« Back to Results

Identification and Estimation of Production Functions

Paper Session

Saturday, Jan. 5, 2019 8:00 AM - 10:00 AM

Atlanta Marriott Marquis, M105
Hosted By: Econometric Society
  • Chair: David Rivers, University of Western Ontario

Constructive Identification of Heterogeneous Elasticities in the Cobb-Douglas Production Function

Tong Li
,
Vanderbilt University
Yuya Sasaki
,
Vanderbilt University

Abstract

This paper presents the identification of heterogeneous elasticities in the Cobb-Douglas production function. The identification is constructive with closed-form formulas for the elasticity with respect to each input for each firm. We propose that the flexible input cost ratio plays the role of a control function under ``non-collinear heterogeneity'' between elasticities with respect to two flexible inputs. The ex ante flexible input cost share can be used to identify the elasticities with respect to flexible inputs for each firm. The elasticities with respect to labor and capital can be subsequently identified for each firm under the timing assumption admitting the functional independence.

Timing Assumptions and Efficiency: Empirical Evidence in the Production Function Context

Daniel Ackerberg
,
University of Michigan

Abstract

Much of the recent empirical work estimating production functions has used methodologies proposed
in two distinct lines of literature: 1) the literature started by Olley and Pakes (1996) on "proxy variable"
techniques, and 2) what is commonly referred to as the "dynamic panel" literature. We illustrate how
timing and Örm information set assumptions are key to both methodologies, and how these assumptions
can be strengthened or weakened almost continously. We also discuss other assumptions that have
utilized in these literatures to increase the precision of estimates. Empirically, we then examine how,
in a number of plant level production datasets, strengthening or weakening the timing/information
set assumptions a§ects the precision of estimates. We compare these impacts on precision to those
achieved by imposing other potential assumptions. This gives the researcher a better idea of the
e¢ ciency tradeo§s between di§erent possible assumptions in the production function context.

Heterogenous Production Functions, Panel Data, and Productivity Dispersion

Jeremy Fox
,
Rice University
Vitor Hadad
,
Boston College
Stefan Hoderlein
,
Boston College
Amil Petrin
,
University of Minnesota
Robert Sherman
,
California Institute of Technology

Abstract

We explore the nonparametric identification of the distribution of heterogenous production functions: production functions where the parameters vary across firms in the same industry for unmeasured reasons. Our production functions have random coefficients that are allowed to be time varying and correlated to the measured inputs used in the production function. We exploit panel data and generalize timing assumptions used in the production function literature following Olley and Pakes (1996). We apply our methods to study productivity dispersion and aggregate productivity in India, a developing country where previous studies have found low levels of productivity and high productivity dispersion. We generalize productivity dispersion and aggregate productivity growth measures to allow for multidimensional unobserved heterogeneity in production functions.

On the Identification of Gross Output Production Functions

Salvador Navarro
,
University of Western Ontario
David Rivers
,
University of Western Ontario

Abstract

We study the nonparametric identification of gross output production functions under the environment of the commonly employed proxy variable methods. We show that applying these methods to gross output requires additional sources of variation in the demand for flexible inputs (e.g., prices). Using a transformation of the firm’s first-order condition, we develop a new nonparametric identification strategy for gross output that can be employed even when additional sources of variation are not available. Monte Carlo evidence and estimates from Colombian and Chilean plant-level data show that our strategy performs well and is robust to deviations from the baseline setting.
JEL Classifications
  • C1 - Econometric and Statistical Methods and Methodology: General
  • C5 - Econometric Modeling