|
on Discrete Choice Models |
| By: | Bruno Pellegrino |
| Abstract: | How does prior information affect discrete choice? In the seminal rational inattention model of Matějka and McKay (2015), multinomial logit arises from the discrete choice of agents who are uncertain about choice payoffs and who have access to a flexible information acquisition technology (RI-logit). A key limitation of this powerful framework is the lack of known solutions that allow the decision maker's prior information to vary across choices: obtaining such solutions has remained an open problem in this literature. In this paper, I solve the RI-logit model analytically for two related families of priors known respectively as Positive Stable and Tempered Stable distributions. In my solution, the decision maker's prior information enters the choice probabilities through a choice-specific parameter, which can be read as shifting either the ex-ante expected utility of a risk-neutral decision maker, or the ex-ante riskiness perceived by a risk-averse one. These results complete the RI-logit framework by separating prior information from information that is endogenously acquired, and expand its empirical applications by making it possible to study how choice probabilities react to changes in prior information. |
| JEL: | D11 D81 D83 |
| Date: | 2026–09 |
| URL: | https://d.repec.org/n?u=RePEc:nbr:nberwo:35702 |
| By: | Bartosz Jusypenko (University of Warsaw, Faculty of Economic Sciences) |
| Abstract: | Economic valuation of heritage conservation depends on what respondents expect to happen without the implementation of a proposed intervention, yet this no-program counterfactual is often left implicit. We examine whether explicitly directing respondents to consider potential heritage loss affects stated willingness to pay (WTP) for renovation of a building and whether the contribution of information about adaptive reuse depends on whether such loss enters the valuation decision. A contingent valuation survey of 452 adult residents living near Pückler Palace, a deteriorating historic building in Poland, used a 2 × 2 randomized design combining a loss-consideration prompt with information about a genuine proposal to adapt the renovated palace as the new premises of the Museum of Ceramics. WTP was elicited as the maximum mandatory one-off levy respondents would support within an advisory-referendum framework. Among respondents not assigned to the loss-prompt treatment, a post-valuation question identified spontaneous consideration of heritage loss. Two-part models separated the probability of positive WTP from the amount stated among positive responses. Respondents assigned to the loss-prompt treatment had approximately twice the odds of reporting positive WTP and stated 61% higher amounts among positive responses. Among unprompted respondents, spontaneous loss consideration—reported by 54%—was associated with 92% higher positive WTP. Reuse information appeared to contribute more when heritage loss was not otherwise considered. The findings show that common information about an asset’s current condition need not produce a common understanding of its expected future without the proposed intervention. Stated preference surveys should therefore specify this trajectory with sufficient clarity and credibility that respondents are evaluating a common underlying policy change. |
| Keywords: | cultural heritage, historic preservation, adaptive reuse, contingent valuation, stated preferences, prospect of loss |
| JEL: | C51 H43 Z18 Q51 |
| Date: | 2026 |
| URL: | https://d.repec.org/n?u=RePEc:war:wpaper:2026-27 |
| By: | Giulia Buccione (CEMFI, Centro de Estudios Monetarios y Financieros); Martín A. Rossi (Universidad de San Andrés) |
| Abstract: | We provide experimental evidence from rural Egypt on the benefits of incorporating sensory familiarity into the design of health interventions in developing countries. Using a “taste-preserving” filtration technology that mimics the local water profile, we find adoption rates of 91 percent, far exceeding the 50 percent ceiling previously documented for chlorinated water. Willingness to pay is 61 percent higher for filtered water than chlorinated water. Mechanism tests show that familiar taste strongly shapes perceptions of water’s healthiness, suggesting that alignment with imprinted sensory preferences improves uptake of health technologies. |
| Keywords: | Chlorination, water-borne diseases, field experiments, Egypt. |
| JEL: | D10 I10 C93 Q53 Z10 |
| Date: | 2026–07 |
| URL: | https://d.repec.org/n?u=RePEc:cmf:wpaper:wp2026_2609 |
| By: | Joel L. Horowitz; Sokbae (Simon) Lee |
| Abstract: | This paper presents a computationally efficient method for binary classification using Manski’s (1975, 1985) maximum score model when covariates are discretely distributed and parameters are partially but not point identified. We establish minimax-regret-optimal classification rules that take account of partial identification of the model’s parameters. We bound misclassification probabilities and expected excess regret induced by sampling uncertainty. We also describe an extension of our method to continuous covariates. Our approach avoids the computational difficulty of maximum score estimation by reformulating the problem as two linear programs. Compared to parametric and nonparametric methods, our method balances extrapolation ability with minimal distributional assumptions. Monte Carlo simulations and empirical applications demonstrate its effectiveness and practical relevance. |
| Date: | 2026–09–08 |
| URL: | https://d.repec.org/n?u=RePEc:azt:cemmap:14/26 |
| By: | Tomomichi Amano; Markus Brobeil; Andrew Rhodes; Stephan Seiler |
| Abstract: | We propose a model of consideration and choice that uses search data to flexibly estimate co-search patterns for every pair of products. Using simulations, we show that estimating co-search flexibly rather than modeling the drivers of consideration comes at a modest cost in precision but guards against misspecification. The method is therefore particularly valuable in online markets, where the drivers of consideration — such as webpage layout or product recommendations — are often unobserved. Relative to using purchase data alone, search data substantially improves the precision of cross-price elasticity estimates, because products that are searched together tend to be close substitutes and because browsing data is far more abundant than purchases. We apply our approach to data from an online retailer and find that most products have a small set of close substitutes and that accounting for substitution patterns when setting prices raises the retailer's average mark-up from 36.5% to 47.5% and increases its profits. |
| Keywords: | demand estimation, consideration sets, consumer search |
| JEL: | D12 D83 M31 |
| Date: | 2026 |
| URL: | https://d.repec.org/n?u=RePEc:ces:ceswps:_12988 |
| By: | Elie Tamer; Christopher D. Walker |
| Abstract: | This paper proposes a nonparametric Bayesian inference framework for partially identified discrete response models. The key observation is that these models map a reduced-form conditional choice probability to an identified set. Consequently, nonparametric Bayesian inference for the conditional probability mass function leads to Bayesian inference for the identified set. The inference framework nests conditional moment inequalities and linear systems with unknown coefficients as special cases. Importantly, our proposal does not require converting conditional moments into unconditional moments or discretizing covariates. We show that the posterior is consistent for the true identified set when the model is correctly specified, show that the posterior can consistently detect model misspecification, and show posterior consistency for a pseudo-identified set that is valid under misspecification. We also verify the assumptions for a class of priors based on Gaussian processes that we use to implement our proposal. These priors offer similar flexibility to frequentist partial identification methods, and are computationally attractive because posterior sampling can be performed in closed-form. We also show that many of the ideas in this paper extend to continuous responses and aggregated discrete responses (e.g., market shares). |
| Date: | 2026–08 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2608.25814 |
| By: | V\'aclav Kratochv\'il; Radim Jirou\v{s}ek; Kl\'ara \v{S}im\r{u}nkov\'a; Simona Ba\v{z}antov\'a |
| Abstract: | This data paper describes an anonymized release derived from controlled behavioral experiments on decision-making under risk (known probabilities) and ambiguous uncertainty (partially specified or unknown probabilities) in Ellsberg-type urn tasks. The release contains 4, 486 decision records from 246 adult participant records across two implementations (laboratory and online). In each task, participants reported their maximum stake to enter a subsequent lottery; these stake responses can be interpreted as proxies for willingness to pay within the implemented incentive mechanism. In a subset of tasks, participants also selected the color they wished to bet on. The online implementation additionally includes ex-post evaluations of perceived uncertainty across situations. The repository further provides detailed game definitions, realized lottery outcomes, participant instructions, and a data dictionary of variables. The source export contained 4, 544 decision records; 58 records belonging to five participants younger than 18 were excluded from this release because documentation of guardian permission for unrestricted public data publication was not available. The dataset supports replication studies, comparisons of ambiguity-attitude models, and analyses of behavioral stability under repeated participation. |
| Date: | 2026–08 |
| URL: | https://d.repec.org/n?u=RePEc:arx:papers:2608.26358 |