Mega Firms Are More Crucial Than Ever for Innovation

A new study, “Mega Firms and Recent Trends in the U.S. Innovation: Empirical Evidence from the U.S. Patent Data” by the National Bureau of Economic Research (NBER), found that big firms have played an increasingly important role in creating new technology over the past two decades.

Data

To measure innovation, the study relied on patent data from the United States Patent and Trademark Office. Since many patents may be filed for purely strategic reasons and are never applied, the authors defined novel patents as those that introduce new combinations of technologies. An example is the 1994 Citibank patent for “trusted agents for open electronic commerce,” which combined commerce and encryption technology to anonymize e-commerce transactions. As the study puts it, this innovation solved the “joint problem of protecting the privacy of buyers and sellers” while allowing e-commerce to persist.

To measure the effect of business size on innovation, the authors distinguished mega firms (the top 50 firms by sales in any given year among all public firms in the Compustat database) from other firms in the database.

Results

Mega firms produced a large—and increasing—share of novel patents. The share of these firms in new patent filings dropped by half from approximately 16 percent in the early 1980s to roughly 8 percent in 2000, only to fully rebound by 2016. In other words, mega firms played a greater role in developing new technology in the mid–2010s than at any time since the early 1980s.

The increasing share of mega firms in novel patents is also associated with more innovation overall. Consistent with studies by Arts et al. (2021) and Kalyani (2022), the study found that:

The share of novel patents in total patent applications had dropped all the way from 12 percent in 1980 (8 percent at the start of the 1990s) to 3 percent in 2007. [...] After 2007, however, the number of novel patent applications doubled to almost 8,000 per year, and their share in total patent applications had accordingly recovered to 6 percent, the level last seen in the mid-1990s, by 2016.

By this account, mega firms are helping the United States’ innovation ecosystem thrive—not holding it back.

Patents by mega firms were more likely to facilitate further innovation than patents by other entities. Whereas mega firms had about 7 percent fewer follow-on patents (subsequent patents that reuse the new technology of the novel patent, including by other firms) than other entities in 1991–2000, the study reports that “they had 6 percent more in 2007–2016.” Mega firms also had a 1.7 percent lower chance of not having follow-on patents than other entities in 2007–2016. In other words, inventions by mega firms were more likely than inventions by other entities to cause further innovation.

There is also evidence that these follow-on patents benefited firms other than the focal firm (the firm that generated the novel patent). Compared to other businesses, mega firms had smaller shares of follow-on patents assigned to the focal firm. In other words, mega firms diffused knowledge beyond their boundaries, not only to consumers but also to other firms.

Patents by mega firms were also more likely to be “hits” (i.e., in the top one percentile of the distribution of follow-on patents within the first five years). If a patent produces many follow-ons, the original invention was likely groundbreaking and had a high potential for transforming its industry. Mega firms’ share in “hits” almost doubled from 11.9 to 21.2 percent from 1991­–2000 to 2007–2016. This is evidence that, in recent years, mega firms created over a fifth of the most influential inventions.

Other recent research confirms the importance of big firms in innovation. This year, economists showed that the share of patents held by the top one percent of firms in patent stocks has been on the rise over the past several decades. These studies provide evidence that mega firms are essential for sustaining American innovation.

Why This Matters

The anti-corporate neo-Brandeisians seek to advance their goal of change in U.S. antitrust, in part, by advancing the narrative that bigness is antithetical to innovation and therefore, large firms need to be broken up. As this NBER study shows, this is unsupported, if not often outright wrong.

An aggressive, neo-Brandeisian approach to antitrust could do lasting damage to the U.S. economy. For example, in Experimental Capitalism: The Nanoeconomics of American High-Tech Industries, economist Steven Klepper argues that antitrust action against RCA—the leading American color TV producer around the 1950s—was one of the triggers that led to the total demise of the U.S. color TV receivers industry. Faced with multiple lawsuits from the Department of Justice, RCA agreed to make all of its color TV patents available to its domestic rivals free of charge. This turned out to be the beginning of the end for RCA and the rest of the U.S. industry. Indeed, similar mistakes were made with AT&T and Bell Labs, Xerox, and others. With U.S. technological dominance facing increasing global challenges, the stakes could not be higher: antitrust enforcers would do well to heed the famous adage that those who do not learn from history are doomed to repeat it.

If it is true that big companies are getting in the way of innovation, tougher antitrust could make sense. But if these big companies are key actors in creating new tech, that plan could backfire. For example, in the landmark Capitalism, Socialism and Democracy, Joseph Schumpeter discussed how that scale can incentivize and facilitate innovation, such as by allowing firms to appropriate their investments in R&D. In this case, heightened antitrust will discourage this risk-taking and R&D that drives innovation. And, as we have discussed, the evidence continues to support a Schumpeterian approach: big firms play a vital role in shaping America’s innovation economy. An antitrust paradigm that ignores this would have the perverse effect of exporting American innovation to other countries with less regulatory interference.