Dealing with panel data in accounting and managerial finance research

Abstract

PurposeGiven the importance of panel datasets in contemporary accounting and managerial finance research, the objective of this paper is to provide practical guidance for researchers who are inexperienced in dealing with panel estimation methodologies.Design/methodology/approachThe paper presents and implements a set of procedures designed to establish whether or not pooled estimation is a viable proposition in a given setting. The paper also explores the suitability of alternative estimation methodologies given the results from these test procedures. To illustrate the key concepts the paper utilises a simple model of the relationship between UK directors' cash compensation and three explanatory variables: accounting earnings; stock returns; and firm growth. This model is used solely for illumination purposes and the paper does not seek to contribute to the compensation literature.FindingsThe results demonstrate the potentially misleading inference in panel settings, which can arise from: pooled OLS, where there is parameter heterogeneity; and firm‐specific OLS, when the impact of unobservable factors is likely to cause omitted variables difficulties.Practical implicationsThe paper provides practical insights to researchers with respect to the appropriate ways of utilising the considerable benefits of panel estimation methodologies while simultaneously avoiding common errors.Originality/valueThis study presents guidance in a relatively non‐technical manner on an issue which has not received sufficient attention in the accounting and managerial finance literature to date, namely the procedures to follow in order to choose appropriate estimation methodologies when dealing with panel datasets.