Abstract
The Small Business Innovation Research (SBIR) program is a U.S. federal initiative that supports innovative activity in small- and medium-sized enterprises (SMEs). Introduced by the U.S. Congress in 1982, the Small Business Innovation Development Act requires all federal agencies with an extramural research and development (R&D) budget exceeding $100 million to set aside 2.5% (as of 2011) of their external research budget to fund innovation projects in SMEs. The Small Business Innovation Research (SBIR) program supports SMEs in a three-phase process. Phase I selects projects based on their scientific and technical merit and feasibility. Each awardee SME is granted up to $150,000 over a 6-month period. Phase II continues to fund these promising Phase I projects with up to $1 million over 2 years. Phase III is a commercialization process that involves non-SBIR federal grants or private investments. SBIR has four objectives related to the technological innovation and commercialization of SMEs. Scholars (e.g., Berger, Little, & Saavedra, 1992; Lerner, 1999) have evaluated SBIR and conclude that this program meets its major objectives. Promoting employment is not among the explicit objectives of SBIR. However, an assessment of SBIR employment effects may influence the future of this small business innovation policy. Link and Scott’s book, Employment Growth From Public Support of Innovation in Small Firms, is to my knowledge the first comprehensive study of SBIR’s impact on employment. In the context of continuous high unemployment rates in the United States since the financial crisis, this book features timely research and deserves attention from policymakers.
The book is divided into eight chapters. In chapter 1, Link and Scott provide evidence that the U.S. federal government has embraced innovation-based economic growth strategies, even during economic crises when job creation is generally the top priority. They also discuss the importance of small firms to innovation and propose a theoretical framework to explain why public support for private innovation is needed, based primarily on market failure in addressing positive externalities. Chapter 2 includes the history of the SBIR program and its role in spurring small business innovation via reducing perceived market risk. In chapter 3, the authors introduce the data source of the empirical policy analysis that is the focus of the following chapter: the National Research Council (NRC) database, which was constructed from a comprehensive survey of Phase II SBIR awards in 2005.
Chapters 4 through 7 represent a comprehensive study of the employment effects of Phase II SBIR grants made by the five federal agencies that comprise the NRC database: Department of Defense, National Institutes of Health, National Aeronautics and Space Administration, Department of Energy, and National Science Foundation. Chapter 4 focuses on SMEs’ job retention after Phase II projects were completed. Link and Scott’s descriptive statistics show that this permanent employment effect of SBIR is small: More than one third of completed projects retained one or two employees and more than two fifths retained no employees at all. They further explore the project-specific factors that are associated with job retention for each agency using negative binomial models. Although regression results exhibit cross-agency variations in the effects of those factors, the authors find that, ceteris paribus, federal acquisition and intellectual property are consistently positively associated with employee retention in SBIR projects.
In chapter 5, Link and Scott use a novel counterfactual model to examine the impact of SBIR on the overall employment change of awardee firms. As the authors indicate in this chapter, the NRC database contains only information of SBIR recipients. It therefore does not allow for some ideal experimental or quasi-experimental methodologies in which employment growth can be compared between SBIR firms and matching non-SBIR firms. Link and Scott creatively introduce an employment model with a set of explanatory variables (including firm age) that might affect a firm’s employment growth but are irrelevant to SBIRs. They then estimate the parameters in the model and predict the firm’s counterfactual employment in 2005 if no SBIR grants were received. Link and Scott subsequently compare each firm’s predicted employment and its actual employment. They find that the mean level of actual employment is substantially higher than the mean level of predicted employment for all the agencies except for the National Science Foundation, supporting the positive employment growth effects of SBIR. However, none of these differences are statistically significant.
In chapter 6, Link and Scott further investigate a firm’s characteristics that may contribute to the difference between an awardee firm’s actual employment and its predicted employment in the absence of SBIR. This may provide insights into the SBIR award criteria if employment growth is added to the explicit objectives of this federal program. Several categories of variables that are not used in chapter 5’s prediction model are considered, including firm characteristics, SBIR research variables, market variables, intellectual property variables, funding variables, commercial agreements, and commercial-type variables. Not surprisingly, Link and Scott’s multivariable analysis suggests that these variables’ associations with the employment growth effects of SBIR vary across the five agencies. The authors eventually highlight, though with caution, that “evidence of investments in intellectual property at the end of Phase I might therefore indicate a greater likelihood that the Phase II research will result in a positive employment trajectory for the firm.”
Although chapters 4 and 5 demonstrate a moderate role of the SBIR program in the employment growth of awardee firms, it is likely that SBIR may lead to job creation in non-SBIR firms as a result of their commercial agreements (e.g., licensing) with SBIR projects. Evidence of this is revealed in chapter 6, where the presence of commercial agreements is found to be negatively associated with SBIR’s job growth effects in awardee firms. Chapter 7 presents some descriptive statistics of various commercial agreements for each agency. Data also support that commercial agreements are most prevalent between U.S. firms, and thus, job creation from commercial agreements involving SBIR does not benefit foreign firms “to any significant extent,” chapter 8 concludes.
This book represents a major step toward assessing SBIR’s employment growth effects. Some may question the value of this research, since employment is not among the explicit objectives of the SBIR program. Evidence of additional positive/negative outcomes of a policy initiative, however, can certainly inform policymakers about a policy’s aggregate value and thus influence the future of that policy. In fact, the role of SBIR in facilitating entrepreneurship, which is also not an explicit objective of the SBIR program, has also been examined, for example, in Audretsch, Weigand, and Weigand (2002) and Qian and Haynes (2011). These studies contribute to a comprehensive policy evaluation of SBIR. My major concern with the book is its lack of discussion on whether the employment model introduced by the authors to predict a firm’s employment in the absence of SBIR is appropriate. Apparently, the success of Link and Scott’s methodology used to evaluate the effect of SBIR on firms’ employment depends on the power of this employment model in predicting the counterfactual employment of firms. I understand the NRC database does not allow the authors to test their employment model using a sample of matching non-SBIR firms, but some discussions on the model’s goodness of fit and some robustness tests would help. Additionally, although only a minor concern, the theoretical foundation of this research is not very strong. A solid literature review on the relationships between small businesses, innovation, and employment growth would help readers better understand the authors’ models.