Association for Public Policy Analysis & Management

The US federal government invest billions of dollars each year in the academic research enterprise. With the aim of spurring innovation, these funds touch graduate students, post-doctoral fellows, and faculty at all stages of their careers. However, evidence from the life sciences has shown a relatively small impact of these funds on the research productivity of graduate students. This muted effect is driven in part by a near zero effect for female recipients.

In this paper, we expand previous efforts on this topic by assessing the impact of federal funding across all of the major scientific divisions, including the life sciences, engineering, math and physical sciences, and social sciences. We utilize the National Science Foundation’s (NSF) Graduate Research Fellowship Program and its unique attributes to assess of early-stage federal funding impacts research productivity, leadership, and contribution in the ten years following the award. With a sample of over 6,000 graduate students, we compare publication activity between awardees and honorable mentions. Honorable mentions are a unique feature of this program in which students are deemed highly qualified by peer review but do not receive funding due to financial and programmatic constraints of NSF. Within this assessment, we examine impact as measured by peer-reviewed publication count and follow-on citations.

We extend the analysis and also examine how the award’s impact varies across gender and race. While it has been shown in the life sciences that the award has a differential effect across gender, it is unknown how gender impacts these measures in other fields. In addition, it is unknown how race may moderate the impacts in any of these divisions. Gender and race are identified and vetted across multiple techniques including first name registrations with the Social Security Administration, facial recognition software, and name recognition software. Then, triple difference and coarsen exact matching (CEM) techniques are used to estimate the impact of gender and race on award impacts of research productivity. Results from these analyses have important implications to understanding how award receipt is viewed across scientific fields and how these impacts vary across gender and race.