Click for next page ( 53


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 52
Pane! Research Perspectives on the SBIR INTRODUCTION Zoltan Acs University of Maryland Dr. Acs opened the session by noting the recent publication by the Small Business Administration (SBA), State of Small Business, that is an assessment and overview of federal programs designed to stimulate the growth and develop- ment of "new-technology-based" firms. Programs examined in the publication included the SBIR program, the Small Business Technology Transfer program, several financing programs managed by SEA, along with the Commerce Depart- ment's Advanced Technology Program (ATP) and Manufacturing Extension Partnership. There is, however, a dearth of academic studies examining the impact of programs under which the federal government and the private sector have formed partnerships to develop and deploy new technologies. Most of these programs, he noted, built on the experience of the states and have been instituted since 1980. In this context, Dr. Acs pointed to the value of work to be presented by the panel's main speaker, Josh Lerner of the Harvard Business School, on the role of the SBIR program. 52

OCR for page 52
PANEL II 53 THE GOVERNMENT AS VENTURE CAPITALIST: AN ANALYSIS OF THE SBIR Joshua Lerner Harvard Business School After spending most of the past decade studying issues associated with tradi- tional venture-capital organizations such as how they finance high-technology companies and what the impacts on innovation are Dr. Lerner has turned his attention to the interaction of public policy and the entrepreneurial sector. This is the area of concern that occupied him during the early years of his career, spent working on technology policy issues in Washington, and a recent surge of interest both in the United States and abroad has drawn him back to it. Public Support for High-Technology Start-Ups Dr. Lerner began the presentation of his work by describing the size of pub- lic-sector support for entrepreneurial firms relative to that provided by private venture capitalists. Although public programs no longer dominate the field the way they did in the 1960s, when their combined size was three times that of the private venture funds, their resources still amount to a "significant fraction" of those of an expanded independent venture sector. In addition, there is much evidence that government programs such as the Small Business Investment Cor- poration (SBIC) program retain considerable importance, whether in the list of firms that have received money from SBICs or SBIR when they were still pri- vately held and relatively unknown or in the experience that many leaders of today's U.S. venture industry received under SBIC several decades ago. The growth of venture-capital sectors in countries such as Israel and Taiwan, where the role of government policy and programs in encouraging entrepreneurial activ- ity has been significant, reinforces this picture. Dr. Lerner concurred in the view that SBIR has received little attention or scrutiny for a program of its size. Qualifying his own study as preliminary in that it concerns a limited set of questions, he said it attempts a systematic look at the growth and evolution over a decade of both the early SBIR awarders and a matched set of firms that have many, although not all, of the same characteristics. In particular, he was interested in ascertaining whether any specific subset of SBIR firms derived exceptional benefit from the awards. Study Summary: An Effective Program Dr. Lerner said that his study suggests that the awards generally make a difference in that the award-winning companies are "associated with greater growth." The greatest growth, however, is associated with awards made to com

OCR for page 52
54 THE SMALL BUSINESS INNOVATION AND RESEARCH PROGRAM panics in regions of the country with significant entrepreneurial activity in high- technology industries. He also found that the firms that won the fewest awards displayed growth equal to or greater than that displayed by firms that won mul- tiple awards. The picture that emerges, he argued, is of a program that is working effectively and seems to be playing an important role in stimulating young firms; much of the evidence is consistent with the argument that SBIR awards are important in providing young firms with certification, a "stamp of approval" they can then use in obtaining access to other resources. At the same time, Dr. Lerner allowed, it may well be possible that "some distortions in the awards process" exist and that they should be further assessed. Rationale for Subsidies Before taking up the data, Dr. Lerner proposed examining the rationale, from a policy perspective, for offering awards and subsidies to small, high-technology companies. He cited two arguments that economists make in favor of them: . . Spillovers or externalities: Many of the benefits of R&D programs do not accrue uniquely to those undertaking the research but are captured also by consumers and by firms developing complementary or competing products. This problem may be particularly acute for small firms with relatively limited resources, which may not have sufficient incentive to undertake research and for which subsidy programs may thus play a sig- nificant role. Information problems: Investors are frequently hampered by a lack of technical expertise in assessing the activities of small, high-technology companies, for which reason viable technologies may not attract the fund- ing they merit. Though theoretically designed to deal with this challenge, venture capital represents a "very narrow and targeted financing mecha- nism." To illustrate, Dr. Lerner cited figures for 1996, during which roughly 1 million firms began operation in the United States: Although this was a record year for venture-capital activity, only about 600 firms received venture financing for the first time. "Of these million compa- nies, many were mom-and-pop convenience stores, solo consulting prac- tices, and so forth, which did not need the financing," he acknowledged. "But even after you throw them out, there is still a very substantial num- ber of companies relative to the very small number that received venture financing." Another manifestation of the venture sector's lack of exper- tise in evaluating information, according to Dr. Lerner, is a tendency to follow investment fashions that has led it to focus narrowly on a succes- sion of "hot" sectors, such as Internet service providers, search engines, and Internet chat channels. This investor behavior, which may leave interesting and potentially viable companies in less fashionable areas out

OCR for page 52
PANEL II 55 in the cold, can to some extent be offset through programs like SBIR that depend on knowledgeable people in many areas of technology to evaluate company proposals. Striking a cautionary note, Dr. Lerner pointed to economic literature that highlights "real dangers" associated with subsidy programs. Work on "political capture" argues that particular interest groups can succeed in steering benefits to themselves and that programs can be used in ways that may not be to the best benefit of those taking part.) Some specific critiques of SBIR have argued that well-connected firms may end up receiving a disproportionate share of awards or that awards may go to companies that are likely to be successful without them, which in turn might distort perceptions of the program's effectiveness. In his own analysis, Dr. Lerner tried to isolate both potential benefits and possible dis tortions, and to discern some distinct patterns in the relative success of awarders. Challenges in Assessing Public Technology Programs Dr. Lerner stated that the two most common approaches to assessing public technology programs surveys in which the companies involved are asked what the direct benefits of participation in a particular program have been, and aca- demic studies of research funding that take place at the level of the general economy have not necessarily provided answers that reliably demonstrate whether a program is reaching its intended goals. Surveys: One problem with the former approach, polling individual award recipients directly, is that a program's advocates may tend to attribute all their success to the program. Conversely, fears of being perceived "to have done too well through the courtesy of public money" might incline an entrepreneur to play down the role of an award that might in fact have been critical; this has been observed in connection with some biomedical inventions, Dr. Lerner noted. Finally, in many instances the award is so closely related to the recipient firm's main activity that ascertaining what can actually be credited to the award is extremely difficult. Similar problems have been associated with academic assessments, which have tended to focus on the question of "crowding out": whether public funding of research discourages the private sector from funding research on its own. Even though attempts have been made to apply empirical techniques to study the im- pact of specific programs such as SEMATECH or the SBIR program so that the question becomes whether a firm's participation in a particular program causes a ~ See for example: Cohen, L. R., and R. G. Noll. 1991. The Technology Pork Barrel. Washington, D.C.: The Brookings Institution.

OCR for page 52
56 THE SMALL BUSINESS INNOVATION AND RESEARCH PROGRAM reduction in its own internal R&D expenditure such techniques may not capture important dynamics of firm behavior.2 The efforts of small entrepreneurial com- panies, Dr. Lerner reminded the audience, often revolve around a key researcher, so that adding staff is not going to cause progress to accelerate significantly. Therefore, he argued, "it may be a very rational and very sensible response for firms in the short run to take the money they receive from a program like SBIR and reduce their own spending" in order to push back the moment at which they will run out of funds. Long-Run Impacts of SBIR Although both methods of analysis provide some clues to the effectiveness of programs, according to Dr. Lerner, neither addresses the "really critical ques- tion" of what happens to the firms and their technologies in the long run. He attempts to elucidate this issue in his own analysis by examining the evolution and success of the firms that received Phase II awards in the first three SBIR program cycles. Relying on data from the U.S. General Accounting Office for an initial survey of these awards, Lerner charted the winning firms' progress over the subsequent decade. He then compared 500 SBIR-winning firms with 900 other firms selected to match them as closely as possible. Referring to Table V of his paper (see Appendix), he pointed to a comparison of employment growth shown by the SBIR awarders and the matching firms over a ten-year period. Although the two sets of firms were roughly at the same size at the end of 1985, a decade later the SBIR awarders had grown by an average of 26 jobs compared with 5 or 6 jobs for the matching firms. On this basis, he saw a significant differential both in job creation and in sales growth between the awarders and the nonawardees. This differential was magnified in geographical areas where the level of venture-capital activity was high in the 1980s. In regions of the country boasting much high-tech entrepreneurial activity, employment at SBIR-winning firms grew by 50 jobs as compared with 3 jobs at non-SBIR firms; the differences discerned in other regions were not statistically significant. From these data, Dr. Lerner posited a "fundamental pattern" according to which the awarders experience far more growth and the growth is concentrated in high-tech regions and industries; scrutiny of multiple awarders revealed no additional growth to be associated with the additional awards. Besides looking at how well individual firms did under SBIR, it would be desirable to examine the impact of the program on society through its effect on the rate of innovation, on the growth of knowledge, and on spillovers to other firms; doing so would be very difficult, however. Also useful would be compar 2 D. A. Irwin and P. J. Klenow, 1994, High-Tech Subsidies: Estimating the Effects of SEMA TECH. NBER Working Paper No. 4974. Washington, D.C.: National Bureau of Economic Research.

OCR for page 52
PANEL II 57 ing crucial aspects of the SBIR program with similar programs, such as ATP. In Dr. Lerner's judgment, the SBIR program, substantial as it has been, has lagged ATP in evaluating the research function. He encouraged members of the SBIR community to reflect on and, perhaps, emulate ATP's efforts in "bringing in people from policy as well as economic perspectives to try to encourage studies- many of which have been thought-provoking and, in some cases, quite critical." The result, he said, has been "a very healthy dialogue." DISCUSSANT Kenneth Flamm* The Brookings Institution Congratulating Dr. Lerner on the quality of his paper, Dr. Flamm began by listing four topics he wished to discuss in his presentation: the fundamental, underlying rationale an economist might offer for in- venting such a program as SBIR; Dr. Lerner's empirical work; Dr. Flamm's personal experience with the program during his service at the DoD, which might suggest further questions for Dr. Lerner's study; and the issue of social valuation of the SBIR program, in particular the issue of crowding out in R&D, with special reference to the study of SEMATECH mentioned by Dr. Lerner, as Dr. Flamm believes that both the study itself and the manner in which it is often interpreted are seriously flawed. The Rationale for SBIR In addressing his first point which he recast as "Why SBIR?" Dr. Flamm recapitulated two answers offered by Dr. Lerner: . that the social return of R&D projects has often been shown to be greater than their private return, in part for reasons connected with the extent to which private investors can appropriate the benefits of innovative activi- ties; and that information asymmetries in financial markets create a role for govern- ment as a certifier of worthwhile projects. *At the time of this symposium, Dr. Flamm was a Senior Fellow at Brookings. He now holds the Dean Rusk Chair in International Affairs at the LBJ School of Public Affairs at the University of Texas at Austin.

OCR for page 52
58 THE SMALL BUSINESS INNOVATION AND RESEARCH PROGRAM Social Returns: Commenting on the former, Dr. Flamm pointed out that if the fundamental and underlying rationale for government involvement in or sub- sidy of technology programs is seen to be the large potential gap between social return and private return and if the market can be counted on to provide informa- tion on private return only, not on social return, then there is a "natural argument" to be made that government agencies, whose objectives are social by definition, might have expertise to offer in identifying areas of high social return and in steering projects toward those areas. In fact, a program like SBIR that targets particular projects has some advan- tage over some other proposals for capturing social return, Dr. Flamm argued. One such proposal, offered by Scott Wallsten of Stanford University, is using R&D tax credits rather than SBIR to subsidize investment in innovations for which social return is expected to exceed private return. The problem with this idea is that a project with a small private return will have a small private return even after a tax credit; in contrast, a project offering a private return but no social return will have a larger private return after the tax credit. So projects for which low private but high social returns can be expected "are not going to be thrust over the [investment] hurdle rate" by the prospect of a tax credit, which instead would encourage "excess" investment in projects that promise reasonably high private returns but little or no social return. Dr. Flamm thus characterized the tax credit as a "blunt instrument" when it comes to identifying areas with high social returns and trying to get projects in those areas funded; in contrast, he argued that, at least in theory, [the SBIR program] has some potential for funding some projects with higher social returns. Information Asymmetries: Praising Dr. Lerner's discussion of the ques- tion of information asymmetries in financial markets, he connected the notion that the role of government as certifier of a project' s value might be particularly important for small firms to another phenomenon that he judged to be worth exploring: While large firms should be able to fend off litigation attacks on their intellectual property, small firms lack the basic resources to defend themselves in court. Suggesting a rationale for the SBIR program not mentioned by Dr. Lerner but often invoked by those who visited him at his former office at the Pentagon, Dr. Flamm posited the existence of an imperfection in capital markets stemming from the limitation of liability under the bankruptcy laws. The only alternative to venture capital for a small company seeking to raise financing is to seek a bank loan; but lenders require collateral, "and if your collateral is your energy and ideas, then because the institution of slavery is illegal there is no collateral if you go bankrupt."

OCR for page 52
PANEL II 59 Empirical Issues Addressing the second of his four original points regarding Dr. Lerner's find- ings, Dr. Flamm commented that Dr. Lerner's paper points to a "thought- provoking" correlation, at least on a regional basis, between receipt of SBIR funding and commercial success. Warning that correlations do not necessarily imply causation, Dr. Flamm pointed to three possible explanations for the corre- lation, although he stressed that none is definitively established on the basis of Dr. Lerner's work: . . Government picks winners. "Contrary to popular belief, government does an excellent job of picking winners; that is, the government bureaucrats who are picking these projects are astute enough to pick the firms that are going to succeed anyway." Government causes winners. There are three different possible mecha- nisms: Government certification causes success in dealing with asymme- tries in the financial market; giving the actual dollars causes success, although Dr. Lerner appears to have debunked that by showing that com- panies receiving more dollars do not seem to have a higher success rate than companies receiving fewer dollars; and further contracts from the government cause success, whether those are procurement contracts or R&D contracts. Winners pick the government. There are certain characteristics that lead firms to apply to the SBIR program among them, perhaps, firms that know they are going to be qualified to win this type of grant end up apply . . . . ng and winning. Experience at DoD Many Programs: Turning to a consideration of what he learned through direct experience at the DoD, his third point, Dr. Flamm stated that SBIR is not a single program but, in fact, "a different program in every government agency." Because the methods used to pick the grantees, although subject to general guide- lines, may vary from agency to agency or even from one period to another within a single agency, it is important to specify the source of the funding when assess- ing the program. An R&D Tax: Second, he recalled that he and his colleagues in the Office of the Secretary of Defense (OSD) saw the SBIR program as a tax on agency R&D budgets. They held as their "deepest, darkest suspicion" the fear that the program was "wired": "Somewhere out there, a director of an R&D lab at the Army or the Navy was coming up with topics that his ax-employees or buddies were going to pursue using SBIR grants; and the topic would be written in such a

OCR for page 52
60 THE SMALL BUSINESS INNOVATION AND RESEARCH PROGRAM way that Dr. X, 'retired from the Naval Surface Tank Division' or whatever, was going to get the grant." This anxiety induced OSD officials to seek ways of increasing the visibility and transparency of the SBIR program. Still, Dr. Flamm conjectured that some of the program' s apparent success may actually be result- ing in the "incubation of 'Beltway Bandits"' that firms getting a start through the SBIR program might subsequently stay alive as contract researchers by culti- vating close relationships with government agencies. He suggested that it might be possible to determine whether this is indeed happening by looking at the share of revenues of SBIR-winning firms that come from the government, and that this subject would be an "interesting" one to consider. Social Valuation As for the last of his four issues, that of social valuation, Dr. Flamm noted Dr. Lerner's even-handedness in referring to a study by Irwin and Klenow that concludes that the main effect of federal funding of the SEMATECH consortium is the reduction by private firms of their spending on R&D.3 He declared, how- ever: "I would like for the record in a public forum to point out that there are some fatal problems with that study that never get mentioned." He said that there are two specific errors in the study, but he had time to speak about only one of them. Using a sample drawn from Standard & Poor's Compustat database of semi- conductor companies, the authors distinguished between those that belonged to SEMATECH and those that did not. But more than half of the non-SEMATECH firms in the sample were makers not of semiconductor devices but of semicon- ductor equipment and materials. The companies are classified together with the chipmakers because there is no separate Standard Industrial Classification cat- egory for semiconductor equipment and materials companies. According to Dr. Flamm, however, if the latter are included, it skews the sample. Until the study is redone making a "proper parsing of the data," he recommended suspend- ing belief on what the potential conclusions might be. DISCUSSION Dr. Acs opened the discussion by recounting, in support of Dr. Lerner's arguments, two findings from a paper he had recently published on local geo- graphic spillover between university research and high-technology innovations: (1) Most innovations by small companies that involve university research take place very close geographically to the university. He cited the distance as 50 3 Irwin and Klenow, 1994, op. cit.

OCR for page 52
PANEL II 61 miles, which he defined as "driving distance in Boston in one day." (2) Regions that have high concentrations of large firms or of branch plants of large firms produce fewer innovations. "So there is reasonably strong evidence," he ob- served, "that small firms, if they dominate an area, will end up being very active in innovation." Dennis Chamot of the National Research Council posited that an ideal suc- cess rate for the SBIR program would be somewhere between 100 percent, which would indicate that selections were overly cautious, and zero, in which case the program would not be yielding any social benefit. He asked what an acceptable rate of success for the program might be. Dr. Lerner suggested that the venture- capital industry provides the best benchmark, because it has had a huge impact on innovation within the industries it has funded, as well as having created a large number of companies that have been able to access the public capital markets. But even though venture-capital firms are extremely selective, funding only one in 100 or 200 of the business plans they consider, they indicate that only one in three to four ventures they fund comes to be regarded as a success. And in the case of even the most established venture-capital groups, the vast majority of returns are generated by a very small number of very highly successful firms, such as Lotus, Netscape, or Genentech. In both the public and the private sectors, "failure is probably going to be the most likely outcome" of investment in small entrepreneurial companies, Dr. Lerner stated. "What one is really looking for is that out of the many failures will come a few that will just be so spectacularly successful that that will undo the many small losses. But that can often be a hard lesson to absorb." James Woo of the Small Business Technology Coalition pointed out that, as some federal agencies have not funded business per se and as the procurement programs of others notably the Pentagon have virtually precluded their going to small business, the SBIR program allows the government to tap into the inno- vative capabilities of small technology companies in connection with specific agency missions. Spin-offs that result may bring wider economic benefits, said Mr. Woo, dismissing the issue of whether public R&D money crowds out private money as irrelevant. Lee Mercer of the National Association of Small Business Investment Com- panies, recalling his involvement with SBIR in its early years, cited two justifica- tions for the program that were current at that time: (1) that small firms could produce innovations more cost-effectively than large firms, and that if the govern- ment was going to procure innovation, it ought to do so in the most cost-effective manner in line with the "Reagan mantra at that time: 'We have to do more with less"'; and (2) that, all other things being equal between two competitors for a particular contract, the government should favor the firm that shows the ability to commercialize the technology not necessarily the potential to succeed as a company.