nep-dcm New Economics Papers
on Discrete Choice Models
Issue of 2020‒02‒10
eight papers chosen by
Edoardo Marcucci
Università degli studi Roma Tre

  1. Fundamental Limits of Testing the Independence of Irrelevant Alternatives in Discrete Choice By Arjun Seshadri; Johan Ugander
  2. Valuing Rural Residents' Attitude Regarding agri-environmental Policy in China: A Best-worst Scaling Analysis By Qinxin Guo; Junyi Shen
  3. Economic evaluation of catch-and-release salmon fishing: impact on anglers’ willingness to pay By Carole Ropars-Collet; Philippe Le Goffe
  4. A Neural-embedded Choice Model: TasteNet-MNL Modeling Taste Heterogeneity with Flexibility and Interpretability By Yafei Han; Christopher Zegras; Francisco Camara Pereira; Moshe Ben-Akiva
  5. Integrated microtransit services with chance-constrained dynamic pricing and demand learning By Tai-Yu Ma; Sylvain Klein
  6. What are Households Willing to Pay for Better Tap Water Quality? A Cross-Country Valuation Study By Olivier Beaumais; Anne Briand; Katrin Millock; Céline Nauges
  7. Reference Dependence in Intertemporal Preference By Zhihua Li; Songfa Zhong
  8. Estimating Marginal Treatment Effects under Unobserved Group Heterogeneity By Tadao Hoshino; Takahide Yanagi

  1. By: Arjun Seshadri; Johan Ugander
    Abstract: The Multinomial Logit (MNL) model and the axiom it satisfies, the Independence of Irrelevant Alternatives (IIA), are together the most widely used tools of discrete choice. The MNL model serves as the workhorse model for a variety of fields, but is also widely criticized, with a large body of experimental literature claiming to document real-world settings where IIA fails to hold. Statistical tests of IIA as a modelling assumption have been the subject of many practical tests focusing on specific deviations from IIA over the past several decades, but the formal size properties of hypothesis testing IIA are still not well understood. In this work we replace some of the ambiguity in this literature with rigorous pessimism, demonstrating that any general test for IIA with low worst-case error would require a number of samples exponential in the number of alternatives of the choice problem. A major benefit of our analysis over previous work is that it lies entirely in the finite-sample domain, a feature crucial to understanding the behavior of tests in the common data-poor settings of discrete choice. Our lower bounds are structure-dependent, and as a potential cause for optimism, we find that if one restricts the test of IIA to violations that can occur in a specific collection of choice sets (e.g., pairs), one obtains structure-dependent lower bounds that are much less pessimistic. Our analysis of this testing problem is unorthodox in being highly combinatorial, counting Eulerian orientations of cycle decompositions of a particular bipartite graph constructed from a data set of choices. By identifying fundamental relationships between the comparison structure of a given testing problem and its sample efficiency, we hope these relationships will help lay the groundwork for a rigorous rethinking of the IIA testing problem as well as other testing problems in discrete choice.
    Date: 2020–01
  2. By: Qinxin Guo (Graduate School of Economics, Kobe University, Japan); Junyi Shen (Research Institute for Economics & Business Administration (RIEB), Kobe University, Japan)
    Abstract: In this study, a stated choice survey was conducted in Anhui Province, China. The best-worst scaling method, an alternative method to the discrete choice experiment, was used to value rural residents' attitude toward agri-environmental policy. Using the multinomial logit and random parameter logit model, the results showed that respondents thought the best policy included protecting underground water quality as the objective, straw recycling as the method, technological support provided by the government, a supervision level of 30% of farmers, and a 6,000 RMB subsidy directly disbursed by the government. Conversely, respondents thought the worst policy included protecting biodiversity as the objective, purchasing pesticides and fertilizers from the prescribed list as the method, no technological support provided by the government, an increased supervision level of 50% of farmers, and a 4,500 RMB subsidy requiring a contract with the government. The results of the latent class logit model suggested the respondents who are older, have fewer children under middle school age, less agree with the rural environment will have a large impact on agriculture production, have more knowledge of agricultural and environmental 2 protection would show more sensitivity to the attributes of agri-environmental policies.
    Keywords: Agri-environmental policy; Best-worst scaling; Latent class model; Random parameter logit model; Multinomial logit model
    Date: 2020–01
  3. By: Carole Ropars-Collet (SMART - Structures et Marché Agricoles, Ressources et Territoires - INRA - Institut National de la Recherche Agronomique - AGROCAMPUS OUEST); Philippe Le Goffe (SMART - Structures et Marché Agricoles, Ressources et Territoires - INRA - Institut National de la Recherche Agronomique - AGROCAMPUS OUEST)
    Abstract: Catch-and-release (C&R) could be an interesting management tool in recreational fisheries as long as mortality remains low and the anglers' well-being does not drop. We used a choice experiment to examine the potential of C&R angling as a monitoring tool for the salmon recreational fishery in Brittany (France). Anglers were asked to choose between hypothetical fishing day trips differing in terms of their combination of relevant attributes and levels. From the analysis of respondents' trade-offs between the fishing trip's attributes, willingness-to-pay were estimated for each level of attribute. Our results show that anglers prefer unrestrictive regulations. All in all, the majority of the anglers nonetheless hold a positive valuation of a C&R fishing day, which could therefore be used to generate economic returns for the river once the TAC is reached. Lastly, the fishing season, and especially the level of river use, impact more on the value of fishing than C&R.
    Abstract: La remise à l'eau des prises peut être une mesure de gestion intéressante dans le cas de la pêche récréative tant que la mortalité demeure faible et que le bien-être des pêcheurs ne diminue pas. Une enquête a été conduite en 2017 auprès des pêcheurs de saumons des trois départements de l'ouest breton, dans le but de leur faire révéler leur consentement à payer pour différents paramètres de gestion de la pêche : saison, total autorisé de capture (TAC), mode de pêche, no-kill, fréquentation. Il était demandé aux pêcheurs de choisir entre des destinations de pêche hypothétiques différant par la combinaison des paramètres de gestion et la distance pour s'y rendre. En moyenne, on observe que le no-kill a un effet dépressif sur la valorisation de la journée de pêche. Cependant, certaines CSP valorisent positivement le no-kill. Au total, il faut retenir que la majorité des pêcheurs conservent néanmoins une valorisation positive de la journée de pêche en no-kill, ce qui permettrait donc de valoriser la rivière après la clôture du TAC. Enfin, la saison de pêche et surtout la fréquentation impactent davantage la valeur de la pêche que le no-kill.
    Keywords: recreational activity,salmon fishing,catch and release,choice experiment,activité récréative,pêche au saumon,no-kill,expérience de choix
    Date: 2020
  4. By: Yafei Han; Christopher Zegras; Francisco Camara Pereira; Moshe Ben-Akiva
    Abstract: Discrete choice models (DCMs) and neural networks (NNs) can complement each other. We propose a neural network embedded choice model - TasteNet-MNL, to improve the flexibility in modeling taste heterogeneity while keeping model interpretability. The hybrid model consists of a TasteNet module: a feed-forward neural network that learns taste parameters as flexible functions of individual characteristics; and a choice module: a multinomial logit model (MNL) with manually specified utility. TasteNet and MNL are fully integrated and jointly estimated. By embedding a neural network into a DCM, we exploit a neural network's function approximation capacity to reduce specification bias. Through special structure and parameter constraints, we incorporate expert knowledge to regularize the neural network and maintain interpretability. On synthetic data, we show that TasteNet-MNL can recover the underlying non-linear utility function, and provide predictions and interpretations as accurate as the true model; while examples of logit or random coefficient logit models with misspecified utility functions result in large parameter bias and low predictability. In the case study of Swissmetro mode choice, TasteNet-MNL outperforms benchmarking MNLs' predictability; and discovers a wider spectrum of taste variations within the population, and higher values of time on average. This study takes an initial step towards developing a framework to combine theory-based and data-driven approaches for discrete choice modeling.
    Date: 2020–02
  5. By: Tai-Yu Ma; Sylvain Klein
    Abstract: The design of integrated mobility-on-demand services requires jointly considering the interactions between traveler choice behavior and operators' operation policies to design a financially sustainable pricing scheme. However, most existing studies focus on the supply side perspective, disregarding the impact of customer choice behavior in the presence of co-existing transport networks. We propose a modeling framework for dynamic integrated mobility-on-demand service operation policy evaluation with two service options: door-to-door rideshare and rideshare with transit transfer. A new constrained dynamic pricing model is proposed to maximize operator profit, taking into account the correlated structure of different modes of transport. User willingness to pay is considered as a stochastic constraint, resulting in a more realistic ticket price setting while maximizing operator profit. Unlike most studies, which assume that travel demand is known, we propose a demand learning process to calibrate customer demand over time based on customers' historical purchase data. We evaluate the proposed methodology through simulations under different scenarios on a test network by considering the interactions of supply and demand in a multimodal market. Different scenarios in terms of customer arrival intensity, vehicle capacity, and the variance of user willingness to pay are tested. Results suggest that the proposed chance-constrained assortment price optimization model allows increasing operator profit while keeping the proposed ticket prices acceptable.
    Date: 2020–01
  6. By: Olivier Beaumais (LISA - Lieux, Identités, eSpaces, Activités - UPP - Université Pascal Paoli - CNRS - Centre National de la Recherche Scientifique); Anne Briand (LASTA - Laboratoire d'Analyse des Sociétés, Transformations et Adaptations - UNIROUEN - Université de Rouen Normandie - NU - Normandie Université); Katrin Millock (PSE - Paris School of Economics); Céline Nauges (LERNA-INRA - TSE - Toulouse School of Economics - UT1 - Université Toulouse 1 Capitole - INRA - Institut National de la Recherche Agronomique - EHESS - École des hautes études en sciences sociales - CNRS - Centre National de la Recherche Scientifique)
    Abstract: We estimate willingness to pay (WTP) for better quality of tap water on a unique cross-section sample from 10 OECD countries. On the pooled sample, households are willing to pay 7.5% of the median annual water bill to improve the tap water quality. The highest relative WTP for better tap water quality was found in the countries with the highest percentage of respondents being unsatisfied with tap water quality because of health concerns. The expected WTP increased with income, education, environmental concern, and health and taste concerns with the tap water.
    Date: 2020–01–07
  7. By: Zhihua Li (University of Birmingham); Songfa Zhong (National University of Singapore)
    Abstract: This paper examines the role of reference dependence in the elicitation of time preference. The dynamic feature of intertemporal choices offers multiple channels through which the reference effect can occur. Naturally, in evaluating future consumption, one’s current consumption serves as a reference point, and this type of reference point is endogenously determined. Yet potentially exogenous factors are present in the choice environment that also influence the decision maker’s revealed preferences. We performed an experiment that allowed us to examine both the endogenous and exogenous reference effects on the revealed tine preferences. Our design also enabled us to separately estimate the discount factor jointly with the utility curvature. The observed behavioral patterns show that the estimated discount factors were biased by reference endogenous and exogenous points. We also propose a mixture model to account for the reference-dependent effects. We demonstrate that after removing the reference effects and also accounting for the preferences toward money via the utility curvature, the elicited discount factor becomes more patient. We further discuss the implications of the reference-dependent effect on recent observations of elicited intertemporal preferences, including underestimation of the discount factor and the issue of subadditivity.
    Keywords: time preference, reference dependence, prospect theory, experiment
    Date: 2020–02
  8. By: Tadao Hoshino; Takahide Yanagi
    Abstract: This paper studies endogenous treatment effect models in which individuals are classified into unobserved groups based on heterogeneous treatment choice rules. Such heterogeneity may arise, for example, when multiple treatment eligibility criteria and different preference patterns exist. Using a finite mixture approach, we propose a marginal treatment effect (MTE) framework in which the treatment choice and outcome equations can be heterogeneous across groups. Under the availability of valid instrumental variables specific to each group, we show that the MTE for each group can be separately identified using the local instrumental variable method. Based on our identification result, we propose a two-step semiparametric procedure for estimating the group-wise MTE parameters. We first estimate the finite-mixture treatment choice model by a maximum likelihood method and then estimate the MTEs using a series approximation method. We prove that the proposed MTE estimator is consistent and asymptotically normally distributed. We illustrate the usefulness of the proposed method with an application to economic returns to college education.
    Date: 2020–01

This nep-dcm issue is ©2020 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 For comments please write to the director of NEP, Marco Novarese at <>. 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.