The surge of invention and industrial change known as the Second Industrial Revolution helped set the stage for the exuberance of the 1920s. New goods, new ways to manufacture them, and rapid diffusion of modern industrial methods fed a widespread belief that technological progress would keep delivering strong growth. In a new study, economists Harold L. Cole, Stefano Cravero, and Jeremy Greenwood ask whether that optimism may also have carried a hidden risk: an economy that poured too much into building capacity even as the pace of new product creation eased and consumers’ appetite for additional goods began to level off.
Their central question is whether the celebrated “Roaring Twenties” could have contained the seeds of the Great Depression through a mechanism of “overaccumulation”—investment that outstrips what later demand and innovation can profitably support. The authors focus on two related possibilities. First, businesses may have continued investing as if a torrent of novel products would keep arriving, when in fact product innovation was slowing and markets for many new goods were becoming saturated. Second, even if fewer breakthrough products were appearing, firms could still have been encouraged to keep spending because production methods were steadily improving. Ongoing process innovation—incremental improvements that raise efficiency—can make new factories and machines look attractive even when the market for final goods is no longer expanding as quickly.
To evaluate these ideas, the researchers build a macroeconomic model designed to capture both sides of technological change: product innovation, which expands the variety of things households and businesses can buy, and process innovation, which improves how efficiently those things are made. By embedding both channels in a single framework, the model can trace how expectations about future technological progress feed into investment decisions, and how mismatches between expected and realized innovation might destabilize the broader economy.
The team then conducts “proof-of-concept” simulations—tests that examine whether the proposed mechanism is capable, in principle, of generating dynamics resembling those seen around the onset of the Great Depression. The results indicate that it can. In their simulations, overly upbeat expectations about product-driven growth lead to excessive investment during the boom. When product innovation slows and demand becomes more fully satisfied, the economy is left with too much capital relative to profitable opportunities, setting the stage for a sharp downturn. Continued process innovation can intensify the buildup by making investment appear worthwhile for longer, even as underlying market conditions deteriorate.
The study does not claim to provide the only explanation for the Great Depression. Instead, it demonstrates that a technology-driven overinvestment narrative is quantitatively plausible within a modern macroeconomic setting—one that distinguishes between inventing new goods and improving the production of existing ones.
As with all National Bureau of Economic Research publications, the paper reflects the authors’ views, not necessarily those of the NBER. The organization also distributes research from its network through free outlets including the NBER Reporter, NBER Digest, and bulletins focused on health, Alzheimer’s economics, and entrepreneurship, alongside research spotlights, lectures, and panel videos.