20 percent of the competitive applicant pool. We augment this sample with data on graduate training, research production, and professional placement from Proquest, web-based data, and the Scopus database.

We rely on a series of bibliometric measures to examine outcomes of graduate training and research productivity by gender. These include measures of total peer-reviewed publication activity, research leadership (measured by first-authored publications), independent research (measured by publications that do not include the student’s advisor as a co-author), and research quality (measured by journal impact and forward citations). The measures of research leadership and independent research offer important insights as to how the graduate student utilizes the R&D resources to develop their own research record.

Our primary analysis utilizes a triple difference model to estimate the treatment effect of being a female awardee. Preliminary results from triple difference analyses indicate that female recipients on average produce 1.659 fewer publications than male awardees in the ten years following the award. In addition, while male awardees are significantly more productive than male honorable mentions, there is no productivity difference between female awardees and female honorable mentions.

Moreover, we find robust and negative effects for both first authored publications and co-authored publications with the student’s advisor for female awardees. Additional analysis will examine how these effects vary by the gender match between the student and their advisor. This study aims to illuminate the varied role of early-career funding on research productivity by gender.