Over the past few decades, economists have tracked two discomforting—and perhaps related—long-term trends for the U.S. economy.
Trend No. 1: The productivity of American workers and businesses is barely growing, at least compared to prior years. Since 2004, annual growth in total factor productivity—a measure of how much we can produce with a given level of inputs—has hovered below 1 percent (Figure 1, bar chart).
Trend No. 2: The business landscape has grown more concentrated. The leading firm in any given industry now commands more than 50 percent of its market, on average (Figure 1, line chart).
Economists can theorize causal arrows pointing in either direction: Are corporate behemoths making us less productive? Or could declining productivity be providing some advantage to large firms? Untangling these connections could indicate the right policies to spark innovation and productivity, increase competition and efficiency, and raise the overall U.S. standard of living.
Rather than one trend driving the other, new work by economist Jane Olmstead-Rumsey points to a third force driving both trends: the declining quality of U.S. patents. Olmstead-Rumsey, now at the London School of Economics, developed the most recent version of her research, “Market Concentration and the Productivity Slowdown,” while serving as a visiting scholar at the Opportunity & Inclusive Growth Institute at the Minneapolis Fed.
Patents just aren’t what they used to be
Scholars have taken various approaches to measure the quality of a patent, such as how often it is cited by subsequent patent applications. Olmstead-Rumsey leverages a formula developed in a 2017 paper1 that focuses on the contribution of a new patent to a company’s stock value. “Firms patent not just because it helps other people develop subsequent stuff,” she said. “They patent because they want a strategic advantage.” A higher-quality patent is one that makes a larger contribution over time to a company’s stock price.
Olmstead-Rumsey analyzed 1.3 million patents issued to 5,700 public companies since 1980 and found patent quality has fallen across the board. From the high-productivity-growth period (1994-2003) to the lower-growth period (2004-2017), the average value of a patent declined by more than half.2
She believes the decline reflects that we are on the tail end of a transformative wave of computing advances that peaked in the 1990s. “When you think about the history of the U.S., times in which patents were of high-average quality are exactly when we had new general-purpose technologies,” she said. “The steam engine, electrification, railroads—technologies that, like computers, were applicable in a wide set of industries.”
Changes unleashed across U.S. businesses by computing and networking in the 1990s and early 2000s were revolutionary. Those since have been more incremental. She offers the example of Amazon, which began a series of patents in 1998 related to a merchant making recommendations based upon items a shopper considered but did not purchase. The notion feels obvious 25 years later. At the time, however, it marked a fundamental shift in online shopping and helped transform Amazon from one of many online bookstores into the world’s largest retailer.
“When another e-commerce firm is innovating now around the shopping cart, it’s just not viewed as having the same impact on that firm as Amazon being the first one to do it,” said Olmstead-Rumsey.
The dwindling advantage of playing catch-up
The decline in patent quality after the year 2000 coincides with the U.S. productivity slowdown. That connection feels intuitive: As innovation falls, so would productivity growth. But what about our second big trend? How could lower-value patents translate into more market concentration?
Olmstead-Rumsey takes a critical step beyond prior research by separating firms into market “leaders” and “followers.” The first notable observation is that the follower firms have much higher-quality patents, on average (Figure 2).
Business and management theories abound for this advantage of innovating from the rear. Some are organizational: The research and development teams of smaller, newer companies might benefit from being “closer to the ground” and more in-tune with their customers. The CEO—imagine the founder of a start-up—is more likely to be intimately involved with R & D and drive an aggressive research vision.
If you’re back in the pack, you also have a clear view of the dominant firms and the opportunity to shape your strategy in response. “You can glean a lot from what the market leader is doing,” said Olmstead-Rumsey. “It’s a lot easier to go from behind to the frontier than it is to innovate on the frontier.”
Follower firms also have a strong incentive to go big or go home. If they ever hope to vault from bit player to market leader—like Amazon—it will take a moonshot approach to find (and patent) truly disruptive innovations. Market leaders, by contrast, “just want to march forward and assure their position,” Olmstead-Rumsey said. All else equal, they generally have less incentive to gamble on high-risk, high-reward R & D.
However, the trend over time in Figure 2 shows the situation changing. Patent quality at the follower firms has taken a much steeper dive. This trend has substantially eroded the traditional edge in patent quality enjoyed by followers.
The result, says Olmstead-Rumsey, is a change in incentives across all firms. For followers, lower patent quality compared with the 1980s and ’90s means they are less likely to strike it big. “The small firms understand that it’s harder to catch up, and they tend to reduce their investments in R & D,” she said. “They’re patenting less, and when they do, those patents are worse than before.”
This cycle alone could lead to more concentrated market share and less leadership turnover in industries—just as we see in the data. But in Olmstead-Rumsey’s economic model, she finds the big firms also react to the changing environment by further cementing their position.
“They are now less likely to be displaced,” she said. “They look into the future and say, ‘Wow, I’m going to be the big player for the next 10 or 20 years. Any improvements I make to my product—which will allow me to increase my markup [profits] and increase my market share—are going to accrue to me for a longer period. I actually might want to invest more in R & D.”
As leading firms do relatively more R & D and follower firms less, Olmstead-Rumsey says the leaders reinforce the effects by gobbling a larger share of the research workforce.
Basic research and the next innovation wave
Olmstead-Rumsey adds to an ever-growing body of research seeking to understand the slow productivity growth of the U.S. economy and the trend toward greater market concentration. As always, when economic models meet the real world, there are likely many factors at play.
What is the effect on innovation, for example, of market leaders buying up competitors and otherwise inhibiting competition from smaller rivals? While anticompetitive activity could play a role in declining innovation, Olmstead-Rumsey says it does not obviously explain the decline in patent quality, which has been particularly severe for the smaller—and traditionally most innovative—businesses.
If this patent decline is at the root of some of our challenges, what policies might help juice our collective creative process? The story told by the research leads back to the power of general-purpose technologies, or GPTs, to drive higher-quality patents. With previous GPTs like the microchip and the internet now receding into the rearview mirror, we would want to invest in the type of basic scientific research more likely to lead to the next big thing.
On its own, however, the private sector will tend to underprovide this kind of R & D. “Companies are very focused on applied research that is directly related to what they already do,” said Olmstead-Rumsey. “Basic research is where we tend to come up with a lot of really important innovations, [and many] have come out of the government side.”
With its lack of immediate commercial benefit, basic research fits the classic profile of a public good. From the U.S. Department of Defense to the National Institutes of Health, the federal government is deep into the basic-research business (Figure 3).
However, the government is even more in the business of funding later-stage (“applied” and “development”) research. And at $41 billion in 2021, spending on basic research remains more or less a rounding error—less than 1 percent—in the context of $6.8 trillion in federal government spending.
Are we investing enough—and in the right places—to turn the tide on some troubling trends? The answer might look different as we learn more about the connection between patents, productivity, and competition.
Endnotes
1 “Technical Innovation, Resource Allocation, and Growth” by Leonid Kogan, Dimitris Papanikolaou, Amit Seru, and Noah Stoffman.
2 Between 1994 and 2003, the average patent contributed 22.9 percent to a company’s share price. Between 2004 and 2017, this figure was 11.2 percent.
Jeff Horwich is the senior economics writer for the Minneapolis Fed. He has been an economic journalist with public radio, commissioned examiner for the Consumer Financial Protection Bureau, and director of policy and communications for the Minneapolis Public Housing Authority. He received his master’s degree in applied economics from the University of Minnesota.