Learning Preferences from Conjoint Data: A Structural Deep Learning Approach

2026-04-12

AI Summary

This study addresses the limitations of traditional conjoint analysis, which often relies on nonparametric causal estimation and struggles to capture heterogeneity in individual preferences. The authors propose a novel structured deep learning framework that embeds deep neural networks within a random utility logit model, allowing preference parameters to flexibly vary with respondent characteristics. By integrating debiased machine learning techniques, the approach enables robust inference of average preference effects while overcoming the constraints of conventional parametric assumptions. Applied to three canonical conjoint experiments, the method reveals nuanced patterns: while the overall gender effect is near zero, 83% of respondents exhibit a preference for female candidates; anti-authoritarian attitudes are widespread yet highly heterogeneous in intensity; and support for progressive taxation is prevalent across partisan groups. This framework substantially enhances the identification of rich preference structures obscured by aggregate averages.

Technology Category

• Machine Learning: Learning Preferences or Rankings
• Knowledge Representation and Reasoning: Preferences
• Humans and AI: Learning Human Values and Preferences

Application Category

• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Conjoint experiments randomize multidimensional profiles, offering a powerful design for recovering structural preference parameters -- including marginal rates of substitution, willingness to pay, and the distribution of preferences across a population. Yet the dominant approach in political science has focused on nonparametric causal estimands that do not leverage this potential. We propose a structural approach that embeds a deep neural network within a random utility logit model, allowing preference parameters to vary as a fully flexible function of respondent characteristics. The neural network addresses the concern that a parametric specification may not capture the true data generating process, while double/debiased machine learning provides valid inference on average preference parameters. We apply our method to three prominent conjoint studies and find rich preference heterogeneity masked by reduced-form averages: a near-zero gender effect coexists with 83% preferring female candidates, opposition to undemocratic behavior is near-universal but varies sharply in intensity, and progressive tax preferences cut across every partisan subgroup.

Problem

• conjoint analysis
• preference heterogeneity
• structural estimation
• random utility model
• willingness to pay

Innovation

• structural deep learning
• conjoint analysis
• preference heterogeneity
• random utility model
• double machine learning