Questions That Blur Political Party Lines
Creating a National Innovation Foundation
In this blueprint, the foundation would build on the few federal programs that already promote innovation and borrow the best public policy ideas from other nations.
The issue of economic growth is on the public agenda in this election year in a way that it has not been for at least 15 years. Policymakers have thus far been preoccupied with providing a short-term economic stimulus to counteract the economic downturn that has followed the collapse of the housing bubble. Yet the problem of how to restart and sustain robust growth goes well beyond short-term stimulus. The nation needs a firm foundation for long-term growth. But as of yet, there has been no serious public debate about how to create one. At best, there has been a rehash of 1990s debates about whether tax cuts or lower federal budget deficits are the better way to increase saving and (it is often assumed) stimulate growth.
A growing number of economists have come to see that innovation—not more saving—is the key to sustained long-term economic growth. Some economists have found that R&D accounts for nearly half of U.S. economic growth, and that R&D’s rate of return to the United States as a whole is as high as 30%. But R&D is not all there is to innovation. Properly conceived, innovation encompasses new products, new processes, and new ways of organizing production, along with the diffusion of new products, technologies, and organizational forms throughout the economy to firms and even entire industries that are not making effective use of leading technologies or organizational practices. Innovation is fundamentally about applying new ideas in organizations (businesses, nonprofits, and governments), not just about creating those ideas.
Innovation has returned to the federal policy agenda, most recently in the form of the America COMPETES Act signed into law in 2007. That law, unfortunately not yet fully funded, provides for much-needed increases in federal support for research and science and engineering education—key inputs into the process of innovation. But it does not go far enough. It does little to promote the demand for those inputs or to organize them in ways that lead to the commercial application of new ideas. More engineers and more R&D funding do not automatically create more innovation or, particularly, more innovation in the United States. In the mid-20th century, the nation could largely rely on leading firms to create research breakthroughs and turn them into new products, leaving the tasks of funding basic research and scientific and technological education to government. But it can no longer do so. Moreover, in the previous era, when the United States was the dominant technology-based economy, both old and new industries were domestic (for example, U.S. semiconductor firms replaced U.S. vacuum-tube firms). But a flat world means that more potential first movers will come from an increasingly large pool of technology-based economies, and that shifts in the locus of global competitive advantage across technology life cycles will occur with increasing frequency.
As a result, it is time for the federal government to make innovation a central component of its economic policy, not just a part of technology or education policy. To do so, it should create a National Innovation Foundation (NIF), which would be funded by the federal government and whose sole responsibility would be to promote innovation.
Growing innovation challenge
Since the end of World War II, the United States has been the world leader in innovation and high-value-added production. But now other nations are posing a growing challenge to the U.S. innovation economy, and increasingly, services as well as goods are subject to international competition. Because the United States cannot and should not try to maintain its standard of living by competing with poorer countries through low wages and lax regulations, it will have to compete in two other ways: by specializing in innovation-based goods and services that are less cost-sensitive, and by increasing productivity sufficiently to offset the lower wages paid in countries such as India and China. Both strategies rely on innovation: the first on product innovation and the second on process and organizational innovation. These same strategies are essential for maintaining the U.S. competitive position relative to other economically advanced countries.
However, there is disturbing evidence that the nation’s innovation lead is slipping. Companies are increasingly shifting R&D overseas. Between 1998 and 2003, investment in R&D by U.S. majority-owned affiliates increased twice as fast overseas as it did at home (52 versus 26%). In the past decade, the share of U.S. corporate R&D sites in the United States declined to 52 from 59%, while the share in China and India increased to 18% from 8%. The United States’ shares of worldwide total domestic R&D spending, new U.S. patents, scientific publications and researchers, and bachelor’s and new doctoral degrees in science and engineering all fell between the mid-1980s and the beginning of this century. The United States ranks only 14th among countries for which the National Science Foundation (NSF) tracks the number of science and engineering articles per million inhabitants. The United States ranks only seventh among countries in the Organization for Economic Co-operation and Development in the percentage of its gross domestic product (GDP) that is devoted to R&D expenditures, behind Sweden, Finland, Japan, South Korea, Switzerland, and Iceland, and barely ahead of Germany and Denmark.
Why has the United States’ innovation lead been slipping? One reason is that the process by which R&D is financed and performed has changed. During the first four decades after World War II, large firms played a leading role in funding and carrying out all stages of the R&D process. Companies such as AT&T and Xerox did a substantial amount of generic technology research, as well as applied R&D, in house. More recently, private funders of R&D have become more risk-averse and less willing to fund long-term projects. U.S. corporations, while investing more in R&D in this country overall, have shifted the mix of that spending toward development and away from more risky, longer-term basic and applied research. Similarly, venture capitalists, who have become a leading source of funding for cutting-edge science-based small firms, have shifted their funding away from startups and early-stage companies, which are riskier investments than later-stage companies, and have even begun to shift funding to other nations. In addition, as short-term competitive pressures make it difficult for even the largest firms to support basic research and even much applied research, firms are relying more on university-based research and industry/university collaborations. Yet the divergent needs of firms and universities can hinder the coordination of R&D between these two types of institutions.
Problems with the diffusion of innovation have also become more important. Outside of relatively new science-based industries such as information technology and biotechnology, many industries, including construction and health care, lag in adopting more productive technologies. Regardless of their industry, many small and medium-sized firms lag in adopting technologies that leading firms have used for decades. This is perhaps most visible in the manufacture of durable goods, where small and medium-sized suppliers have lagged behind their larger customers in adopting waste-reducing lean production techniques. Although smaller firms have long been late adopters of new technologies, the problem was less serious in an era when large firms manufactured most of their own components and designed products for their outside suppliers. With today’s more elaborate supply chains, technological lag by suppliers is a more serious problem for the U.S. economy.
Finally, geographic clustering—the tendency of firms in the same or related industries to locate near one another—enables firms to take advantage of common resources, such as a workforce trained in particular skills, technical institutes, and a common supplier base. Clustering also facilitates better labor-market matching and facilitates the sharing of knowledge, thereby promoting the creation and diffusion of innovation. It exists in such diverse industries and locations as information technology in Silicon Valley, autos in Detroit, and insurance in Hartford, Connecticut. Evidence suggests that geographic clustering may have become more important for productivity growth during the past three decades. Yet because the benefits of clustering spill over beyond the boundaries of the firm, market forces produce less geographic clustering than society needs, and firms have little incentive to collaborate to meet needs that they share in common, such as worker training to support new technologies or ways of organizing work.
Federal innovation policy should respond directly to these innovation challenges. It should help fill the financing gaps in the private R&D process, particularly for higher-risk, longer-term, and more-generic research. It should spur collaboration between firms and research institutions such as universities, colleges, and national laboratories. It should help speed the diffusion of the most productive technologies and business practices by subsidizing the training of workers and managers in the use of those technologies and practices and by giving firms (especially small and medium-sized ones) the information and assistance they need to adopt them. There is also a growing need for government to encourage the development of industry clusters, as governments such as that of China have deliberately done as a way of reducing costs and improving productivity. The federal government should do all of these things in an integrated way, taking advantage of complementarities that exist among activities to create an integrated, robust innovation policy that can make a real contribution to long-term economic growth.
Current federal policy does little to address the nation’s innovation challenges. Most fundamentally, the federal government does not have an innovation policy. It has a basic science policy (supporting basic scientific research and science and technology education). It has an intellectual property policy, carried out through the Patent and Trademark Office. It has agencies and programs that promote innovation in specific domains as a byproduct of agencies and missions that are directed at other goals (for example, national defense, small business assistance, and energy production). It even has a few small programs that are designed to promote various types of commercial innovation. But this activity does not add up to an innovation policy. Innovation-related programs are fragmented and diffuse, scattered throughout numerous cabinet departments, including Commerce, Labor, Energy, and Defense, and throughout a host of independent agencies, such as NSF and the Small Business Administration. There is no federal agency or organization that has the promotion of innovation as its sole mission. As a result, it is not surprising that innovation is rarely thought of as a component of national economic policy.
Existing federal innovation efforts are underfunded as compared with efforts in other economically advanced nations. In fiscal year 2006, the U.S. government spent at most a total of $2.7 billion, or 0.02% of GDP, on the principal programs and agencies that are most centrally concerned with commercial innovation. If the federal government were to invest the same share of GDP in these programs and agencies as many other nations do in comparable organizations, it would have to invest considerably more: $34 billion per year to match Finland, $9 billion to match Sweden, $5.4 billion to match Japan, and $3.6 billion to match South Korea. Some U.S. programs, particularly the Technology Innovation Program and its predecessor, the Advanced Technology Program, and the Manufacturing Extension Partnership Program, have had their budgets drastically reduced (from already low levels), largely because the current administration has tried to have them abolished.
The only federal support for technology diffusion comes through the Manufacturing Extension Partnership Program, an outstanding but underfunded program whose existence has been threatened during the current administration. But although federal support for technology diffusion in manufacturing is meager, in services it is almost nonexistent, as is federal support for innovation in services generally. Yet services, which include everything other than agriculture, mining, and manufacturing, account for four out of every five civilian jobs.
Likewise, there is little federal support for regional industry clusters. In fact, few federal innovation promotion programs engage in any way with state or regional efforts to spur innovation. Yet state governments and regional partnerships of businesses, educational institutions, and other actors involved in innovation have developed many effective efforts to promote innovation. These efforts are relatively small-scale, amounting to only $1.9 billion annually, and they are, understandably, undertaken with only the interests of states and regions, not the interests of the nation as a whole, in mind. Federal support, with appropriate federal incentives, could remedy both of these defects.
Charting a new federal approach
Establishing a NIF should lie at the heart of federal efforts. NIF would be a nimble, lean, and collaborative entity devoted to supporting firms and other organizations in their innovative activities. The goal of NIF would be straightforward: to help firms in the nonfarm U.S. economy become more innovative and competitive. It would achieve this goal by assisting firms with such activities as joint industry/university research partnerships, technology transfer from labora tories to businesses, technology-based entrepreneurship, industrial modernization through the adoption of best-practice technologies and business procedures, and incumbent worker training. By making innovation its mission, funding it adequately, and focusing on the full range of firms’ innovation needs, NIF would be a natural next step in advancing the innovation agenda that Congress put in place when it passed the America COMPETES Act.
Because flexibility should be one of NIF’s key characteristics, it would be counterproductive now to overspecify NIF’s operational details. NIF would determine how best to organize its activities; it would not be locked into a particular programmatic structure. Nonetheless, there are some core functions that NIF should undertake.
- Catalyze industry/university research partnerships through national-sector research grants. To begin, NIF would offer competitive grants to national industry consortia to conduct research at universities—something the government does too little of now. These grants would enable federal R&D policy to break free of the dominant but unproductive debate over science and technology policy, which has tended to pit people who argue that the federal government should fund industry to conduct generic precompetitive R&D against those who maintain that money should be spent on curiosity-directed basic research at universities. This is a false dichotomy. There is no reason why some share of university basic research cannot be oriented toward problems and technical areas that are more likely to have economic or social payoffs for the nation. Science analyst Donald Stokes has described three kinds of research: purely basic research (work inspired by the quest for understanding, not by potential use); purely applied research (work motivated only by potential use); and strategic research (work inspired by both potential use and fundamental understanding). Moreover, there is widespread recognition in the research community that drawing a hard line between basic and applied research no longer makes sense. One way to improve the link between economic goals and scientific research is to encourage the formation of industry research alliances that fund collaborative research, often at universities.
Currently, the federal government supports a few sector-based research programs, but they are the exception rather than the rule. As a result, a key activity of NIF would be to fund sector-based research initiatives. NIF would offer competitive Industry Research Alliance Challenge Grants to match funding from consortia of businesses, businesses and universities, or businesses and national labs. These grants would resemble the National Institute of Standards and Technology’s (NIST’s) Technology Innovation Partnership programs and NSF’s innovation programs (Partnerships for Innovation, Industry-University Cooperative Research Centers, and Engineering Research Centers). However, NIF grants would have an even greater focus on broad sectoral consortia and would allow large firms as well as small and mid-sized ones to participate. Moreover, like the NIST and NSF innovation programs, NIF’s work in this area would be industry-led, with industry coming to NIF with proposals.
To be eligible for NIF matching funding, firms would have to form an industry-led research consortium of at least five firms, agree to develop a mid-term (3- to 10-year) technology roadmap that charts out generic science and technology needs that the firms share, and provide at least a dollar-for-dollar match of federal funds.
This initiative would increase the share of federally funded university and laboratory research that is commercially relevant. In so doing, it would better adjust the balance between curiosity-directed research and research more directly related to societal needs.
NIF would also support a productivity enhancement research fund to support research into automation, technology-enabled remote service delivery, quality improvement, and other methods of improving productivity. Automation (robotics, machine vision, expert systems, voice recognition, and the like) is a key to boosting productivity in both manufacturing and services. Technology-enabled remote service delivery (for example, home health monitoring, remote diagnosis, and perhaps even remote surgery) has considerable potential to improve productivity in health care and other personal service industries. A key function of NIF would be to fund research at universities or joint business/university projects focused on increasing the efficiency of automated manufacturing or service processes. NIF would support early-stage research into processes with broad applications to a range of industries, not late-stage research focused on particular companies. It also would fund a service-sector science initiative to conduct research into productivity and innovation in the nearly 80% of the economy that is made up of service industries.
- Expand regional innovation promotion through state-level grants to fund activities such as technology commercialization and entrepreneurial support. The design of a more robust federal innovation policy must consider, respect, and complement the plethora of energetic state and local initiatives now under way. Although the federal government has taken only very limited steps to promote innovation, state governments and state- and metropolitan-level organizations have done much more. They engage in a variety of different technology-based economic development activities to help spur economic growth. They spur the development of cutting-edge science-based industries by boosting research funding. Moreover, they try to ensure that research is commercialized and good jobs created in both cutting-edge science-based industries and industries engaging in related diversification. States have established initiatives to help firms commercialize research into new business opportunities. They also promote upgrading and project-based innovation by helping existing firms become more competitive.
Although already impressive, these entities could do even more, and their current efforts could be made more effective. Because the benefits of innovation often cross state borders or take at least a few years to result in direct economic benefits, or both, state elected officials have less incentive to invest in technology-based economic development activities than in other types of activities, such as industrial recruitment, that lead to immediate benefits in the state.
Moreover, any effective national innovation initiative will need to find a way to assist the tens of thousands of innovation-focused small and mid-sized firms as well as larger firms that have specific regionally based innovation needs that they cannot meet on their own. Unlike small nations, the United States is too big for the federal government to play an effective direct role in helping these firms. State and local governments and regional economic development organizations are best positioned do this.
As a result, without assistance from the federal government, states will invest less in these kinds of activities than is in the national interest. NIF would compensate for this political failure by offering state Innovation-Based Economic Development (IBED) Partnership Grants to help states expand their innovation-promotion activities. The state IBED grants would replace part of the grantmaking that the NIST and NSF innovation programs currently perform but would operate exclusively through the states.
To be eligible for NIF funding, states would need to provide at least two dollars in actual funding for every NIF dollar they receive. Rotating panels of IBED experts would review proposals. NIF staff would also work in close partnership with states to help ensure that their efforts are effective and in the national as well as the state interest.
- Encourage technology adoption by assisting small and mid-sized firms in implementing best-practice processes and organizational forms that they do not currently use. Although NIF’s national-sector grants and state IBED grants would largely support new-to-the-world, sometimes radical product and process innovation, its technology diffusion work would focus more on the diffusion of existing processes and organizational forms to firms (mostly small and mid-sized) that do not currently use them. This effort would incorporate and build on NIST’s Manufacturing Extension Partnership (MEP) program, the only federal program whose primary purpose is to promote technology diffusion among such firms. NIF effort would follow the MEP model of a federal/state partnership. One or more technology diffusion centers would be located in each state. Like existing MEP centers, the centers could be operated by state or private organizations. States would submit proposals to NIF for the operation of these centers, and NIF would evaluate the centers periodically. Some specific changes in the current MEP program would enable NIF to serve as a more comprehensive and more effective promoter of technology diffusion for both manufacturing and service industries. NIF would expand the scope of the current MEP beyond its current emphasis on applying waste-reducing, quality-improving lean production techniques to the direct production of manufactured goods. It would do so by helping improve productivity in some service activities where lean production could be applied.
In addition to supporting efforts that assist firms directly, NIF would analyze opportunities and challenges regarding technological, service-delivery, and organizational innovation in service industries, such as health care, construction, residential real estate, financial, and transportation services. It also might recommend steps that federal and state governments could take to help spur innovation, including the digital transformation of entire sectors through the widespread use of information technology and e-business processes. Such steps might include revising procurement practices, modifying regulations, and helping spur standards development.
To guide its own work and provide firms and government agencies with the information they need to promote innovation, NIF would create methods of measuring innovative activity and carry out research on innovation. It would be the primary entity for conceptualizing how innovation should be measured and the primary advocate within the federal government for measuring innovation. It would help the major federal statistical agencies (the Census Bureau, Bureau of Economic Analysis, and Bureau of Labor Statistics) and NSF develop operational measures of innovation that can be included in new or existing economic data sources.
NIF would also work with other agencies to improve the measurement of productivity and innovation, particularly in better measuring output in the service sector; total factor productivity (the most comprehensive measure of productivity, which accounts for capital, materials, energy, and purchased services, in addition to labor, as productive inputs); and better bottom-up estimates of gross product and productivity for counties and metropolitan areas.
In addition, NIF would be the federal government’s major advocate for innovation and innovation policy. As a key step, it would produce an annual Innovation Report, akin to the annual Economic Report of the President. More generally, NIF’s advocacy role in support of innovation would resemble the Small Business Administration’s role as a champion for small business. NIF would seek input into other agencies’ decisions on programs that are likely to affect innovation. However, unlike the Small Business Administration, NIF would not have any authority to intervene in other agencies’ decisions.
Compelling need, but obstacles
In the current fiscal climate, it will be difficult for the federal government to launch major new investment initiatives, especially because strong political forces on either side of the aisle oppose raising taxes or cutting other spending. Nevertheless, the compelling need to boost innovation and productivity merits a substantial investment in NIF. The federal government should fund it at an initial level of $1 billion per year, but approximately 40% of this funding would come from consolidating existing innovation programs and their budget authority into NIF. (Rolled up would be the NIST and NSF innovation programs, as well as the Department of Labor’s WIRED program. Federal expenditures on all of the programs that NIF would replace or incorporate total $344 million. In addition, the America COMPETES Act provides a total of about $88 million more in 2010 than in 2006 for the programs that will be folded in. Therefore, current and already-planned expenditures on the programs whose work would be included in NIF total $432 million.) After several years, NIF could easily be ramped up to a budget of $2 billion, a level that would make its budget approximately one-third the size of NSF’s. In addition, because of its strong leveraging requirements from the private sector and state governments, NIF would indirectly be responsible for ensuring that states and firms spent at least one dollar on innovation for every dollar that NIF spent.
NIF could be organized in several ways. It could be organized as part of the Department of Commerce, as a government-related nonprofit organization, as an independent federal agency, or as an arm of the Office of the President. But whatever way it is organized, it should remain relatively lean, employing a staff of approximately 250 individuals. It should recruit the best practitioners and researchers whose expertise overlaps in the areas of productivity, technology, business organization and strategy, regional economic development, and (to a lesser extent) trade. Like NSF, NIF would be set up to allow some staff members to be rotated into the agency for limited terms from outside of government and to allow some permanent NIF staff members to go on leave for limited terms to work for private employers.
Already there is legislation in the Senate to create an NIF-like organization. The National Innovation Act, introduced by Senators Hillary Clinton (D-NY) and Susan Collins (R-ME), would create a National Innovation Council, housed in the Office of the President and consolidating the government’s primary innovation programs.
Now more than ever, the U.S. standard of living depends on innovation. To be sure, companies are the engines of innovation, and the United States has an outstanding market environment to fuel those engines. Yet firms and markets do not operate in a vacuum. By themselves they do not produce the level of innovation and productivity that a perfectly functioning market would. Even indirect public support of innovation in the form of basic research funding, R&D tax credits, and a strong patenting system—important as it is—is not enough to remedy the market failures from which the nation’s innovation process suffers. At a time when the United States’ historic lead in innovation is shrinking, when more and more high-productivity industries are in play globally, and when other nations are using explicit public policies to foster innovation, the nation cannot afford to remain complacent. Relying solely on firms acting on their own will increasingly cause the United States to lose out in the global competition for high-value-added technology and knowledge-intensive production.
The proposed NIF would build on the few federal programs that already succeed in promoting innovation and borrow the best public policy ideas from other nations to spur innovation in the United States. It would do so through a combination of grants, technical assistance, information provision, and advocacy. It would address the major flaws that currently plague federal innovation policy and provide the United States with a state-of-the-art initiative for extending its increasingly critical innovation prowess.
Yet NIF would neither run a centrally directed industrial policy nor give out “corporate welfare.” Rather than taking the view that some industries are more important than others, NIF is based on the idea that innovation and productivity growth can happen in any industry and that the nation benefits regardless of the industry in which they occur. It would work cooperatively with individual firms, business and business/university consortia, and state governments to foster innovation that would benefit the nation but would not otherwise occur. In a world of growing geographic competition for innovative activities, these economic and political actors are already making choices among industries and technologies to serve their own interests. NIF would give them the resources they need to make those choices for the benefit of the nation as a whole.
Without the direct federal spur to innovation that NIF would offer, productivity growth will be slower. Wages will not rise as rapidly. U.S. companies will introduce fewer new products and services. Other nations have realized this and established highly effective national innovation-promotion agencies. It is time for the United States to do the same. By combining the nation’s world-class market environment with a world-class public policy environment, the United States can remain the world’s innovation leader in the 21st century.
Robert Atkinson and Howard Wial, Boosting Productivity, Innovation, and Growth through a National Innovation Foundation (Washington, DC: Brookings Institution and Information Technology and Innovation Foundation, 2008), available online at http://www.brookings.edu/~/media/
Files/rc/reports/2008/04_federal_role_atkinson_wial/NIF%20Report.pdf or http://www.itif.org/files/NIF.pdf.
Robert Atkinson ([email protected]) is president of the Information Technology and Innovation Foundation in Washington, DC. Howard Wial is an economist at the Brookings Institution in Washington, DC.