Productivity Paradox

The emergence of the paradox and classic explanations

US labor productivity growth slowed sharply from the early 1970s, even as firms invested heavily in computers over the same period. Solow's (1987) one-line remark came to symbolize this mismatch. Brynjolfsson (1993) set out four explanations. The first is mismeasurement, in which improvements in output quality and variety are not properly captured in statistics. The second is lags before investments pay off. The third is redistribution, in which IT is used merely to divide profits among competing firms. The fourth is mismanagement, the failure to use the technology well. David (1990) offered a historical analogy: the electric dynamo, invented in the 1880s, took about 40 years to raise factory productivity substantially, because factories built around the steam engine had to be redesigned for electric motors.

Firm-level evidence and the late-1990s revival

Analyzing firm-level data, Brynjolfsson and Hitt (1996) showed that information systems spending yields substantial returns and declared the paradox lost. Bresnahan, Brynjolfsson and Hitt (2002) showed that IT has large effects when combined with complementary organizational changes, such as decentralized organization and skilled workers. This fits the features of general purpose technologies as defined by Bresnahan and Trajtenberg (1995): technologies that spread across many industries and spur complementary innovation. US productivity growth did rebound in the late 1990s, and Oliner and Sichel (2000) and Jorgenson (2001) found that the production and use of IT explained much of the revival. The episode illustrates the interplay of technology and organization emphasized by creative destruction and endogenous growth theory.

A new slowdown and the AI debate

Productivity growth slowed again after about 2005, reviving the paradox. Gordon (2016) argued that the economic impact of recent digital innovations is small compared with great inventions of the past, such as electricity and the internal combustion engine. Some attributed the slowdown to mismeasurement, with the value of free digital services missing from the statistics, but Byrne, Fernald and Reinsdorf (2016) and Syverson (2017) countered that mismeasurement alone cannot explain its scale. Brynjolfsson, Rock and Syverson (2019, 2021) argued that general purpose technologies such as AI trace a productivity J-curve. Because they first require investment in unmeasured intangible assets, productivity is at first underestimated and later overestimated. The productivity paradox thus shows that the effects of new technologies appear late and unevenly, depending on organizational change and measurement systems. It also relates to Baumol's cost disease, which addresses stagnant productivity in services.

FAQ

What is the productivity paradox in simple terms?

Companies spent enormous sums on computers and IT, yet the effect barely showed up in national productivity statistics. It became famous through Solow's quip that the computer age is everywhere except in the productivity statistics.

Will AI face a productivity paradox too?

That possibility is being debated. Researchers argue that general purpose technologies like AI first require organizational change and investment in intangible assets, so their effects may follow a J-curve and take time to show up in the statistics.

Sources reviewed: 2026-10-26

References

1. Solow, R. M. (1987, July 12). We'd better watch out. New York Times Book Review, 36.
2. David, P. A. (1990). The dynamo and the computer: An historical perspective on the modern productivity paradox. American Economic Review, 80(2), 355–361.
3. Brynjolfsson, E. (1993). The productivity paradox of information technology. Communications of the ACM, 36(12), 66–77.
4. Bresnahan, T. F., & Trajtenberg, M. (1995). General purpose technologies: 'Engines of growth'? Journal of Econometrics, 65(1), 83–108.
5. Brynjolfsson, E., & Hitt, L. (1996). Paradox lost? Firm-level evidence on the returns to information systems spending. Management Science, 42(4), 541–558.
6. Oliner, S. D., & Sichel, D. E. (2000). The resurgence of growth in the late 1990s: Is information technology the story? Journal of Economic Perspectives, 14(4), 3–22.
7. Jorgenson, D. W. (2001). Information technology and the U.S. economy. American Economic Review, 91(1), 1–32.
8. Bresnahan, T. F., Brynjolfsson, E., & Hitt, L. M. (2002). Information technology, workplace organization, and the demand for skilled labor: Firm-level evidence. Quarterly Journal of Economics, 117(1), 339–376.
9. Byrne, D. M., Fernald, J. G., & Reinsdorf, M. B. (2016). Does the United States have a productivity slowdown or a measurement problem? Brookings Papers on Economic Activity, 2016(1), 109–182.
10. Gordon, R. J. (2016). The Rise and Fall of American Growth. Princeton: Princeton University Press.
11. Syverson, C. (2017). Challenges to mismeasurement explanations for the US productivity slowdown. Journal of Economic Perspectives, 31(2), 165–186.
12. Brynjolfsson, E., Rock, D., & Syverson, C. (2019). Artificial intelligence and the modern productivity paradox: A clash of expectations and statistics. In A. Agrawal, J. Gans, & A. Goldfarb (Eds.), The Economics of Artificial Intelligence (pp. 23–57). Chicago: University of Chicago Press.
13. Brynjolfsson, E., Rock, D., & Syverson, C. (2021). The productivity J-curve: How intangibles complement general purpose technologies. American Economic Journal: Macroeconomics, 13(1), 333–372.