Abstract
The recent expansion in the availability of longitudinal and granular panel data has led to a significant increase in the adoption of dynamic panel data (DPD) models in empirical Operations Management (OM) research. To explore this trend, we conducted a comprehensive literature review of empirical OM studies published from 2012 to 2024 in four leading OM journals. Our findings reveal a marked rise in the utilization of DPD models in recent years, with 49 out of 66 (74.2%) reviewed studies published between 2020 and 2024. Despite this growing adoption, the estimation of DPD models presents notable methodological challenges compared to standard panel fixed‐effects models, creating barriers to their wider use in empirical OM research. Our simulation shows that the widely used least‐squares dummy variable (LSDV) estimator can be significantly biased when estimating DPD models. Variations in methodological choices can lead to differing results and conclusions, raising concerns about the rigor of current practices. To address these issues, we provide a technical review of recent estimation methods for DPD models, highlighting their underlying assumptions and initial conditions. Among these, we recommend the Generalized Method of Moments (GMM) for its flexibility and less restrictive assumptions, which have contributed to its popularity across various research fields. However, the application of GMM requires a careful, iterative process, which many OM researchers may not be familiar with. To support the effective use of GMM in DPD models in empirical OM research, we propose a practical framework that guides researchers through this process and demonstrates its practical use through real‐world data from supply base management. This framework aims to enhance the reliability and accuracy of estimating DPD models, facilitating more robust empirical research in the field of OM.