Why the US could lead
without ossifying

Essay #2 · Part 1 · June 11, 2026

Every economy that has held the world's lead eventually ran its money from risk to safety, and lost the lead. The only open question is whether the United States will be the first that does not.

The American century is usually dated to 1945. That is the wrong date. By the measure that decides who stays on top, productivity, the United States had already passed Britain around the turn of the century1; the Second World War only ratified a lead that was forty years in the making. And this was not the first time the frontier of the world economy changed address. From roughly 1450 to 1650 it sat in the city-states of northern Italy2, until Italian capital pulled out of risky trade into safe land and lending, and the Dutch Republic, pushing hard into shipping, public banking, and Atlantic trade, took it. The Dutch then did the same: by the 18th century their capital had drifted into finance, lending abroad to Britain itself, while Britain built industry across an entire national market rather than a single city, and pulled ahead. Britain repeated it once more, its money moving into the City of London and imperial finance while the United States scaled the newer industries and, decade by decade, out-produced it. The pattern does not vary. Every holder of the frontier eventually ossified the same way, its money running from risk to safety, and every time a hungrier challenger inherited the lead. So the real question is not whether the United States leads now. It does. The question is the one every predecessor failed: each ran to safety at its peak. Will this one?

To measure whether a country is leading, the right indicator is Total Factor Productivity (TFP): the share of output growth that added labor and capital cannot explain. Growth accounting consistently finds that this share, not the accumulation of inputs, dominates long run growth3. Picture two bakeries that each sell more bread this year than last. The first hired more bakers and bought more ovens. Its extra bread is fully explained by its extra inputs. The second sold more bread with the same bakers and ovens: it found a better recipe and a smarter workflow. Its extra bread cannot be explained by counting inputs at all. That second kind of growth is what TFP measures. TFP is the share of a country's output growth that the growth of its labor and capital cannot explain. It is the measure of getting smarter, not just bigger.

Between 1880 and 1920, US TFP grew at only 0.46% per year4. The frontier had already moved to the US by this period, but the loop that would convert that position into a productivity surge had not yet been built. From 1920, TFP growth jumped to 1.89% per year and held that pace through 19704. Counterintuitively, the surge ran straight through the Great Depression: private industrial R&D matured, electrification finally diffused through factories, the federal road build-out reshaped logistics, and capital-saving innovations meant more output per unit of input5. WWII industrial mobilization added another push in the 1940s, and the postwar golden age carried it through to 1973.

A surge of that size does not build itself. During World War II, Vannevar Bush ran the federal science effort that produced radar, mass-produced penicillin, and the atomic bomb: proof that government money aimed at hard problems paid off. In 1945 he argued the same engine should run in peacetime6, and the government built it: the National Science Foundation in 1950, a vastly expanded NIH, and, after the Soviet Union launched Sputnik in 1957, NASA and DARPA in 1958. By 1964 the federal government funded 67% of all US R&D, about 1.86% of GDP. This was the failure-tolerant funder Britain never had. Decades later, when much of that research still sat unused (by commonly cited estimates, only ~5% of federally funded patents were ever commercialized, and some 28,000 inventions were left on the shelf), the Bayh-Dole Act of 1980 let universities, small businesses, and nonprofits own and license what public money discovered, and commercialization rose7.

Measured the same way, as social returns, government nondefense R&D returns 140 to 210% on the dollar8 against about 55% for private R&D9. Publicly funded R&D is one of the key drivers of economic growth and of turning new science into real technology. The reason it earns the higher return is structural. A private firm has to capture the return to justify the cost, and risky basic research has distant, uncertain, non-ownable payoffs, so it rationally skips them. A public fund does not need to capture the return, so it can fail nine out of ten and the one survivor still pays for the rest. Failure tolerance is not a trait of the researchers; it is that the funder does not need each bet to pay off.

Britain, for example, did not run out of smart people. Its share of world manufacturing fell from ~23% in 1880 to ~14% in 191310, while the US rose from 15% to 32% over the same span. Britain has won more than 90 science Nobels, and in the early window, 1900 to the 1930s, Germany and Britain split the top of the table while the US barely registered before WWI. The inventors stayed; the money that turns invention into output left. British capital chose safe colonial bonds over risky new industry, and no public funder existed to underwrite the failures private capital would not. That is the loop dying: not a shortage of genius, a shortage of patient money.

Every previous holder of the frontier ossified the same way, and lost it. The pattern is too consistent to be bad luck. Once a country has won, its capital is large, and large capital behaves differently from hungry capital. It has more to lose, the frontier's returns are uncertain, and safe assets pay reliably. For each individual holder of money, the smart move is to leave risk for safety. Nobody decides to ossify. Millions of individually rational choices simply sum to it. Only one force has ever broken that sum: a funder that does not share the incentive, that does not need its own bets to pay off. Every predecessor lacked it and followed the law. America built it. Whether it is still running is the question.

Part 2 follows: is the loop still running in 2026?

references

  1. Broadberry, Stephen N., and Douglas A. Irwin. "Labor productivity in the United States and the United Kingdom during the nineteenth century." Explorations in Economic History 43.2 (2006): 257-279. ↩
  2. Braudel, Fernand. Out of Italy: Two Centuries of World Domination and Demise. Europa Editions, 2019. ↩
  3. Solow, Robert M. "Technical change and the aggregate production function." The Review of Economics and Statistics 39.3 (1957): 312-320. ↩
  4. Gordon, Robert J. The Rise and Fall of American Growth: The U.S. Standard of Living Since the Civil War. Princeton University Press, 2016 (TFP figures, Fig. 17.2). ↩
  5. Field, Alexander J. "The most technologically progressive decade of the century." American Economic Review 93.4 (2003): 1399-1413. ↩
  6. Bush, Vannevar. Science, the Endless Frontier. United States Government Printing Office, 1945. ↩
  7. Tseng, Ampere A., and M. Raudensky. "Performances of technology transfer activities of US universities after the Bayh-Dole Act." Journal of Economics, Business and Management 3.6 (2015): 661-667. ↩
  8. Fieldhouse, Andrew J., and Karel Mertens. "The Social Returns to Public R&D." Federal Reserve Bank of Dallas Working Paper, 2025. ↩
  9. Bloom, Nicholas, Mark Schankerman, and John Van Reenen. "Identifying technology spillovers and product market rivalry." Econometrica 81.4 (2013): 1347-1393. ↩
  10. Bairoch, Paul. "International industrialization levels from 1750 to 1980." Journal of European Economic History 11.2 (1982): 269-333. ↩