With growth averaging about 2% annually, the recovery from the recession that ended in 2009 has experienced the slowest growth of any American economic recovery in almost a century. The critical factor behind slow economic development lies in declining productivity growth data. Productivity growth has been about 1% annually since 2010, with productivity growth declining to negative in 2016.
What is behind the productivity growth data? There appears to be a causal connection between productivity growth declines and a decline in business investment. One argument suggests that with weak demand, companies have little incentive to invest in businesses. Without more investment, productivity growth declines and, with it, aggregate economic growth.
Economists parse out the components of productivity growth into capital, labor and total factor productivity (TFP). TFP represents a cluster of factors under the umbrella of technology innovation. The levels of business investment in the main components of productivity have generally been stagnant since the end of the recession, and appear to be weakening. Although productivity growth increased in the aftermath of the recession as reduced aggregate employment increased efficiency for the remaining workers, the long-term trend shows that productivity growth began to slow before the recession, in the mid-2000s, suggesting a structural economic transformation.
Economists have a number of explanations of productivity growth declines in recent years. First, some argue that innovations have peaked in recent generations and that innovation is simply not robust enough to increase productivity. Second, some argue that we are in an era of secular stagnation driven by limited demand, which results in limited production. Third, some argue that the productivity data are mismeasured, particularly since industrial nations like the U.S. no longer maintain a substantial manufacturing industry. Fourth, some argue that there are lags in productivity data in which major productivity growth gains will result from recent innovations as the technology works its way through the economy. Finally, some argue that the concentration of wealth for the top Americans constrains investment since they have no compelling reason to invest.
While all economic views of productivity growth may have a grain of truth, I argue that all of these theories of productivity growth are fundamentally wrong. The economic evidence shows clear relationships between the trend of decline in productivity growth and the trend in declines of the U.S. patent system. As the patent system has been weakened in the last decade, there is less investment by start-ups in technology R&D. At the same time, the technology industry has become highly concentrated, with big tech incumbents dominating their industries. The combination of increasingly concentrated competitive configurations in the technology industry with the reduction in key rights embedded in patents, has resulted in substantially reduced incentives to invest in R&D innovations.
The reduction of patent rights is not new. In the 1970s, before the creation of the U.S. Court of Appeals for the Federal Circuit, patent law was weakened. During this period, productivity growth was also reduced along with reduced business investment. The tremendous growth of the venture capital industry during the 1980s and 1990s, spurred by an era of strong patents, in turn generated a technological revolution that catalyzed productivity growth.
There are two major forces that contribute to a reduction in patent rights. First, on the far right, big tech incumbents have an incentive to protect their monopoly profits by attacking patents since patents are the main tool to protect innovations by market entrants. Second, on the far left, progressives have attacked the property right in a patent in order to develop a system for public benefits. Progressives ignore ex ante costs of R&D and seek to mandate technology for the public interest. They see the patent system, and its private rights, as blocking a socialist agenda. Together, radicals from both the left and the right, have succeeded in dramatically transforming patent law. These changes began in the mid-2000s, about the time of the decline in productivity growth.
Today, the patent system is a shadow of its former self. Critical elements of patent law have been changed, including the examination system in the U.S. Patent and Trademark Office (PTO). The new system consists of a two-part examination approach that includes a very expensive and onerous reexamination of an issued patent. In addition to changes in examinations, patent remedies have been restricted. For instance, injunctions to stop others from using a patented invention are now very rare, thereby shifting the system to one of instituting liability rules rather than property rules. In addition, patent damages are generally reduced from a decade ago because patents are narrowly apportioned to a specific feature of a system, without considering qualitative weighting of specific innovations that provide a market advantage.
These changes in patent law have had the effect of actually encouraging infringement and eliminating the voluntary licensing market. Rather than license a patent, big tech incumbents ignore patents and refuse to deal with patent holders. This reverse hold out by incumbents forces patent holders to enforce patents in the courts.
One main effect of the changes in patent law is that transaction costs of patent examinations and patent enforcement have now increased dramatically. It is precisely these much higher transaction costs that are influencing the decisions by innovators not to invest in new technologies, particularly at the margins.
However, the degradation of rights in patents and the dramatic increase in transaction costs to enforce patents have had adverse effects. Since big tech incumbents can use others’ technologies without fear of an injunction, they have engaged in the practice of “efficient infringement” in which they infringe patents with impunity and then pay a licensing fee only if a patent holder meets the barriers of high transactions costs.
The patent system worked well from 1790 to about 2000. During this two hundred year period, the U.S. was the location for the industrial revolutions that propelled economic growth to historic levels. The core impetus for this massive industrialization was the incentives provided in the intellectual property clause of the U.S. Constitution, which embedded “for limited times” an “exclusive right” to an inventor’s discoveries. Patents are disclosures of inventions that enable others to learn about original work and provide the opportunity to build on it with other novel work. In this way, the patent system both encourages investment in hard technical problems and supplies economic progress.
Because of these high transaction costs for patent enforcement, only a few market entrants can enforce patents. Also, since the big tech incumbents are collectively refusing to deal with market entrants, the voluntary licensing market has receded, bringing nearly all matters to the courts. Without a voluntary licensing market, the time to obtain returns on investments has dramatically extended, thereby effectively blocking the innovation market. Since market entrants are not compensated significantly for their work, they do not have capital to invest in more R&D, so the system of continued invention is stopped.
For market entrants, high transaction costs for patent reexams and enforcement, combined with reduced licensing settlements, result in a squeeze on limited capital resources. Given these higher costs and reduced rewards, incentives to invest in risky research are diminished. This describes the process behind the persistent slowdown of investment into innovation for market entrants. These enhanced challenges in a weak patent regime are evident in historically low business starts.
Ironically, the large tech incumbents have enjoyed remarkable profits in recent years and have stored most of these profits as cash. The cash hoard of the top five technology companies was recorded at over $530B. In addition, the four highest capitalized corporations in the world are big tech incumbents. Yet, they have not substantially invested these record profits into more investment. This diminished investment can be explained because they do not need to invest in new technology if they can infringe others’ technologies and only pay a nominal fee.
When both market entrants and incumbents restrict innovation investment, productivity growth naturally declines. Consequently, the patent system degradation may be a major component in the mechanism explaining the decline of productivity growth in the last decade.
The good news is that if the changes in the patent system of the last decade are a core cause of the decline in productivity growth, a repair of key components of the patent system may restore productivity growth. The policy recommendations to change the patent system include modifying PTO rules so as to substantially increase the bar to institute reexams. This is straightforward and increases certainty by restoring rules for the presumption of patent validity. Second, restore a targeted injunction in a patent that only focuses on the relevant component of an invention. By restoring a limited injunction to protect the patent property right, the era of compulsory licensing will be ended. Third, penalize technology companies that repeatedly infringe patents in order to encourage them to voluntarily license technology under fair and reasonable terms. Doing so would eliminate the majority of enforcement matters driven to the courts. Finally, enforcing competition laws against large technology companies would even the playing field for smaller competitors.
These simple modifications will restore strong patents that protect inventor rights and encourage investment into risky research projects. Until the patent system is modified to restore key rights, there is no reason to believe that incentives will encourage critical investment, particularly by market entrants that are responsible for a majority of innovation.
Without implementing these changes, there is no reason to expect a change in dismal productivity growth or aggregate economic growth data for a generation.