Production Function Estimation With Resource Misallocation
We show that the proxy variable method fails to yield consistent estimates of production function coefficients in the presence of resource misallocation. This failure arises because unobserved firm‐specific distortions violate the assumptions of scalar unobservability and strict monotonicity. We propose a novel identification strategy that does not rely on these assumptions. Our method is robust to various distortions and production function specifications, and can be extended to accommodate serial correlation in unexpected productivity shocks. Monte Carlo experiments confirm the efficacy of our approach in consistently estimating production function coefficients under resource misallocation. Using a large panel of Chinese manufacturing firms—and employing the share of state‐owned firms as a proxy for industry‐level resource misallocation—we find that estimates obtained via the proxy variable method more closely align with those from our proposed approach when resource misallocation is reduced.
No keywords indexed for this article. Browse by subject →
Daniel A. Ackerberg, Kevin Caves, Garth Frazer
John Asker, Allan Collard-Wexler, Jan De Loecker
Richard Blundell, Stephen Bond
Loren Brandt, Johannes Van Biesebroeck, Yifan Zhang
Chang-Tai Hsieh, Peter J. Klenow
James Levinsohn, Amil Petrin
G. Steven Olley, Ariel Pakes
Diego Restuccia, Richard Rogerson
- Published
- Jan 21, 2026
- Vol/Issue
- 41(3)
- Pages
- 280-294
- License
- View
You May Also Like
M. Hashem Pesaran, Yongcheol Shin · 2001
12,917 citations
M. Hashem Pesaran · 2007
9,732 citations
Jushan Bai, PIERRE PERRON · 2002
3,992 citations
Leslie E. Papke, Jeffrey M. Wooldridge · 1996
2,856 citations