Identification Properties for Estimating the Impact of Regulation on Markups and Productivity
This paper addresses several shortcomings in the productivity and markup estimation literature. Using Monte-Carlo simulations, the analysis shows that the methods in Ackerberg, Caves and Frazer (2015) and De Loecker and Warzynski (2012) produce bia...
| Main Authors: | , , |
|---|---|
| Format: | Working Paper |
| Language: | English |
| Published: |
World Bank, Washington, DC
2021
|
| Subjects: | |
| Online Access: | http://documents.worldbank.org/curated/en/560821611247482947/Identification-Properties-for-Estimating-the-Impact-of-Regulation-on-Markups-and-Productivity http://hdl.handle.net/10986/35070 |
| Summary: | This paper addresses several
shortcomings in the productivity and markup estimation
literature. Using Monte-Carlo simulations, the analysis
shows that the methods in Ackerberg, Caves and Frazer (2015)
and De Loecker and Warzynski (2012) produce biased estimates
of the impact of policy variables on markups and
productivity. This bias stems from endogeneity due to the
following: (1) the functional form of the production
function; (2) the omission of demand shifters; (3) the
absence of price information; (4) the violation of the
Markov process for productivity; and (5) misspecification
when marginal costs are excluded in the estimation. The
paper addresses these concerns using a quasi-maximum
likelihood approach and a generalized estimator for the
production function. It produces unbiased estimates of the
impact of regulation on markups and productivity. The paper
therefore proposes a work-around solution for the
identification problem identified in Bond, Hashemi, Kaplan
and Zoch (2020), and an unbiased measure of productivity, by
directly accounting for the joint impact of regulation on
markups and productivity. |
|---|