The Patent System: Mend, Don’t End

Patents provide inventors with stronger incentives to search for and commercialize new ideas than would be otherwise. As long as the patent system can insure that others can legally develop and offer competing products using a different method, the system will spur innovation.  The present system fails to completely reach that standard, but, even its flawed form, succeeds enough to warrant its existence.

In our recent report on the role of invention in regional economic growth—which uses patents as a proxy for inventive activity—we conclude that metropolitan areas with more patents attain higher worker productivity. Though not the focus of the report, we highlight the rise of litigation perpetuated by patent monetization entities (or “trolls” in the vernacular) and offer a policy response that could curtail abusive legislation. Following scholar Mark Lemley, we also suggest that the legal analysis of software patents needs to be clarified so that broad “functional” patents are not allowed.

Others would go so far as to abolish the patent system, and they have champions in economists Michael Boldrin and David Levine. Their most recent article published in the Journal of Economic Perspective, “The Case against Patents,” raises many worthwhile concerns, but I am not persuaded by their central argument that the patent system is deterring innovation compared to what would be the case if patents were abolished.

Boldrin and Levine argue that there is no empirical relationship between patents and productivity. Instead they argue that competition drives productivity growth and innovation, and patents block competition by design. The first claim can be tested empirically. The second is a matter of interpretation.

Patents increase the wages of inventors and the value of firms, and my analysis suggests a clear aggregate connection: From 2011 to 1977, high-tech industries—an aggregation of R&D and patent intensive sectors—added $1.2 trillion more in measured GDP to the U.S. economy—making a larger contribution than any single industry or the entire health care sector over that period. With private-funded R&D per capita steadily increasing, the high-tech sector has delivered 5 percent annual productivity growth since 1977, compared to just 1 percent for the total U.S. economy.

Of course, measured productivity growth has also been high in a few highly regulated non-competitive industries with huge barriers to entry—like petroleum refining and banking, but that has little to do with innovation. In any case, research by economist Erik Brynjolfsson and others finds that IT investments by companies outside the tech sector (like banking) have raised their productivity growth.

Without intellectual property protection, it is doubtful that high-tech firms would have found this level of R&D worthwhile. Studies that compare patent regimes across countries generally conclude that patent protection is growth enhancing, as a report from scholars at the United Nations Industrial Development Organization summarizes. From another angle, Petra Moser concludes a new literature review article by arguing that strong patent systems deter growth relative to more informal methods to protect ideas, like trade secrets. Yet, in her earlier work, also based on 19th century trade shows, she concluded the opposite and showed the U.S. patent system shifted R&D to manufacturing, spurring long-run growth. Others have found that throughout U.S. history, patents have provided an important spark to innovation, and most of the great inventions in U.S. technology history have been patented.

Our report took another approach: By systematically comparing metropolitan areas within the same country, we were able to control for institutional and historic factors along with other causes of innovation. We found that patents are strongly correlated with productivity from 1980 to 2010, even controlling for high-tech workers. We cannot observe the counterfactual of regional productivity growth in the absence of patents, but this at least reinforces the underlying connection between patenting and regional productivity growth.

I believe Boldrin and Levine are correct that competition drives innovation. Yet, patenting is compatible with competition. Patents deter a potential competitor—under threat of litigation—from making the same product using the same method, but in principle, and often in practice, patents do not prohibit a competitor from making the same product using a different method. (To the extent that they do, they should be struck down in court or nullified by the patent office).

To see how competition can work with patenting look at smart phones, which embody an astonishing 250,000 patents according to one estimate. On Amazon, as of writing, one can choose between 2,759 varieties of unlocked smartphones across 22 manufacturers, and 189 different varieties that come with plans; prices are as low as one penny. Perhaps, there would be more smartphones without patents with better open-source software, but they couldn’t be any cheaper unless companies paid you to buy them, and many industry analysts argue that there the market is already too crowded to sustain the number of competitors. If patents impose a major barrier to entry into an industry segment, as some have found for software, then the most heavily patented segment of all should be dramatically less competitive.

Broader evidence can be found in our report. In recent decades, there has a drop in the share of patents granted to the top 100 or top 50 patent owners, and many companies have emerged near the top only recently. To illustrate the breadth of patent ownership, in 1976, 2,677 entities owned exactly one patent granted that year to a U.S. inventor; by 2011, that number had soared to 9,909. This dispersion is true even for software patents, where the top 50 and top 100 shares have fallen.

Even Boldrin and Levine acknowledge that patents are more common in competitive industries, as convincing empirical  research has shown. Yet, they do not seem to consider that intellectual property protection can spur entry and safeguard entrepreneurial development. This is the case for many small businesses that win federal SBIR grants, for example, and the patent-venture capital relationship is well documented. Such competition may be why prices have fallen the fastest in the most patent intensive tech industries.

There are reform points where everyone should agree: No one should be able to own a function or threaten companies with litigation over specious claims without cost; research collaboration should be more readily facilitated; and greater clarity of property rights should be an overarching goal. Abolishing intellectual property rights should not be.