Estimating Poisson pseudo-maximum-likelihood rather than log-linear model of a log-transformed dependent variable

Abstract

Purpose The purpose of this study is to account for a recent non-mainstream econometric approach using microdata and how it can inform research in business administration. More specifically, the paper draws from the applied microeconometric literature stances in favor of fitting Poisson regression with robust standard errors rather than the OLS linear regression of a log-transformed dependent variable. In addition, the authors point to the appropriate Stata coding and take into account the possibility of failing to check for the existence of the estimates – convergency issues – as well as being sensitive to numerical problems. Design/methodology/approach The author details the main issues with the log-linear model, drawing from the applied econometric literature in favor of estimating multiplicative models for non-count data. Then, he provides the Stata commands and illustrates the differences in the coefficient and standard errors between both OLS and Poisson models using the health expenditure dataset from the RAND Health Insurance Experiment (RHIE). Findings The results indicate that the use of Poisson pseudo maximum likelihood estimators yield better results that the log-linear model, as well as other alternative models, such as Tobit and two-part models. Originality/value The originality of this study lies in demonstrating an alternative microeconometric technique to deal with positive skewness of dependent variables.

Tourism and Economic Development: Evidence from Mexico’s Coastline

“…In the calibration, we refer to ~ A i = A i Z T,i β as the tourism attractiveness shifter in region i, which captures both a productivity and an amenity shifter. It follows that tourism trade shares can be written as (7) λ…”

“…Step: Calibration of Regional Fundamentals.-Using information on ( w n , L M,n , L S,n , L T,n , G n ) with parameters ( ν j j′ , α MT , α G , β, θ, ρ ) , we calibrate the baseline equilibrium according to equations (9), (10), (7), (8), (13), (14), and (15). Following Redding (2016), we invert the calibrated model to recover the unique tourism and manufacturing shifters ˜ A n and M n (up to scale) that are consistent with the data.…”

“…Methodologically, the paper follows a recent but growing literature that uses quantitative spatial equilibrium models to analyze the welfare consequences of aggregate or local shocks, taking into account the frictions to trade and mobility between regions within countries (e.g., Redding 2016;Caliendo et al 2018;Monte, Redding, and Rossi-Hansberg 2018;Bryan and Morten 2015;Caliendo, Dvorkin, and Parro 2015;Fajgelbaum et al 2019;Adao, Arkolakis, and Esposito 2017;and Galle, Rodríguez-Clare, and Yi 2014). 7 We build on the framework developed by Allen and Arkolakis (2014) and Redding (2016) and extend the model and methodology in several dimensions to study the role of within-and cross-sector agglomeration externalities, a novel dimension in this class of quantitative frameworks. We combine the structure of the model with observed empirical moments to identify the strength of the agglomeration forces, close to the approach followed in a one-sector model in Ahlfeldt et al (2015).…”

Inclusive Finance, Environmental Regulation, and Public Health in China: Lessons for the COVID-19 Pandemic

“…Silva and Teneyro (18,19) proposed the PPML estimation method to solve the above two problems and exemplified the gravity model used in international trade research. Note that the PPML estimator is scale-invariant since we have used the logtransformed dependent variable (20). Besides, the PPML method can solve a possible endogeneity bias.…”

Research methods in management: advances and applications

“…The fifth quantitative article was written by Motta (2019), who argues that researchers in different fields within business administration often estimate models with a log-transformed dependent variable. Failure to account for adjustments for heteroskedasticity and normality of residuals may lead to biased estimates of the conditional mean and the slope on its original scale.…”