nep-dcm New Economics Papers
on Discrete Choice Models
Issue of 2021‒02‒08
nine papers chosen by
Edoardo Marcucci
Università degli studi Roma Tre

  1. Uninsured by choice? A choice experiment on long term care insurance By Akaichi, Faical; Costa-Font, Joan; Frank, Richard
  2. Willingness to Pay and Attitudinal Preferences of Indian Consumers for Electric Vehicles By Prateek Bansal; Rajeev Ranjan Kumar; Alok Raj; Subodh Dubey; Daniel J. Graham
  3. New York, Abu Dhabi, London or Stay at Home? Using a Cross-Nested Logit Model to Identify Complex Substitution Patterns in Migration By Michel Beine; Michel Bierlaire; Frédéric Docquier
  4. Informational Content of Factor Structures in Simultaneous Binary Response Models By Khan, Shakeeb; Maurel, Arnaud; Zhang, Yichong
  5. Rationalizing choice functions with a weak preference By Yuta Inoue
  6. Happiness and Air Pollution By Arik Levinson
  7. Consumer acceptance and willingness-to-pay for insect-based foods: The role of proximity of insects in the food chain By Giotis, Thomas; Drichoutis, Andreas C.
  8. Additive and decomposable conjoint measurement with ordered categories By Denis Bouyssou; Thierry Marchant
  9. Choice modelling in the age of machine learning By S. Van Cranenburgh; S. Wang; A. Vij; F. Pereira; J. Walker

  1. By: Akaichi, Faical; Costa-Font, Joan; Frank, Richard
    Abstract: We examine evidence from two unique discrete choice experiments (DCE) on long term care insurance and values several of its relevant attributes. More specifically, the experiments examine choices made by a large sample made of 15,298 individuals in the United States with and without insurance. We study the valuation of a number of insurance attributes, namely the daily insurance benefit, insurance coverage, the compulsory and voluntary nature of the insurance policy design, alongside the costs (insurance premium) and health requirements. This paper elicits both respondents’ preferences and willingness to pay (WTP) for these care insurance's attributes using a random parameter logit model, and assesses the heterogeneity of choice responses using demographic, socioeconomic and attitudinal motivations to segment response to insurance choices. We find that an increase in the insurance premium by an additional $100 would reduce insurance uptake by 1pp. Insurance policy uptake is higher when it provides benefits for the lifetime (the monthly marginal WTP being $178.64), and voluntary (the monthly marginal WTP increases by an extra $74.71) as opposed to universal, and when it forgoes health checks (the monthly marginal WTP increases by an extra 28US$).
    Keywords: long term care insurance; constrained choices; self-insurance; behavioural constraints; insurance design
    JEL: I18 I31
    Date: 2020–05–01
  2. By: Prateek Bansal; Rajeev Ranjan Kumar; Alok Raj; Subodh Dubey; Daniel J. Graham
    Abstract: Consumer preference elicitation is critical to devise effective policies for the diffusion of electric vehicles in India. For this purpose, we use stated preferences of over 1000 Indian consumers and present the first estimates of the additional purchase price that Indian consumers are willing to pay to improve the electric vehicle attributes (e.g., driving range). We adopt a hybrid choice modelling framework to simultaneously model the willingness to pay and the effect of consumers' latent attitudes (e.g., hedonic and symbolic values) on their likelihood to adopt electric vehicles. We also account for reference dependence by considering a conventional vehicle as the reference alternative and illustrate that the resulting willingness to pay estimates are more realistic. Based on our results, Indian consumers are willing to pay additional USD 10-34 in purchase price to reduce the fast charging time by 1 minute, USD 7-40 to add a kilometre to the driving range of electric vehicle at 200 kilometres, and USD 104-692 to save USD 1 per 100 kilometres in operating cost. These estimates and the effect of attitudes on the likelihood to adopt electric vehicles provide insights about electric vehicle design, marketing strategies, and pro-electric-vehicle policies (e.g., specialised lanes and reserved parking for electric vehicles) to expedite the adoption of electric vehicles in India.
    Date: 2021–01
  3. By: Michel Beine; Michel Bierlaire; Frédéric Docquier
    Abstract: The question of how people revise their decisions about whether to emigrate, and where to, when facing changes in the global environment is of critical importance in migration literature. We propose a cross-nested logit (CNL) approach to generalize the way deviations from the IIA (independence from irrelevant alternatives)) hypothesis can be tested and exploited in migration studies. Compared with the widely used logit model, the structure of a CNL model allows for more sophisticated substitution patterns between destinations. To illustrate the relevance of our approach, we provide a case study using migration aspiration data from India. We demonstrate that the CNL approach outperforms standard competing approaches in terms of quality of fit, has stronger predictive power, implies stronger heterogeneity in responses to shocks, and highlights complex and intuitive substitution patterns between all possible alternatives. In particular, we shed light on the low degree of substitutability between the home and foreign alternatives as well as on the subgroups of countries that are considered by potential Indian movers as highly or poorly substitutable.
    Keywords: International migration; Discrete choice modelling; Independence from irrelevant alternatives; Cross-nested logit; Migration aspirations
    JEL: J61
    Date: 2020–01
  4. By: Khan, Shakeeb (Boston College); Maurel, Arnaud (Duke University); Zhang, Yichong (Singapore Management University)
    Abstract: We study the informational content of factor structures in discrete triangular systems. Factor structures have been employed in a variety of settings in cross sectional and panel data models, and in this paper we formally quantify their identifying power in a bivariate system often employed in the treatment effects literature. Our main findings are that imposing a factor structure yields point identification of parameters of interest, such as the coefficient associated with the endogenous regressor in the outcome equation, under weaker assumptions than usually required in these models. In particular, we show that a "non-standard" exclusion restriction that requires an explanatory variable in the outcome equation to be excluded from the treatment equation is no longer necessary for identification, even in cases where all of the regressors from the outcome equation are discrete. We also establish identification of the coefficient of the endogenous regressor in models with more general factor structures, in situations where one has access to at least two continuous measurements of the common factor.
    Keywords: factor structures, discrete choice, causal effects
    JEL: C14 C31 C35
    Date: 2020–12
  5. By: Yuta Inoue (Graduate School of Economics, Waseda University)
    Abstract: This paper develops revealed preference analysis of an individual choice model where an agent is a weak preference maximizer, under the assumption that a choice function, rather than a choice correspondence, is observed. In particular, we provide a revealed preference test for such model, and then provide conditions under which we can surely say whether some alternative is indifferent / weakly preferred / strictly preferred to another, solely from the information of the choice function. Furthermore, interpreting a choice correspondence as sets of potential candidates of alternatives that could be chosen from each feasible set, we analyze which alternatives must be, or cannot be a member of the choice correspondence: sharp lower and upper bounds of this underlying choice correspondence are given. As an assumption on observability of data, we assume that the choice function is defined on a non-exhaustive domain, so our results are applicable to data analysis even when only a limited data set is available.
    Keywords: Revealed preference; Choice function; Choice correspondence; Weak preference; Bounded rationality
    JEL: D1 D6
    Date: 2020–05
  6. By: Arik Levinson (Department of Economics, Georgetown University)
    Abstract: I pose three questions: Does pollution make people unhappy? How much? And is the effect proportional to pollution’s estimated effects on mortality and productivity? Answers to those three questions must overcome three obstacles: unobserved characteristics of locales correlated with both pollution and happiness; selection by pollution-averse individuals to less polluted areas; and habituation by residents to local circumstances. Since 2010, when the initial few studies relating happiness to pollution were last surveyed, thirty more have been published. I discuss how the new studies tackle each of those three problems and I devise a method of comparing their findings despite their different measures of both happiness and pollution. I combine the happiness and income coefficients from each study into a willingness-to-pay measure, for a one-day, one-standard-deviation pollution reduction. Finally, I document a surprising concordance between those calculated willingness-to-pay measures and new research assessing the effects of pollution on mortality and productivity
    Keywords: Stated well-being, willingness-to-pay, habituation, short-termism
    JEL: Q51 Q53 H41
    Date: 2020–09–30
  7. By: Giotis, Thomas; Drichoutis, Andreas C.
    Abstract: Over the last few years, the interest on alternative protein sources, such as edible insects, has been growing rapidly. However, Western consumers' acceptance of insects as a food source is very low, mainly due to unfamiliarity with insect-based food. We investigate consumers' attitude and behavior and estimate their willingness to pay (WTP) a premium for three products that vary on a between-subjects basis, the proximity of insects in the food chain. The data were collected through an online questionnaire of 451 consumers in Greece and WTP was elicited using the Contingent Valuation (CV) method. Our results show that the majority of Greek consumers are not willing to pay a premium for an insect-based energy bar and cookie; on the contrary, they would require a discount to acquire such products. On the other hand, consumer acceptance is higher for a gilt-head bream that is fed with insect-based feed. Consumers with positive WTP are on average willing to pay a premium of 15.8%, 17% and 31.8% for the energy bar, cookie and gilt-head bream, respectively, while consumers that are not WTP a premium would require discounts of 43.8%, 42.4% and 30.7%, respectively.
    Keywords: Consumer acceptance; willingness-to-pay; contingent valuation; cheap talk; insect-based products; insect-based feed
    JEL: C90 D12 Q13
    Date: 2020–12–18
  8. By: Denis Bouyssou (LAMSADE - Laboratoire d'analyse et modélisation de systèmes pour l'aide à la décision - Université Paris Dauphine-PSL - PSL - Université Paris sciences et lettres - CNRS - Centre National de la Recherche Scientifique); Thierry Marchant (Ghent University - Research Group on Combinatorial Algorithms and Algorithmic Graph Theory - UGENT - Ghent University [Belgium])
    Abstract: Conjoint measurement studies binary relations defined on product sets and investigates the existence and uniqueness of numerical representations of such relations. It has proved to be quite a powerful tool to analyze and compare MCDM techniques designed to build a preference relation between multiattributed alternatives and has been an inspiring guide to many assessment protocols. These MCDM techniques lead to a relative evaluation model of the alternatives through a preference relation. Such models are not always appropriate to build meaningful recommendations. This has recently lead to the development of MCDM techniques aiming at building evaluation models having a more absolute character. In such techniques, the output of the analysis is, most often, a partition of the set of alternatives into several ordered categories defined with respect to outside norms, e.g., separating "Attractive" and "Unattractive" alternatives. In spite of their interest, the theoretical foundations of such MCDM techniques have not been much investigated. The purpose of this paper is to contribute to this analysis. More precisely, we show how to adapt classic conjoint measurement results to make them applicable for the study of such MCDM techniques. We concentrate on additive models. Our results may be seen as an attempt to provide an axiomatic basis to the well-known UTADIS technique that sorts alternatives using an additive value function model.
    Date: 2021–01–05
  9. By: S. Van Cranenburgh; S. Wang; A. Vij; F. Pereira; J. Walker
    Abstract: Since its inception, the choice modelling field has been dominated by theory-driven models. The recent emergence and growing popularity of machine learning models offer an alternative data-driven approach. Machine learning models, techniques and practices could help overcome problems and limitations of the current theory-driven modelling paradigm, e.g. relating to the ad-hocness in search for the optimal model specification, and theory-driven choice model's inability to work with text and image data. However, despite the potential value of machine learning to improve choice modelling practices, the choice modelling field has been somewhat hesitant to embrace machine learning. The aim of this paper is to facilitate (further) integration of machine learning in the choice modelling field. To achieve this objective, we make the case that (further) integration of machine learning in the choice modelling field is beneficial for the choice modelling field, and, we shed light on where the benefits of further integration can be found. Specifically, we take the following approach. First, we clarify the similarities and differences between the two modelling paradigms. Second, we provide a literature overview on the use of machine learning for choice modelling. Third, we reinforce the strengths of the current theory-driven modelling paradigm and compare this with the machine learning modelling paradigm, Fourth, we identify opportunities for embracing machine learning for choice modelling, while recognising the strengths of the current theory-driven paradigm. Finally, we put forward a vision on the future relationship between the theory-driven choice models and machine learning.
    Date: 2021–01

This nep-dcm issue is ©2021 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.