Innovative Resources Moving to Large Firms Likely Isn’t the Reason for Slowed Productivity Growth

A recent International Monetary Fund article titled The Innovation Paradox claims that increased research and development spending has not boosted productivity because innovative resources have shifted toward large, established companies rather than smaller, more innovative ones. The author notes that “while the U.S. has been increasing overall R&D spending relative to GDP, the shift of inventive talent toward large, old businesses has not led to the expected boost in productivity.” However, his assertion that the shift in innovative resources to large firms has led to foregone growth in productivity is questionable as it rests on flawed evidence and logic. Moreover, the analysis seems to overlook the broader societal value of R&D spending, which extends beyond direct productivity gains—such as research on orphan drugs. Instead of attempting to interfere with the allocation of innovative resources among firms, policymakers should restore R&D expensing to ensure both large and small firms are incentivized to continue investing.

The article’s assertion that shifting innovative resources to large firms has stifled productivity growth is questionable due to flawed supporting evidence. The author claims that “many innovators…switch jobs and join larger, well-established companies. However, these inventors’ innovativeness dropped by 6 percent,” citing a study by Akcigit and Goldschlag. That said, this study uses 2000 as a base year when examining the allocation of inventors to younger, small businesses compared to larger, incumbent firms. This is problematic because 2000 was the year the dot-com bubble peaked and burst, meaning many startups reached a high and subsequently went bankrupt. As such, studies on young firms will overestimate their changes, for example, the number of inventors at young firms.

Additionally, the study measures the impact of an inventor’s innovation using patents, which primarily capture product rather than process innovation—where large firms excel. A study by Knott and Vieregger examining R&D productivity and firm size explains that patents “miss a good deal of large firm innovation. This is because process and incremental innovation (the two forms favored by large firms) … are less likely to be patented.” Thus, Akcigit’s claim that inventors at large firms are less innovative than those at younger firms may not necessarily be accurate.

In supporting his assertion, Akcigit further notes that small businesses “use R&D resources more efficiently,” citing an article by Akcigit and Kerr. However, this article also relies on patents as a proxy for innovation, which fails to capture a significant share of large companies’ innovative output. Indeed, as Acs and Audretsch highlight, studies testing the Schumpeterian hypothesis that firm size is related to innovative activity are flawed because they use patents or other proxies to measure innovation despite the fact that not all inventions are patented. Another study corroborates this, noting that “patents capture a small subset of corporate innovation occurring in industries.” As such, Akcigit’s supporting evidence that small businesses use R&D resources more efficiently than larger firms is questionable at best, making his assertion unsubstantiated.

Finally, domestic R&D is not the primary driver of national productivity growth. Increased R&D spending may still generate significant societal value that outweighs the foregone gains in productivity. For example, the pharmaceutical industry may invest in orphan drug research that benefits only a handful of patients in the United States, while aerospace companies may fund theoretical physics research that can be applied to future products. These efforts may not have an immediate impact on productivity growth, but R&D spending in these areas still provides immense value to society by advancing scientific progress, driving medical breakthroughs, and much more. Additionally, many productivity-enhancing innovations that companies adopt do not originate from domestic R&D efforts, but rather from the pool of knowledge generated by global R&D. As such, the correlation between national R&D spending and national productivity growth should be limited.

Policymakers should not attempt to reallocate innovative resources from large to small firms. Instead, they should pass the Tax Relief for American Families and Workers Act of 2024, which will restore R&D expensing. Congress should also expand the R&D tax credit. These measures will encourage firms of all sizes to continue investing in research and development, subsequently ensuring both productivity-enhancing and socially valuable contributions are made.