Estimating production functions using inputs to control for unobservables

James Levinsohn, Amil Petrin

Research output: Contribution to journalArticlepeer-review

2195 Scopus citations


We add to the methods for conditioning out serially correlated unobserved shocks to the production technology. We build on ideas first developed in Olley and Pakes (1996). They show how to use investment to control for correlation between input levels and the unobserved firm-specific productivity process. We show that intermediate inputs (those inputs which are typically subtracted out in a value-added production function) can also solve this simultaneity problem. We discuss some theoretical benefits of extending the proxy choice set in this direction and our empirical results suggest these benefits can be important.

Original languageEnglish (US)
Pages (from-to)317-341
Number of pages25
JournalReview of Economic Studies
Issue number2
StatePublished - Apr 2003

Bibliographical note

Funding Information:
Acknowledgements. We would like to thank our referees and Editor for helpful suggestions that led to a substantially revised paper. We would like to thank seminar participants at UC Berkeley, University of Toronto, Yale University, Harvard University, University of Chicago, and NBER for helpful suggestions on earlier work on this project. Jason Abrevaya, Dan Ackerberg, Susanto Basu, Joel Horowitz, Peter Klenow, Steve Olley, Ariel Pakes and Mark Roberts provided especially helpful suggestions. Wendy Petropoulos provided splendid research assistance and many helpful ideas. We are grateful to the Russell Sage Foundation and the Centel FoundationIRobert P.Ruess Faculty Research Fund at the GSB, the University of Chicago for support.


Dive into the research topics of 'Estimating production functions using inputs to control for unobservables'. Together they form a unique fingerprint.

Cite this