Getting to the "How" of Innovation
Innovation—everyone is for it. In fact, one of the most firmly established principles in economics is the importance of innovation in driving growth and prosperity. The ability to produce knowledge and use it commercially is the main determinant of whether—and how well—nations and regions thrive.
Yet how to drive innovation remains elusive. We really don’t know much about the production, diffusion, and application of knowledge. At the same time, economics is pretty good at providing some rules of thumb—many of which, unfortunately, tend to be ignored. With that in mind, Forefront asked some economic experts for their take on how best to capture the fruits of innovation.
Why innovation is important for growth
Let’s start with the question of how we know that innovation powers economic growth. This key insight came from Nobel economist Robert Solow, who found that standard economic models failed to account for almost half of the economy’s output. There was something beyond the accumulation of physical capital—a residual that most people now ascribe to technological innovation. We at the Cleveland Fed have defined it as “finding a better way to get things done with what you already have.”
Over the years, economists have refined Solow’s growth theory to clarify the role of technological innovation. Real-world evidence on the importance of innovation is also mounting. Looking over a 75-year period, Cleveland Fed economists and their co-authors have found that patents were the single largest factor explaining income differences across U.S. states. High levels of patents in states like Connecticut, Massachusetts, and New Jersey explain most of their income growth going back to 1939. This implies strong innovation and knowledge-building in places with vigorous income growth. Overall, our economists concluded, innovation is the most important factor in determining people’s welfare.
Why we think policies can help innovation
Once we understand the importance of innovation, the next logical question is whether public policy can help raise innovation levels. Our nation’s go-to policies include tax credits for research and development, robust patent protections for innovating firms, grants and subsidies to entrepreneurs and academic researchers, and so on. Are they working?
The litmus test for most economists is whether the free market would drive innovation more efficiently than government policy.
According to Andy Atkeson, an economist at the University of California, Los Angeles, and a visiting scholar at the Minneapolis Fed, the litmus test for most economists is whether the free market would drive innovation more efficiently than government policy. After all, the main reason firms might want to innovate is to boost their profits. If firms are going to innovate anyway, why should the government step in and risk distorting the market?
The answer is that firms may not invest in innovation as much as the rest of us would like. That’s because, as Atkeson notes, the benefits of innovation don’t flow just to the innovators. Society gains as well.
Ideas generally can’t be stopped from spilling over into the wider world, and firms learn from each other. In fact, it’s the social element of innovation that makes it beneficial to the wider economy. To take a classic example: Synthetic fiber was invented in the chemical industry, but the knowledge that created it spilled over to the textile industry.
Atkeson offers a more recent example: “Apple’s investment in R&D has been amply rewarded in the company’s profits. But they can’t patent the market categories they created, like the smartphone. They showed people what it should look like, and now lots of others are copying it. That means they didn’t get the full return from their investment. In the end, there is a big intellectual component from innovation that’s not captured by the innovator.”
The very existence of knowledge spillovers suggests that we might not be getting as much from them as we could. All the same, economists strongly suspect that companies will invest in research and development only to the extent that they can profit from it. The part that spills over is extra, not something they factor into the equation.
You can see the stagnation in productivity data. Since 1973, the rate of productivity growth has fallen below the trend in the post–WWII period.
But from a societal standpoint, we want as much of the innovation spillover as possible. That’s where policy can make a difference.
Why innovation is so important right now
The nation’s struggle to claw back from a severe recession would be motivation enough to think about ways to boost innovation. But another motivation comes from the long-run trend of productivity growth—in the wrong direction. As economist Tyler Cowen explains in his 2011 e-book, The Great Stagnation, America has enjoyed a long period of picking low-hanging fruit, in the form of new technology to fuel a fast-growing labor force. But since the 1970s, the lowest branches have become increasingly bare. “That’s a sign that the pace of technological development has been slowing down,” Cowen notes. “It’s not that something specific caused the slowdown, but rather we started to exhaust the benefits of our previous momentum without renewing them.”
You can see the stagnation in productivity data. Since 1973, the rate of productivity growth has fallen below the trend in the post–WWII period.
The starting point
Innovation is not a linear process; it’s messy and iterative. Ideas bop around until they are fine-tuned into something with real market value. Failure is a key part of the process.
For that reason, there seem to be many potential entry points where policies can affect innovation. These begin with basic research—the most fundamental stage of innovation, where ideas are beginning to germinate and most likely have no specific commercial use in mind. Then there is applied research, in which commercial entities transform ideas into prototypes and processes. Also crucial is funding—entrepreneurs need financial backing to get their ideas into development and production. Intellectual property rights and associated patent protection policies are also important. And of course, educational institutions perform a number of roles, from idea and business generators to workforce preparation.
Scott Shane, BusinessWeek blogger and economist at Case Western Reserve University, suggested where innovation policy could be most useful.
The starting point, Shane believes, is weighing the costs and benefits. Often, he says, it’s not clear that policies aimed directly at firms would create better outcomes than the private market would on its own. In other cases, it seems that the government provides windfall subsidies beyond the point where companies would have invested otherwise.
“It’s very hard to link what economic developers are saying should be done to some specific evidence of what would work,” Shane explains. “Why should we advocate policies to create innovations if we don’t even know whether they work?”
Innovation “maybe nots”
With that yardstick in mind, here are the innovation policy staples that some economists have identified as needing careful cost–benefit analysis.
Incubators are a favorite “innovation-friendly” government program. Incubators typically feature a building whose small-business tenants share space, computer and office equipment, and onsite counsel. Putting innovators in the same place, it is hoped, will encourage them to swap ideas, contacts, and funding sources. The same principle is behind the notion of agglomeration—the idea that people and firms working within the same region increase their ability to share and profit from knowledge.
The National Business Incubation Association counts about 1,200 business incubators across the country. Many are supported by local governments and universities.
The way Shane sees it, you most often end up with promising young firms that choose to operate in the incubator because they can get free or inexpensive space. The clustering and specialization that firms can experience as part of an incubator can help their performance. He wonders, “If the private sector benefits from providing space for new firms, why doesn’t it do it?”
Direct financial subsidies to companies are premised on the theory that young firms in particular don’t have the resources to engage in serious research and development on unproven ideas, and they can’t attract investors in their very early stages. Subsidies can take the form of grants or loans, sometimes funneled through public venture-capital funds. Local policymakers often use multipliers to justify these investments, arguing that every $1 invested turns into a certain number of new jobs, for example.
The problem with the direct subsidies approach, several economists contend, is that it may distort market outcomes. The government could get itself into a situation of picking winners and losers, and prematurely at that.
Private markets are much better at sorting out winners from losers than anyone trying to foresee which idea may eventually gain traction.
Patent protection, counterintuitively, may sometimes stifle innovation. This is the argument made by George Mason University’s Alex Tabarrok, author of Launching the Innovation Renaissance (2011). Patents are supposed to reward innovators by handing them a monopoly on their product for a certain number of years. Profit-seeking investors will be more likely to put their money into projects they think will be insulated from competition, the thinking goes. Patent protection certainly makes sense in situations where the costs of innovation greatly exceed the costs of imitation—like the pharmaceutical industry.
But Tabarrok sees too many situations in which patent protection is overused. The problem is particularly acute with innovations that produce intermediate goods, that is, goods that are used to produce other goods and innovations. Tabarrok’s example is the “oncomouse,” a genetically engineered mouse used in biomedical research. For years, he says, Harvard and DuPont wielded virtual control over the oncomouse, even though others could have greatly enhanced biomedical research with access to it. The upshot was that strong patent protections increased the cost of building on previous research and thus discouraged further innovation.
“The patent system is now being used as a weapon for innovative firms to attack other innovative firms and slow them down,” Tabarrok said.
The tax code is a critically important consideration in innovation policy. To Atkeson, the tax code is in fact the most important consideration because it bears so directly on whether entrepreneurs choose to enter the market.
Consider the offsetting effects of the U.S. tax code on research and development tax credits. R&D credits were established in 1981 as an incremental subsidy—a tax credit to supplement a defined base amount of spending. In recent years, the value of the subsidy has ranged between $4 billion and $8 billion a year. A 2009 Government Accountability Office report found that the R&D tax credit reduced the business costs of new research by about 7 percent. The GAO also noted a number of shortcomings in the R&D tax-credit system, including disparities between the amount of the credit and its actual incentive effects.
But there is a wider issue to consider: Any innovation-investment payoffs will be counterbalanced by the taxes businesses will pay on their profits. Atkeson explains it this way:
When you’re considering creating a new firm, you tally up all the revenue you think you can make versus the expenses. That includes projections of a subsidy for innovation that you might get, as well as payment of corporate taxes on profits. As anyone who evaluates business plans knows, if the present value of your investment is positive, you should enter. If not, you shouldn’t. So, from a purely business plan perspective, a subsidy is dollars coming in and taxes are dollars going out. I would think those two policies would just cancel each other out. But in fact, our research has found that they don’t cancel each other out, because the corporate tax is bigger than the subsidies!
So the discouragement of entry that we have with the corporate tax is much larger than the encouragement of innovation with the R&D tax credit.
“If policies discourage entry, that’s not good,” Atkeson said. “The rule should be to evaluate the universe of policies based on whether they affect an entrepreneur’s decision to enter.”
Just as economists have reservations about certain policies, they are enthusiastic about others. These include:
The discouragement of entry that we have with the corporate tax is much larger than the encouragement of innovation with the R&D tax credit.
Investment in basic research. The federal government is the main source of funding for basic research, which is mostly conducted at academic institutions. Meanwhile, according to the National Science Board, only about 3.8 percent of industry-sponsored R&D can be classified as basic. The federal government’s annual contribution is about $37 billion of the $62 billion total spent on basic research in the United States.
Experts agree that when it comes to spurring long-term innovation, the federal government should do much more investing in basic research, which is the lifeblood of innovation. Basic research increases the technologies that lead to new industries and new products in the future.
The internet is Exhibit A. Without coordinated effort and investment by federal agencies in the 1960s and 1970s, the internet infrastructure never would have been built. Most of these dollars went directly to universities and research institutions, well before it was clear that the “packet switching” technology dreamt up at MIT would blossom into the internet of today.
“The best way policymakers can fund basic research is through government labs or to support academic research,” Shane contends.
Improve education. To bring ideas from the lab to the marketplace requires more than a single genius. But in a country where one of every four American men doesn’t even graduate high school, access to that help is a challenge.
Tabarrok advocates a range of efforts to improve educational outcomes that drive innovation. These include rewarding good teachers, creating more vocational educational programs and apprenticeships, and encouraging more students to go into math and science fields.
A body of other research points to early childhood education as a potential source of fuel for innovation. As Art Rolnick, co-director for the Human Capital Research Collaborative at the University of Minnesota, puts it, “Early childhood education is economic development, and the research shows it’s economic development with a high public return—very high.”
Let high-skilled immigrants work here. Another way to improve the labor force is to open our doors more fully to high-skilled immigrants. As Tabarrok argues, the United States has a workforce of 150 million but annually allows only 120,000 employment visas. And each country has a limit on the number of immigrants allowed each year, regardless of the needed skills.
We ought to grab up smart people who want to come here, and do it now when they still want to come.
Tabarrok describes this as low-hanging fruit with an expiration date. Wait too long to change policy, and eventually conditions will improve in the immigrants’ home countries to the point where they have little incentive to emigrate. “We ought to grab up smart people who want to come here, and do it now when they still want to come.”
Ken Simonson, chief economist of the Associated General Contractors of America, agrees, saying current U.S. immigration policy sends a mixed message: “We welcome students from abroad into our science, engineering, and technology programs, and then we say, ‘Sorry, we don’t want you working here.’ That’s just at the point when they could be our innovators!”
Don’t forget the private sector
Beyond these policy recommendations, where does the private sector fit in? Eugene Fitzgerald, an engineering professor at MIT and formerly a scientist at AT&T Bell Labs, brings a perspective from the trenches. He invented something called “strained silicon electronics,” a way to improve the performance of integrated circuits that was a huge technological advancement.
The interaction between corporate and public America was central to this process, in Fitzgerald’s telling. Back in the 1950s and 1960s, he says, the back-and-forth between corporate labs, government labs, and university researchers was robust. Ideas were swapped and knowledge spread to benefit end users.
Today, large corporate labs have largely disappeared. They went away, Fitzgerald says, because global competitors sprouted up to take on the virtual monopoly firms in the United States. For example, Kodak suddenly had to deal with Fuji, and pressure shifted to maintaining short-term profits. Before, big U.S. firms could wait a decade for a return on their research investments, but that’s too long in today’s hypercompetitive environment.
With mainly government and university research left to carry the load, the crucial feedback loop with the private sector was severed, and innovation became less efficient. Fitzgerald cites this trend to explain the drop-off in productivity growth starting in the 1970s.
Fitzgerald sees the absence of corporations from the innovation pipeline as contributing to an “innovation gap” that slows the overall pace of innovation. The innovation pipeline may currently produce a lot of research and patents, he says, but not a whole lot of economic growth.
Figuring out the mechanism for bridging the gap is another matter. Theoretically, the R&D tax credit should induce long-term investment in innovation. Fitzgerald says he is working on an index that measures “innovation capacity” in a company. The IRS could eventually use this metric to confirm tax credits are being properly channeled to long-term investments.
“There is no way the free market alone can get corporations to invest 15 years out,” Fitzgerald adds. “You need long-term government funding. They’re the only ones who can recover their investment over the long term, because they get it back in tax revenues and growth.”
The waiting is the hardest part
The final ingredient in spurring innovation in America may be patience. All of these efforts will take time before any return is evident. In today’s results-now world, waiting a decade for the payoff seems like an eternity. Politically, a multipronged approach to increasing innovation would probably have to survive at least two presidential administrations and several more Congresses.
The challenge is to create linkages between our long-term innovation goals and the short-term needs of the institutions that play a part in the innovation process. It’s a lot simpler to articulate that challenge than to actually address it.
Fortunately, the stakes are high enough to give policymakers plenty of incentive to make it happen. The future of innovation depends on their determination.