A new working paper from researchers affiliated with the U.S. Census Bureau and several universities revisits one of the biggest questions in innovation policy: why has productivity growth slowed even as research and development spending continues to rise? For the technology-based economic development (TBED) community, the answer matters because it shapes how states, regions, and federal agencies think about innovation investments. If the problem is weak idea generation, policy options may be limited. But if firms are still producing ideas and struggling to scale them, the challenge becomes commercialization, workforce capacity, and diffusion. For example, a region may already be producing promising university research, but still struggle if startups cannot find early customers, manufacturers cannot find skilled technicians, or small firms lack the supplier relationships needed to scale.
A widely cited 2020 paper by Nicholas Bloom and colleagues argued that firms must devote increasing amounts of research effort just to maintain existing rates of technological progress. Other researchers have pointed to declining business dynamism, growing market concentration, and weaker knowledge spillovers as possible explanations for slower productivity growth. The new study, Growth Is Getting Harder to Find, Not Ideas revisits these questions using a much larger, longer-running dataset than many previous analyses.
Using Census Bureau business data linked to patents, R&D spending, employment, payroll, and sales records from 1977 through 2021, the study separates innovation into two related but distinct processes. First is the production of ideas themselves, measured primarily through patents. Second is the ability of those ideas to generate broader economic growth through productivity gains and firm expansion. The authors argue that the bottleneck is not necessarily generating ideas but turning them into broad-based growth. They find that firms are still producing innovations; patent output relative to R&D spending has not fallen over time, and in some cases appears to have improved. Companies with stronger patent growth generally continue to experience stronger sales per worker growth. Over time, the economy appears to have become less effective at translating invention into broad-based growth. The issue may be less about invention than about diffusion and scale. Promising technologies do not generate broad economic gains if firms cannot expand production, find skilled workers, or spread adoption across supply chains.
The study also highlights how the geography and sectoral composition of innovation have changed over time. Manufacturing firms account for a smaller share of breakthrough patents than they once did, while information technology, professional services, and management-related sectors now play a much larger role. Large firms continue to dominate many high-impact innovations, even as patenting activity itself has spread more broadly across sectors.
For TBED practitioners, much of this will sound familiar: research activity alone is not enough. As SSTI noted in the May 21, 2026, Digest, regions benefit economically when they have the systems and institutions needed to move ideas into the marketplace and spread those benefits across firms and workers. In practice, that can mean better technology transfer, proof-of-concept funding, support for entrepreneurs, training programs aligned with industry needs, and partnerships that help smaller firms connect with universities, suppliers, and customers.
The findings are especially relevant as states and regions continue implementing place-based innovation strategies tied to initiatives such as the CHIPS and Science Act, NSF Regional Innovation Engines, and EDA Tech Hubs. Many of these initiatives understandably focus on research capacity and attracting technology firms. This paper suggests that TBED practitioners should pay equal attention to the less visible parts of the innovation system, including supplier networks, workforce development, commercialization support, and technology diffusion. In some regions, that may be the difference between funding a research center and building the conditions that give startups a better chance to survive, help suppliers qualify for new work, and create enough opportunities that trained workers do not leave for larger labor markets.
This page was prepared by SSTI using Federal funds under award ED22HDQ3070129 from the Economic Development Administration, U.S. Department of Commerce. The statements, findings, conclusions, and recommendations are those of the author(s) and do not necessarily reflect the views of the Economic Development Administration or the U.S. Department of Commerce.