Missing Growth from Creative Destruction
GAICIL Insights
The statistics may be undercounting progress precisely where progress is fastest.1 Missing Growth from Creative Destruction, by Philippe Aghion, Antonin Bergeaud, Timo Boppart, Peter J. Klenow and Huiyu Li, identifies a systematic bias in how official statistics treat products that disappear. When products exit, statistical agencies often impute inflation from surviving products. That understates growth if the products replacing them through creative destruction improve more than survivors. The newest, best things are the ones the price index handles worst.
The authors build a model with incumbent and entrant innovation that yields an explicit expression for missing growth as a function of the frequency and size of creative destruction relative to other innovation. The key insight is testable: if creatively destroyed products improve more than survivors, the market share of surviving products should systematically shrink. Using market shares of surviving, entering and exiting plants in US Census data on non-farm businesses from 1983 to 2013, they estimate missing growth's magnitude. Market shares become a lie detector for price indices.
The findings are substantial. Missing growth from imputation was about half a percentage point per year, or over one-fifth of measured productivity growth. It arose mostly from hotels and restaurants, with less from manufacturing. It accelerated only modestly after 2005, accounting for only about one-tenth of the sharp growth slowdown since then. Mismeasurement is real and important, and it does not explain the recent slowdown. Both the optimists and the pessimists get something here.
The estimate may be conservative. The authors assume no errors in measuring incumbents' quality improvements on their own products. They view the bias as over and above, and amplified by, the quality bias emphasized by the Boskin Commission. Extensions include measuring missing growth in other countries and revisiting optimal innovation policy, since the optimal R&D subsidy may be larger if true growth exceeds measured growth.
Now the AI angle. Artificial intelligence is about to stress-test this bias. AI is expected to generate waves of creative destruction, with new products and services replacing existing ones. If statistical methods fail to capture improvements embodied in new products, official statistics may understate the growth AI generates. Debates over whether AI is delivering productivity gains should account for missing growth. The productivity paradox of AI may be partly a measurement paradox.
Digital and AI-enabled products make measurement harder still. Many are new, rapidly improving or offered at zero price, and hard to capture in conventional price indices. Using market share dynamics to infer missing growth offers a promising way to measure AI's contribution to economic welfare. When a product is free, the price index registers nothing at all.
Policy stakes follow from the numbers. If growth is understated, returns to innovation may be larger than they appear, strengthening the case for supporting research and development. A higher true growth rate implies greater benefits from innovation, so the optimal R&D subsidy may be larger. Policymakers relying on measured growth may underinvest in the activities driving progress. A thermometer that reads low leads to the wrong prescription.
The post-2005 result keeps the argument honest. Missing growth does not explain the slowdown, which appears real, so its causes must be sought elsewhere. Measurement error and genuine slowdown can coexist, and distinguishing them is essential.
Statistical agencies face a clear question: how should methodology change in light of these findings? Better methods for valuing new products and accounting for creative destruction would improve growth statistics. As AI accelerates product turnover, the question becomes more pressing.
A hard question for statistical agencies: modernizing price indices for an economy of rapid product turnover may be one of the highest-return AI policies available. Without it, the debate over AI's productivity payoff may be settled by a measurement artifact.
Introduction
For exiting products, statistical agencies often impute inflation from surviving products. This understates growth if creatively-destroyed products improve more than surviving ones. If so, then the market share of surviving products should systematically shrink. Using entering and exiting establishments to proxy for creative destruction, we estimate missing growth in US Census data on non-farm businesses from 1983 to 2013. We find missing growth (i) equaled about one-half a percentage point per year; (ii) arose mostly from hotels and restaurants rather than manufacturing; and (iii) did not accelerate much after 2005, and therefore does not explain the sharp slowdown in growth since then. (JEL E23, E31, L14, L15, O30, O41)
Takeaway
In this paper we lay out a model with incumbent and entrant innovation to assess the unmeasured TFP growth resulting from creative destruction. Crucial to this missing growth is the use of imputation by statistical agencies when producers no longer sell a product line.1
Our model generates an explicit expression for missing TFP growth as a function of the frequency and size of creative destruction vs. other types of innovation. Based on the model and U.S. Census data for all nonfarm businesses, we estimated the magnitude of missing growth from creative destruction over the period 1983–2013. Our approach uses the market share of surviving, entering, and exiting plants. We found that: (i) missing growth from imputation was substantial at around half a percentage point per year — or over one-fifth of measured productivity growth; and (ii) it has accelerated modestly since 2005, so that it accounts for only about one-tenth of the sharp growth slowdown since then. We may be understating missing growth because we assumed there were no errors in measuring quality improvements by incumbents on their own products. We think missing growth from imputation is over and above (and amplified by) the quality bias emphasized by the Boskin Commission. Our analysis could be extended in several interesting directions. One would be to look at missing growth in countries other than the U.S. A second extension would be to revisit optimal innovation policy. Based on Atkeson and Burstein (2018), the optimal subsidy to R&D may be bigger if true growth is higher than measured growth.2
Conversely, our estimates give a more prominent role to creative destruction with its attendant business stealing. A natural question is how statistical offices should alter their methodology in light of our results, presuming our estimates are sound. The market share approach would be hard to implement without a major expansion of BLS data collection to include market shares for entering, surviving, and exiting products in all sectors. The indirect inference approach is even less conducive to high frequency analysis. A feasible compromise might be for the BLS to impute quality growth for disappearing products based on its direct quality adjustments for those surviving products that have been innovated upon. Our missing growth estimates have other implications which deserve to be explored further. First, ideas may be getting harder to find, but not as quickly as official statistics suggest if missing growth is sizable and relatively stable.
Economists at the BLS and BEA recently estimated that quality bias from health and ICT alone was about 40 basis points per year from 2000–2015 Groshen et al. (2017).3
Notes
- Philippe Aghion et al., Missing Growth from Creative Destruction, 109 Am. Econ. Rev. 2795 (2019), https://doi.org/10.1257/aer.20171745. Takeaway source: Conclusion, reproduced from a free author or repository copy (working-paper version). ↩
- Atkeson, Andrew, and Ariel Burstein. 2018. “Aggregate implications of innovation policy.” forthcoming.. ↩
- Groshen, Erica L. Brian C. Moyer, Ana M. Aizcorbe, Ralph Bradley, and David M. Friedman. 2017. “How Government Statistics Adjust for Potential Biases from Quality Change and New Goods in an Age of Digital Technologies: A View from the Trenches.” Journal of Economic Perspectives, 31(2): 187–210.. ↩
Cite as: Philippe Aghion et al., Missing Growth from Creative Destruction, 109 Am. Econ. Rev. 2795 (2019), https://doi.org/10.1257/aer.20171745
Header image: Old digital media - Computer History Museum (30781540312).jpg by Ik T from Kanagawa, Japan, CC BY 2.0, via Wikimedia Commons.
.jpg){kind=link}