Panel Data in Economic Research

Abstract

This article reviews recent literature (2020 – 2025) on the use of panel data econometrics in economic research, with a focus on growth-related studies in Latin American economies. Using an exploratory-descriptive review design and searches in Google Scholar, Elsevier, Dialnet, and Redalyc, the paper synthesises core panel specifications and estimation approaches. It outlines the general linear panel model and discusses pooled OLS, fixed effects (including LSDV, within, and first-difference estimators), and random effects (error-components) models, highlighting their assumptions, strengths, and limitations. The review also summarises key specifications and diagnostic tests commonly used in applied work (e.g., Breusch–Pagan/LM, Hausman, Wooldridge, and tests for heteroscedasticity). Finally, the article briefly introduces dynamic panel models (Arellano–Bond type GMM) and panel cointegration frameworks, which are frequently, employed extensions in growth-oriented empirical research.