A Deep Learning Approach to Heterogeneous Consumer Aesthetics in Fast Fashion (Working Paper)
Aesthetics drives product differentiation in industries such as fashion, interior decor, luxury goods, real estate and hospitality. However, visual differentiation is hard to encode in formal economic analysis. This paper develops a methodology to parse, process and model the aesthetic dimension and applies it to fast fashion. It works with millions of purchase records from H&M in the Netherlands, including product images, text descriptions, prices, and consumer demographics. I fine-tune Fashion CLIP embeddings with a three-tower approach that builds separate channels for product visuals and text, consumer history, and price, which makes downstream analysis tractable and scalable. The embeddings feed a latent-class deep demand system that captures price and taste sensitivities through deep nets, recovers rich substitution patterns, reveals meaningful heterogeneity, and performs much better than competing alternatives. Then, a supply-side inversion recovers sensible markups and costs and supports conduct tests and counterfactuals on sustainability practices. I also estimate machine learning hedonic pricing models that perform much better than competing alternatives. This model allows us to construct quality-adjusted price indices, make it possible to price completely new designs, and with an Oaxaca-Blinder decomposition reveal the underlying sources of price changes. Finally, a Poisson event study around the COVID-19 lockdown shows that the range of demand responses across embedding-based product and user clusters exceeds anything recoverable from simple text-based attributes or demographic labels alone. All these findings show that high-dimensional embeddings can be first class citizens in disciplined economic and statistical analysis. The methodology is portable to any market where products are differentiated along sensory dimensions that are hard to encode but meaningfully important for consumer choices.
Who is more Bayesian: Humans or ChatGPT? (with John Rust, Chengjun Zhang, Tianshi Mu and Aaron Zhong)
We compare the performance of human and artificially intelligent (AI) decision makers in simple binary classification tasks where the optimal decision rule is given by Bayes Rule. We reanalyze choices of human subjects gathered from laboratory experiments conducted by El-Gamal and Grether and Holt and Smith. We confirm that while overall Bayes Rule represents the single best model for predicting human choices, subjects are heterogeneous and a significant share of them make suboptimal choices that reflect judgment biases described by Kahneman and Tversky. These biases include the "representativeness heuristic" (excessive weight on the evidence from the sample relative to the prior) and "conservatism" (excessive weight on the prior relative to the sample). We compare the performance of AI subjects gathered from recent versions of large language models (LLMs), including several versions of ChatGPT. These general-purpose generative AI chatbots are not specifically trained to excel in narrow decision-making tasks but are instead trained as "language predictors" using a large corpus of textual data from the web. We show that ChatGPT is also subject to biases that result in suboptimal decisions. However, we document a rapid evolution in the performance of ChatGPT, transitioning from sub-human performance in early versions (ChatGPT 3.5) to superhuman and nearly perfect Bayesian classifications in the latest versions (ChatGPT 4.0).
Structural Econometrics and Reinforcement Learning (Oxford Research Encyclopedia, with John Rust)
This survey article explores the synergies between structural econometrics and reinforcement learning. Structural econometrics interprets observed economic choices as optimal decisions under constraints, enabling counterfactual prediction of behavior under rule changes. Reinforcement learning offers a framework for learning optimal policies in complex multi-step problems through exploration and exploitation. We identify opportunities for cross-fertilization between these fields. Structural econometrics can leverage reinforcement learning algorithms to solve previously intractable high-dimensional economic models and games. Inverse reinforcement learning provides econometricians with new methods to recover agents' objective functions from observed behavior. Reinforcement learning, in particular bandits, can be enhanced by incorporating economic theory and structural assumptions, accelerating learning and improving sample complexity by orders of magnitude. We review methodological connections, demonstrate applications across finance, industrial organization, public policy, and marketing. Both fields create new tools for inference and decision-making while tackling the shared challenges of the curse of dimensionality, equilibrium multiplicity, and identification.
A Survey of Reinforcement Learning for Economics
This survey (re)introduces reinforcement learning methods to economists. The curse of dimensionality limits how far exact dynamic programming can be effectively applied, forcing us to rely on suitably "small" problems or our ability to convert "big" problems into smaller ones. While this reduction has been sufficient for many classical applications, a growing class of economic models resists such reduction. Reinforcement learning algorithms offer a natural, sample-based extension of dynamic programming, extending tractability to problems with high-dimensional states, continuous actions, and strategic interactions. I review the theory connecting classical planning to modern learning algorithms and demonstrate their mechanics through simulated examples in pricing, inventory control, strategic games, and preference elicitation. I also examine the practical vulnerabilities of these algorithms, noting their brittleness, sample inefficiency, sensitivity to hyperparameters, and the absence of global convergence guarantees outside of tabular settings. The successes of reinforcement learning remain strictly bounded by these constraints, as well as a reliance on accurate simulators. When guided by economic structure, reinforcement learning provides a remarkably flexible framework. It stands as an imperfect, but promising, addition to the computational economist's toolkit. A companion survey (Rust and Rawat, 2026b) covers the inverse problem of inferring preferences from observed behavior.
Algorithmic Collusion in Auctions: Evidence from Controlled Laboratory Experiments (WEAI Conference 2025)
Algorithms increasingly automate bidding in online auctions, raising concerns about tacit bid suppression and revenue shortfalls. Prior work identifies individual mechanisms behind algorithmic bid suppression, but it remains unclear which factors matter most and how they interact, and policy conclusions rest on algorithms unlike those deployed in practice. This paper develops a computational laboratory framework, based on factorial experimental designs and large-scale Monte Carlo simulation, that addresses bid suppression across multiple algorithm classes within a common methodology. Each simulation is treated as a black-box input-output observation; the framework varies inputs and ranks factors by association with outcomes, without explaining algorithms' internal mechanisms. Across six sub-experiments spanning Q-learning, contextual bandits, and budget-constrained pacing, the framework ranks the relative importance of auction format, competitive pressure, learning parameters, and budget constraints on seller revenue. The central finding is that structural market parameters dominate algorithmic design choices. In unconstrained settings, competitive pressure is the strongest predictor of revenue; under budget constraints, budget tightness takes over. The auction-format effect is context-dependent, favouring second-price under learning algorithms but reversing to favour first-price under budget-constrained pacing. Because the optimal format depends on the prevailing bidding technology, no single auction format is universally superior when bidders are algorithms, and applying format recommendations from one algorithm class to another leads to counterproductive design interventions.
Approximating Auction Equilibria with Reinforcement Learning (Working Paper)
Traditional methods for computing equilibria in auctions become computationally intractable as auction complexity increases, particularly in multi-item and dynamic auctions. This paper introduces a self-play based reinforcement learning approach that employs advanced algorithms such as Proximal Policy Optimization to approximate Bayes-Nash equilibria. This framework allows for continuous action spaces, high-dimensional information states, and delayed payoffs. Through self-play, these algorithms can learn robust and near-optimal bidding strategies in auctions with known equilibria, including those with symmetric and asymmetric valuations, private and interdependent values, and multi-round auctions.