 Corresponding author: Shuvo Kumar Mallik Copyright © 2025 Author(s) retain the copyright of this article. This article is published under the terms of the Creative Commons Attribution Liscense 4.0. Econometric advances in causal inference: The machine learning revolution Shuvo Kumar Mallik 1, * , Imran Uddin 2 , Sadia Maliha Trisha 3 , Md. Morshedul Hasan 4 and M Abeedur Rahman 5 1 Department of Economics, Southeast University, Dhaka, Bangladesh. 2 A2Z Finance Australia (Easy Mortgage Solutions Australia), Australia. 3 Dublin Business School, Dublin, Ireland. 4 School of Business Roya University, Dhaka, Bangladesh. 5 Assistant Professor, Department of Economics, Southeast university, Dhaka, Bangladesh. GSC Advanced Research and Reviews, 2025, 22(03), 229-244 Publication history: Received on 08 February 2025; revised on 15 March 2025; accepted on 17 March 2025 Article DOI: https://doi.org/10.30574/gscarr.2025.22.3.0082 Abstract This is one of the challenges that new and fast-growing econometric literature is beginning to tackle in addressing causal inference problems with machine learning methods. Yet, empirical economics still has not really made use of the strengths of these modern approaches. Here, we revisit groundbreaking empirical work through the perspective of causal machine learning methods to connect econometric theory with applied economics. In particular, we will cover double machine learning, causal forests, and more general machine learning methodologies, both in the setting of average treatment effects and heterogeneous treatment effects. We demonstrate the application of these methods in diverse settings and discuss their significance and additional benefits relative to classical approaches that were utilized in the original studies. Keywords: Econometric Literature; Machine Learning Methods; Value; Applied economics; Empirical work 1. Introduction Estimating the causal effect of some intervention on an outcome of interest is one of the main goals of empirical research in economics. However, omitted variable bias can lead to misleading estimates in observational studies, so it is critical to include a wide array of control variables. Even with a relatively small number of raw covariates, interactions and transformations can be added that dramatically increase the number of controls in a regression. In such settings, machine learning (ML) methods are instrumental. However, standard ML prediction models are, at a fundamental level, not designed to achieve the same goals as most empirical economic research. Since ML techniques are optimized to predict outcomes for test samples, model selection is performed so that prediction errors in hold-out test sets are minimized. In contrast, most empirical economic research aims to learn about the underlying relationships that underlie policy decisions, not to improve predictive accuracy. This fundamental distinction is why you cannot unthinkingly use techniques from the ML world, which is fundamentally about predicting, to causal inference problems, a misuse that can yield biased estimates. Still, converging econometric literature is making impressive progress in applying ML methods to causal inference (Athey et al., 2018; Chernozhukov, Chetverikov, Facchinei et al., 2018; Chernozhukov, Demirer, et al., 2018; Wager & Athey, 2018). This literature offers new insights and theoretical results that speak to both ML and econometrics/statistics. Yet, so far, modern causal estimation techniques have not been fully adopted by empirical economics. This paper aims to showcase researchers with empirical evidence on the benefits of causal machine learning context in the real world. We do this by revisiting a number of influential studies applying causal ML methods, and we