nep-dcm New Economics Papers
on Discrete Choice Models
Issue of 2025–10–13
seventeen papers chosen by
Edoardo Marcucci, Università degli studi Roma Tre


  1. Random preference model By Mohammad Ghaderi; Kamel Jedidi; Miłosz Kadziński; Bas Donkers
  2. Functional effects models: Accounting for preference heterogeneity in panel data with machine learning By Nicolas Salvad\'e; Tim Hillel
  3. Beyond Softmax: A New Perspective on Gradient Bandits By Emerson Melo; David M\"uller
  4. A Tale of Two News: The Impact of Media Outlets on Consumption Choices By Juan Carlos Angulo; Aldo Gutierrez-Mendieta
  5. A data fusion approach for mobility hub impact assessment and location selection: integrating hub usage data into a large-scale mode choice model By Xiyuan Ren; Joseph Y. J. Chow
  6. How do rising temperatures affect inflation expectations? By Georgarakos, Dimitris; Kenny, Geoff; Meyer, Justus; van Rooij, Maarten
  7. Hiring Preferences for Military Veterans: Evidence from a Stated Choice Experiment By Bäckström, Peter
  8. Non-Bayesian Learning in Misspecied Models By Sebastian Bervoets; Mathieu Faure; Ludovic Renou
  9. Deterring Industrial Vessels from African Coastal Fisheries By Agarwal, Aishwarya; Englander, Gabriel
  10. Inducing State Anxiety in LLM Agents Reproduces Human-Like Biases in Consumer Decision-Making By Ziv Ben-Zion; Zohar Elyoseph; Tobias Spiller; Teddy Lazebnik
  11. The evolution of insurance purchasing behavior: an empirical study on the adoption of online channels in Poland By Gabriela Wojak; Ernest G\'orka; Micha{\l} \'Cwi\k{a}ka{\l}a; Dariusz Baran; Rafa{\l} \'Swiniarski; Katarzyna Olszy\'nska; Piotr Mrzyg{\l}\'od; Maciej Frasunkiewicz; Piotr R\k{e}czajski; Daniel Zawadzki; Jan Piwnik
  12. Elicitability By Yaron Azrieli; Christopher Chambers; Paul Healy; Nicolas Lambert
  13. The Bayesian Origin of the Probability Weighting Function in Human Representation of Probabilities By Xin Tong; Thi Thu Uyen Hoang; Xue-Xin Wei; Michael Hahn
  14. Can Machine Learning Algorithms Outperform Traditional Models for Option Pricing? By Georgy Milyushkov
  15. Forecasting Inflation Based on Hybrid Integration of the Riemann Zeta Function and the FPAS Model (FPAS + $\zeta$): Cyclical Flexibility, Socio-Economic Challenges and Shocks, and Comparative Analysis of Models By Davit Gondauri
  16. The Choice of Political Advisors By Migrow, Dimitri; Park, Hyungmin; Squintani, Francesco
  17. Uncovering Representation Bias for Investment Decisions in Open-Source Large Language Models By Fabrizio Dimino; Krati Saxena; Bhaskarjit Sarmah; Stefano Pasquali

  1. By: Mohammad Ghaderi; Kamel Jedidi; Miłosz Kadziński; Bas Donkers
    Abstract: We introduce the Random Preference Model (RPM), a non-parametric and flexible discrete choice model. RPM is a rank-based stochastic choice model where choice options have multi-attribute representations. It takes preference orderings as the main primitive and models choices directly based on a distribution over partial or complete preference orderings over a ï¬ nite set of alternatives. This enables it to capture context-dependent behaviors while maintaining adherence to the regularity axiom. In its output, it provides a full distribution over the entire preference parameter space, accounting for inferential uncertainty due to limited data. Each ranking is associated with a subspace of utility functions and assigned a probability mass based on the expected log-likelihood of those functions in explaining the observed choices. We propose a two-stage estimation method that separates the estimation of ranking-level probabilities from the inference of preference parameters variation for a given ranking, employing Monte Carlo integration with subspace-based sampling. To address the factorial complexity of the ranking space, we introduce scalable approximation strategies: restricting the support of RPM to a randomly sampled or orthogonal basis subset of rankings and using partial permutations (top-k lists). We demonstrate that RPM can effectively recover underlying preferences, even in the presence of data inconsistencies. The experimental evaluation based on real data conï¬ rms RPM variants consistently outperform multinomial logit (MNL) in both in-sample ï¬ t and holdout predictions across different training sizes, with support-restricted and basis-based variants achieving the best results under data scarcity. Overall, our ï¬ ndings demonstrate RPM’s flexibility, robustness, and practical relevance for both predictive and explanatory modeling.
    Keywords: choice models, nonparametric modeling, rankings, context-dependent preference, random utility
    JEL: C35 C14 C15
    Date: 2025–07
    URL: https://d.repec.org/n?u=RePEc:upf:upfgen:1913
  2. By: Nicolas Salvad\'e; Tim Hillel
    Abstract: In this paper, we present a general specification for Functional Effects Models, which use Machine Learning (ML) methodologies to learn individual-specific preference parameters from socio-demographic characteristics, therefore accounting for inter-individual heterogeneity in panel choice data. We identify three specific advantages of the Functional Effects Model over traditional fixed, and random/mixed effects models: (i) by mapping individual-specific effects as a function of socio-demographic variables, we can account for these effects when forecasting choices of previously unobserved individuals (ii) the (approximate) maximum-likelihood estimation of functional effects avoids the incidental parameters problem of the fixed effects model, even when the number of observed choices per individual is small; and (iii) we do not rely on the strong distributional assumptions of the random effects model, which may not match reality. We learn functional intercept and functional slopes with powerful non-linear machine learning regressors for tabular data, namely gradient boosting decision trees and deep neural networks. We validate our proposed methodology on a synthetic experiment and three real-world panel case studies, demonstrating that the Functional Effects Model: (i) can identify the true values of individual-specific effects when the data generation process is known; (ii) outperforms both state-of-the-art ML choice modelling techniques that omit individual heterogeneity in terms of predictive performance, as well as traditional static panel choice models in terms of learning inter-individual heterogeneity. The results indicate that the FI-RUMBoost model, which combines the individual-specific constants of the Functional Effects Model with the complex, non-linear utilities of RUMBoost, performs marginally best on large-scale revealed preference panel data.
    Date: 2025–09
    URL: https://d.repec.org/n?u=RePEc:arx:papers:2509.18047
  3. By: Emerson Melo; David M\"uller
    Abstract: We establish a link between a class of discrete choice models and the theory of online learning and multi-armed bandits. Our contributions are: (i) sublinear regret bounds for a broad algorithmic family, encompassing Exp3 as a special case; (ii) a new class of adversarial bandit algorithms derived from generalized nested logit models \citep{wen:2001}; and (iii) \textcolor{black}{we introduce a novel class of generalized gradient bandit algorithms that extends beyond the widely used softmax formulation. By relaxing the restrictive independence assumptions inherent in softmax, our framework accommodates correlated learning dynamics across actions, thereby broadening the applicability of gradient bandit methods.} Overall, the proposed algorithms combine flexible model specification with computational efficiency via closed-form sampling probabilities. Numerical experiments in stochastic bandit settings demonstrate their practical effectiveness.
    Date: 2025–10
    URL: https://d.repec.org/n?u=RePEc:arx:papers:2510.03979
  4. By: Juan Carlos Angulo (Department of Economics, Universidad Iberoamericana); Aldo Gutierrez-Mendieta (School of Social Sciences and Government, Tecnologico de Monterrey)
    Abstract: How does exposure to multiple, sometimes conflicting, news stories influence individuals’ consumption choices? We address this question through a survey experiment in which participants are randomly assigned to receive one or two news headlines related to avocado consumption.These vignettes present either a positive framing (health benefits) or negative consequences, such as environmental damage or links to organized crime. After each exposure, we measure participants’ willingness to pay (WTP) for a standard product using a contingent valuation method. Our findings show that the content and emotional framing of information matter more than the number of exposures. The strongest effect comes from conflict-related news, which significantly reduces WTP. When participants receive two headlines, the second exposure tends to drive the response, especially in cases with conflicting information. Overall, these results suggest that consumers react not only to the presence of information but also to how it is framed. Even when the product remains unchanged, the tone and content of the message influence economic decisions in subtle but measurable ways.
    JEL: D91 Q18 D83
    Date: 2025–09–17
    URL: https://d.repec.org/n?u=RePEc:smx:wpaper:2025005
  5. By: Xiyuan Ren; Joseph Y. J. Chow
    Abstract: As cities grapple with traffic congestion and service inequities, mobility hubs offer a scalable solution to align increasing travel demand with sustainability goals. However, evaluating their impacts remains challenging due to the lack of behavioral models that integrate large-scale travel patterns with real-world hub usage. This study presents a novel data fusion approach that incorporates observed mobility hub usage into a mode choice model estimated with synthetic trip data. We identify trips potentially affected by mobility hubs and construct a multimodal sub-choice set, then calibrate hub-specific parameters using on-site survey data and ground truth trip counts. The enhanced model is used to evaluate mobility hub impacts on potential demand, mode shift, reduced vehicle miles traveled (VMT), and increased consumer surplus (CS). We apply this method to a case study in the Capital District, NY, using data from a survey conducted by the Capital District Transportation Authority (CDTA) and a mode choice model estimated using Replica Inc. synthetic data. The two implemented hubs located near UAlbany Downtown Campus and in Downtown Cohoes are projected to generate 8.83 and 6.17 multimodal trips per day, reduce annual VMT by 20.37 and 13.16 thousand miles, and increase daily CS by $4, 000 and $1, 742, respectively. An evaluation of potential hub candidates in the Albany-Schenectady-Troy metropolitan area with the estimated models demonstrates that hubs located along intercity corridors and at urban peripheries, supporting park-and-ride P+R patterns, yield the most significant behavioral impacts.
    Date: 2025–10
    URL: https://d.repec.org/n?u=RePEc:arx:papers:2510.08366
  6. By: Georgarakos, Dimitris; Kenny, Geoff; Meyer, Justus; van Rooij, Maarten
    Abstract: Global temperatures are rising at an alarming pace and public awareness of climate change is increasing, yet little is known about how these developments affect consumer expectations. We address this gap by conducting a series of experiments within a large-scale, population-representative survey of euro area consumers. We randomly assign consumers to hypothetical global temperature change scenarios, after which we elicit their expectations for inflation and key macroeconomic indicators under these conditions. We find that a 0.5°C rise in global temperatures leads to a 0.65 percentage point increase in five-year-ahead inflation expectations, with effects particularly pronounced among consumers with greater awareness of climate change. Additionally, respondents expect adverse impacts of global warming on economic growth, employment, public debt, tax burdens, and their well-being. Despite these pessimistic expectations, many consumers demonstrate limited willingness to pay for mitigating further temperature increases. Instead, they place primary responsibility for climate action on governments. Our findings underscore the interplay between climate change and economic expectations, highlighting the potential implications for monetary and fiscal policy in a warming world. JEL Classification: D12, E31, E52, H31, Q54
    Keywords: climate change, consumer expectations, Consumer Expectations Survey (CES), global warming, Randomized Control Trial (RCT)
    Date: 2025–10
    URL: https://d.repec.org/n?u=RePEc:ecb:ecbwps:20253132
  7. By: Bäckström, Peter (Department of Economics, Umeå University)
    Abstract: Successfully finding and maintaining civilian employment is crucial for military veterans' reintegration into civilian society. This paper examines how formerly deployed Swedish military veterans are treated in civilian hiring situations using data from a stated-choice experiment. Approximately 1, 000 survey respondents were asked to imagine advising on recruitment for a job similar to their own and to choose repeatedly between two hypothetical job applicants, some of whom were described as military veterans. The results indicate that former military deployment can be both an advantage and a disadvantage in job applications, depending on the occupational field and recruiter characteristics. Military veterans who served as infantry soldiers face obstacles when applying for jobs in social professions, especially if the recruiter is female or lacks personal experience with military veterans. Conversely, veterans who served in military staff positions are more likely to be called for an interview, particularly if the recruiter is male, has personal experience with military veterans, or if the job is for a managerial position. These findings suggest that the impact of military deployment on job prospects is highly context-dependent.
    Keywords: military veterans; hiring; choice experiment; hiring bias
    JEL: C91 H56 J24 J71
    Date: 2025–10–02
    URL: https://d.repec.org/n?u=RePEc:hhs:umnees:1037
  8. By: Sebastian Bervoets (Aix-Marseille Univ., CNRS, AMSE, Marseille, France); Mathieu Faure (Aix-Marseille Univ., CNRS, AMSE, Marseille, France); Ludovic Renou (ASU, QMUL and CEPR)
    Abstract: Deviations from Bayesian updating are traditionally categorized as biases, errors, or fallacies, thus implying their inherent “sub-optimality.” We offer a more nuanced view. In learning problems with misspecified models, we show that some non-Bayesian updating can outperform Bayesian updating.
    Keywords: learning, Bayesian, consistency
    JEL: C73 D82
    Date: 2025–09
    URL: https://d.repec.org/n?u=RePEc:aim:wpaimx:2513
  9. By: Agarwal, Aishwarya; Englander, Gabriel
    Abstract: Most African coastal nations prohibit industrial vessels from fishing near their shores; these Inshore Exclusion Zones (IEZs) reserve the most productive locations for small-scale, artisanal fishers. However, previous descriptive research suggests that non-compliance by industrial vessels prevents IEZs from benefiting African economies, food security, and fish stocks. Radar data released in 2024 detect industrial vessels without selection, enabling the first causal evaluation of African IEZs. First, regression discontinuity estimates reveal that 6 of 20 African countries successfully deter industrial fishing vessels (Nigeria, Sierra Leone, Liberia, Mauritania, Ghana, and Guinea). Second, bunching estimators obtain counterfactual vessel distributions that are uncontaminated by spillovers outside IEZ boundaries. Third, extensive‑margin effects are captured by calibrating a discrete choice vessel location model to the bunching estimates. Back‑of‑the‑envelope bioeconomic calculations indicate that IEZs increase annual artisanal fisher catch by 324 thousand tons—enough to meet key micronutrient requirements for 6.3 million people—without reducing industrial catch.
    Date: 2025–09–29
    URL: https://d.repec.org/n?u=RePEc:wbk:wbrwps:11222
  10. By: Ziv Ben-Zion; Zohar Elyoseph; Tobias Spiller; Teddy Lazebnik
    Abstract: Large language models (LLMs) are rapidly evolving from text generators to autonomous agents, raising urgent questions about their reliability in real-world contexts. Stress and anxiety are well known to bias human decision-making, particularly in consumer choices. Here, we tested whether LLM agents exhibit analogous vulnerabilities. Three advanced models (ChatGPT-5, Gemini 2.5, Claude 3.5-Sonnet) performed a grocery shopping task under budget constraints (24, 54, 108 USD), before and after exposure to anxiety-inducing traumatic narratives. Across 2, 250 runs, traumatic prompts consistently reduced the nutritional quality of shopping baskets (Change in Basket Health Scores of -0.081 to -0.126; all pFDR
    Date: 2025–08
    URL: https://d.repec.org/n?u=RePEc:arx:papers:2510.06222
  11. By: Gabriela Wojak; Ernest G\'orka; Micha{\l} \'Cwi\k{a}ka{\l}a; Dariusz Baran; Rafa{\l} \'Swiniarski; Katarzyna Olszy\'nska; Piotr Mrzyg{\l}\'od; Maciej Frasunkiewicz; Piotr R\k{e}czajski; Daniel Zawadzki; Jan Piwnik
    Abstract: This paper examines how Polish consumers are adapting to online insurance purchasing channels and what factors influence their preferences. Drawing on a structured survey of 100 respondents with varied demographic profiles, the study explores purchasing frequency, channel usage, price sensitivity, trust, and decision-making behaviors. Results indicate a clear shift toward digital tools, with many consumers valuing the speed, convenience, and transparency of online platforms, particularly for simple insurance products. However, barriers remain, including concerns about data security, lack of personal guidance, and difficulty understanding policy terms. A hybrid model is emerging, where online tools are used for research and comparison, while traditional agents are consulted for complex decisions. Respondents emphasized the importance of trust and personal contact, showing that emotional and psychological factors still play a role in digital adoption. Price was the dominant decision factor, but many consumers also prioritized service quality and reliability. The study concludes that insurers should invest in user-friendly digital experiences while maintaining human support options. Strategic omnichannel integration is recommended to meet diverse customer needs and reduce digital exclusion. Limitations of the study include a modest sample size and focus on the Polish market. Future research should investigate the role of AI in digital distribution, segment preferences by insurance type, and analyze trends across different regions or age groups. This paper adds empirical value to the understanding of insurance distribution and consumer behavior in digitally transforming financial markets.
    Date: 2025–10
    URL: https://d.repec.org/n?u=RePEc:arx:papers:2510.07933
  12. By: Yaron Azrieli; Christopher Chambers; Paul Healy; Nicolas Lambert
    Abstract: An analyst is tasked with producing a statistical study. The analyst is not monitored and is able to manipulate the study. He can receive payments contingent on his report and trusted data collected from an independent source, modeled as a statistical experiment. We describe the information that can be elicited with appropriately shaped incentives, and apply our framework to a variety of common statistical models. We then compare experiments based on the information they enable us to elicit. This order is connected to, but different from, the Blackwell order. Data preferred for estimation are also preferred for elicitation, but not conversely. Our results shed light on how using data as incentive generator in payment schemes differs from using data for statistical inference.
    Date: 2025–10
    URL: https://d.repec.org/n?u=RePEc:arx:papers:2510.00879
  13. By: Xin Tong; Thi Thu Uyen Hoang; Xue-Xin Wei; Michael Hahn
    Abstract: Understanding the representation of probability in the human mind has been of great interest to understanding human decision making. Classical paradoxes in decision making suggest that human perception distorts probability magnitudes. Previous accounts postulate a Probability Weighting Function that transforms perceived probabilities; however, its motivation has been debated. Recent work has sought to motivate this function in terms of noisy representations of probabilities in the human mind. Here, we present an account of the Probability Weighting Function grounded in rational inference over optimal decoding from noisy neural encoding of quantities. We show that our model accurately accounts for behavior in a lottery task and a dot counting task. It further accounts for adaptation to a bimodal short-term prior. Taken together, our results provide a unifying account grounding the human representation of probability in rational inference.
    Date: 2025–10
    URL: https://d.repec.org/n?u=RePEc:arx:papers:2510.04698
  14. By: Georgy Milyushkov
    Abstract: This study investigates the application of machine learning techniques, specifically Neural Networks, Random Forests, and CatBoost for option pricing, in comparison to traditional models such as Black-Scholes and Heston Model. Using both synthetically generated data and real market option data, each model is evaluated in predicting the option price. The results show that machine learning models can capture complex, non-linear relationships in option prices and, in several cases, outperform both Black-Scholes and Heston models. These findings highlight the potential of data-driven methods to improve pricing accuracy and better reflect market dynamics.
    Date: 2025–10
    URL: https://d.repec.org/n?u=RePEc:arx:papers:2510.01446
  15. By: Davit Gondauri
    Abstract: Inflation forecasting is a core socio-economic challenge in modern macroeconomic modeling, especially when cyclical, structural, and shock factors act simultaneously. Traditional systems such as FPAS and ARIMA often struggle with cyclical asymmetry and unexpected fluctuations. This study proposes a hybrid framework (FPAS + $\zeta$) that integrates a structural macro model (FPAS) with cyclical components derived from the Riemann zeta function $\zeta(1/2 + i t)$. Using Georgia's macro data (2005-2024), a nonlinear argument $t$ is constructed from core variables (e.g., GDP, M3, policy rate), and the hybrid forecast is calibrated by minimizing RMSE via a modulation coefficient $\alpha$. Fourier-based spectral analysis and a Hidden Markov Model (HMM) are employed for cycle/phase identification, and a multi-criteria AHP-TOPSIS scheme compares FPAS, FPAS + $\zeta$, and ARIMA. Results show lower RMSE and superior cyclical responsiveness for FPAS + $\zeta$, along with early-warning capability for shocks and regime shifts, indicating practical value for policy institutions.
    Date: 2025–10
    URL: https://d.repec.org/n?u=RePEc:arx:papers:2510.02966
  16. By: Migrow, Dimitri (University of Edinburgh); Park, Hyungmin (University of Warwick); Squintani, Francesco (University of Warwick)
    Abstract: We study a leader’s choice of advisors, balancing political alignment, informational competence, and diversity of views. The leader can consult one or two advisors : one is politically aligned but less informed or shares potentially redundant information; the other is better informed but more biased. The leader’s optimal strategy can exhibit reversals. If both advisors are initially consulted, increasing the bias of the more biased advisor may cause the leader to exclude the aligned advisor to preserve truthfulness from the informed one. As bias rises further, the leader ultimately replaces the informed advisor if his bias becomes too large. When the leader is uncertain about the bias of the more informed advisor, increasing the chance of alignment can justify consulting both advisors.
    Date: 2025
    URL: https://d.repec.org/n?u=RePEc:wrk:warwec:1582
  17. By: Fabrizio Dimino; Krati Saxena; Bhaskarjit Sarmah; Stefano Pasquali
    Abstract: Large Language Models are increasingly adopted in financial applications to support investment workflows. However, prior studies have seldom examined how these models reflect biases related to firm size, sector, or financial characteristics, which can significantly impact decision-making. This paper addresses this gap by focusing on representation bias in open-source Qwen models. We propose a balanced round-robin prompting method over approximately 150 U.S. equities, applying constrained decoding and token-logit aggregation to derive firm-level confidence scores across financial contexts. Using statistical tests and variance analysis, we find that firm size and valuation consistently increase model confidence, while risk factors tend to decrease it. Confidence varies significantly across sectors, with the Technology sector showing the greatest variability. When models are prompted for specific financial categories, their confidence rankings best align with fundamental data, moderately with technical signals, and least with growth indicators. These results highlight representation bias in Qwen models and motivate sector-aware calibration and category-conditioned evaluation protocols for safe and fair financial LLM deployment.
    Date: 2025–10
    URL: https://d.repec.org/n?u=RePEc:arx:papers:2510.05702

This nep-dcm issue is ©2025 by Edoardo Marcucci. It is provided as is without any express or implied warranty. It may be freely redistributed in whole or in part for any purpose. If distributed in part, please include this notice.
General information on the NEP project can be found at https://nep.repec.org. For comments please write to the director of NEP, Marco Novarese at <director@nep.repec.org>. Put “NEP” in the subject, otherwise your mail may be rejected.
NEP’s infrastructure is sponsored by the School of Economics and Finance of Massey University in New Zealand.