Resolving the Conflict on Conduct Parameter Estimation in Homogeneous Goods Markets between Bresnahan (1982) and Perloff and Shen (2012)

TL;DR

This paper clarifies the correct approach to estimating conduct parameters in homogeneous goods markets, resolving conflicting claims by demonstrating the importance of demand shifters and sample size for accurate estimation.

Contribution

It corrects the previous proof by Perloff and Shen (2012) and shows that proper inclusion of demand shifters and larger samples lead to accurate conduct parameter estimation, supporting Bresnahan (1982).

Findings

Perloff and Shen's proof is incorrect.

Including demand shifters improves estimation accuracy.

Larger samples lead to better conduct parameter estimates.

Abstract

We revisit conduct parameter estimation in homogeneous goods markets to resolve the conflict between Bresnahan (1982) and Perloff and Shen (2012) regarding the identification and the estimation of conduct parameters. We point out that Perloff and Shen's (2012) proof is incorrect and its simulation setting is invalid. Our simulation shows that estimation becomes accurate when demand shifters are properly added in supply estimation and sample sizes are increased, supporting Bresnahan (1982).