Using Machine Learning to Generate, Clarify, and Improve Economic Models

2025-08-26

AI Summary

While machine learning models achieve high predictive accuracy for economic behavior, they lack causal interpretability and thus fail to support theoretical development or policy counterfactual analysis. To address this, we propose an “interpretability-theory-driven” paradigm that systematically embeds economic priors—such as theoretical constraints and causal structural assumptions—into the ML pipeline, integrating adversarial training, symbolic induction, and large language model–assisted modeling to enable automated model discovery, clarification, and validation. Methodologically, our approach transcends purely data-driven learning: it preserves high predictive performance while explicitly generating interpretable, mechanism-based models that uncover causal pathways and robust regularities among variables. Our key contributions are threefold: (1) the first systematic integration of ML with theory construction objectives; (2) significantly enhanced out-of-distribution generalizability and policy counterfactual reasoning capability; and (3) advancement of data-informed economic theory innovation.

Technology Category

• Game Theory and Economic Paradigms: Adversarial Learning
• Machine Learning: Causal Learning
• Humans and AI: Human-in-the-loop Machine Learning

Application Category

• Economics, Online Markets and Human Computation: Economic ramifications for generative AI infrastructure and applications
• Semantics and Knowledge: Methods, algorithms and applications for the development of semantic models, knowledge graphs and other forms of structured data models with machine-interpretable semantics
• User Modeling, Personalization and Recommendation: Explainable and interpretable methods for personalization

Abstract

Machine learning algorithms can now outperform classic economic models in predicting quantities ranging from bargaining outcomes, to choice under uncertainty, to an individual's future jobs and wages. Yet this predictive accuracy comes at a cost: most machine learning algorithms function as black boxes, offering little insight into emph{why} outcomes occur. This article asks whether machine learning can guide the development of new economic theories. Economic models serve an important purpose beyond prediction -- they uncover the general mechanisms behind observed behaviors. A model that identifies the causal pathways of economic development is more valuable than one that merely predicts which countries will escape poverty, because it enables policymakers to encourage that development in countries where it might not have happened otherwise. Similarly, a model that predicts imperfectly across many domains can be more valuable than one that is highly accurate in a specific domain, since the former allows insights and data obtained from one setting to inform decisions and policy in another. Applying machine learning algorithms off-the-shelf is unlikely to yield such models. But recent work shows that, when reconceived with the aims of an economic modeler in mind, machine learning methods can improve both prediction and understanding. These approaches range from adversarially training algorithms to expose the limits of existing models, to imposing economic theory as a constraint on algorithmic search. Advances in large language models complement these strategies and open new research directions.

Problem

• Using machine learning to generate new economic theories
• Overcoming black box limitations for economic insight
• Improving prediction and understanding in economic models

Innovation

• Machine learning generates and improves economic models
• Adversarial training exposes limits of existing models
• Economic theory constrains algorithmic search for insights