Recently, the research result of Researcher Zhang Ziyao from the Center for International Cooperation and Disciplinary Innovation of Income Distribution and Public Finance and his collaborator (Huang Wei, a tenured associate professor at the China Center for Economic Research and the National School of Development, Peking University), titled "Control Variable Selection in Empirical Research: Theory and Principles", was officially published in the top-notch Journal of Management World, No. 10, 2025.
Highlights
The selection of control variables is a critical determinant of the validity of causal inference in empirical research. This paper develops a systematic framework that clarifies the conceptual foundations, selection criteria, and practical guidelines for their use. Leveraging the potential outcomes framework and a decomposition of linear regression estimators, we distinguish two essential functions of controls: causal identification and statistical inference. For identification, good controls enable observational studies to approximate randomized experiments by conditioning on relevant strata, whereas bad controls may generate selection bias. For inference, suitable controls attenuate noise and enhance precision, while poor choices inflate estimation error and weaken statistical power. Whether a control is "good" or "bad" depends on its position within the causal structure, which should be grounded in theory and institutional context rather than data-driven selection. Building on this insight, we classify common types of controls and articulate five principles for applied research: grounding control selection in causal structure, rigorously addressing bad controls, ensuring overlap, weighing trade-offs in complex settings, and avoiding misinterpretation of coefficients. Practical recommendations are provided accordingly. This framework offers an operational methodology to improve research design and strengthens the credibility, transparency, and replicability of empirical work in the social sciences.
About the Author
Zhang Ziyao is an associate professor at the School of Public Finance and Taxation, Zhongnan University of Economics and Law, a Young Scholar of the Wenlan Scholar Program, and a researcher at the Center for International Cooperation and Disciplinary Innovation of Income Distribution and Public Finance. He has been selected for the High-Level Talent Program of Hubei Province. He holds a doctorate in economics from Renmin University of China. His research areas include public economics and industrial economics, with a focus on structural estimation methods for firm heterogeneity, resource allocation efficiency, factor income distribution, public policy, and corporate behavior. His research findings have been published in major academic journals such as Economic Research Journal, Journal of Management World, China Economic Quarterly, and The Journal of World Economy. He has presided over the young scientists fund projects of the National Natural Science Foundation of China and general programs of the China Postdoctoral Science Foundation, and participated in several major projects of the National Social Science Fund of China and general programs of the National Natural Science Foundation of China. Furthermore, he has received multiple honors, including the Hubei Provincial Social Science Outstanding Achievement Award, the Mundell-Huang Prize on Economics, the Contemporary Economics Doctoral Innovation Project, and the Tan Chongtai Development Economics Scholarship.