A Tax Policy for Innovation

Where does innovation come from? The conventional wisdom is that it comes from corporate spending on research. That's why one of the biggest corporate tax breaks is the credit given to corporations for spending on research and experimentation (R&E). In December 2006, Congress voted to increase the subsidy from $7.3 billion dollars to an estimated $9.3 billion. And now, separate Senate and House bills are under consideration that would raise the credit again. These bills are popular with companies that spend heavily on research, such as technology companies and defense contractors; Hewlett-Packard and Microsoft are lobbying for the increase.

In our increasingly competitive global economy, innovation is absolutely critical to our continued success, and we all have an interest in a tax code that fosters innovation. The problem is that it's been extremely difficult to prove any connection between increased research and development spending and increased innovation. In 2005, a Booz Allen Hamilton study analyzed the world's 1,000 top spenders on research and development, and found no relationship between the amount spent and the usual measures of performance: sales growth, gross profit, operating profit, total shareholder return. The same study found no relationship between the number of patents issued to a company and its business results. To take just one high-profile example, in the last few years Microsoft has spent around $5 billion a year on research, but most software innovations have come from other companies.

The problem with the tax credit is that most of the benefit goes to a small number of large corporations, but research on innovation shows that the most radical innovations emerge from smaller organizations. Although 10,000 companies received a credit in 2004, over half of the $5.5 billion credit that year went to 100 companies. We should keep the R&E tax credit, but it should be revised based on what we know about how innovation works.

First of all, the credit should be revised so that more of the benefit flows to smaller, more nimble organizations--the entrepreneurial startups and university lab spin-offs that are often the source of new innovations.

Second, in the most innovative organizations, research and innovation activities are deeply woven into the fabric of every activity. This can make it hard to document for the IRS exactly which dollars were spent on "research" or "experimentation." The idea that innovation always originates in a separate lab is out of date; innovation today emerges from activities throughout the company, when scientists work side by side with engineers and plant workers. The credit should be revised so that it no longer forces companies to comply with this dated model of innovation.

Third, corporations that receive the benefit ought to give us all something back in return. When R&D results in a new innovative product, the company can, for example, receive a patent and retain the ownership of the new innovation. The ability to protect intellectual property is an important incentive for innovation. But when the tax credit was first introduced in 1981, intellectual property laws did not change. When the government provides the credit, we citizens are like investors in the company, and we should share in some of that value.

The best way to share that value is with new intellectual property laws that align with the way innovation today works. New research has shown that overall innovation grows when ideas are shared; that's because innovations are always the result of combinations of previous innovations, and innovations typically emerge from networks that involve multiple players. The problem is that many companies use patent protection to keep new competitors from introducing better products. The solution? If you take the tax credit, you should agree to reduce the length of time you retain legal right to the innovations that emerged from tax credit activities, and you should be willing to agree to license the new innovations to other businesses at a reasonable cost.

The government should provide incentives for innovation. But it doesn't make sense to do it simply by writing a bigger check; we should provide incentives that align with the way innovation works.

Keith Sawyer is a professor at Washington University. His latest book is GROUP GENIUS: THE CREATIVE POWER OF COLLABORATION (Basic Books).