Do public R&D subsidies produce jobs? Evidence from the SBIR/STTR program

• Published in: Research policy 2021-09, Vol.50 (7), p.104286, Article 104286
• Main Authors: Lanahan, Lauren, Joshi, Amol M., Johnson, Evan
• Format: Article
• Language: English
• Subjects: Employment; Innovation; R&D; Small business
• ISSN: 0048-7333; 1873-7625
• DOI: 10.1016/j.respol.2021.104286

•We examine the impact of public R&D on job creation with a two-study research design.•First, we find federal recipients hire fewer employees than matched non-recipients.•Second, we confirm this effect with a natural experiment using noncompetitive funds.•Firm recipients outsource a portion of R&D projects to external partners.•Firms face competing expectations to stimulate innovation while also creating jobs. Governments make substantial investments in small businesses to promote innovation and create jobs. While there is extensive evidence that R&D policies catalyze innovation, there is less empirical evidence on their employment effects. We examine the impact of public R&D subsidies on net job creation with a two-study empirical research design that draws upon first-time recipients of the U.S. SBIR/STTR program from the five leading federal agencies from 2000 to 2015. This two-study design enables us to analyze the extensive and intensive margins of the program. For the first study, we compare federal recipients to a matched set of non-recipients and find awardees hire fewer employees than their counterparts. For the second study, we leverage a natural experiment to investigate the employment effects of additional matching funds from state governments. We find evidence that federal recipients that receive noncompetitive matching funds also hire fewer employees than federal recipients located in states without the match. We investigate a series of features that may accelerate or attenuate the total effect and find that ventures substitute away from hiring toward partnering and gaining access to external resources. Moreover, firms with heightened levels of risk, indicated by medium-to-low credit records, account for the greatest difference. These results highlight tensions for ventures as they balance competing expectations of the program to stimulate innovation while also creating jobs. Policymakers should reconsider whether emphasizing direct job growth is appropriate when evaluating this type of R&D program.