nep-dcm New Economics Papers
on Discrete Choice Models
Issue of 2021‒04‒12
fifteen papers chosen by
Edoardo Marcucci
Università degli studi Roma Tre

  1. Comparing hundreds of machine learning classifiers and discrete choice models in predicting travel behavior: an empirical benchmark By Shenhao Wang; Baichuan Mo; Stephane Hess; Jinhua Zhao
  2. Duality in dynamic discrete-choice models By Khai Xiang Chiong; Alfred Galichon; Matt Shum
  3. The propensity to adaptation under the new era of climate changes By Ary José A. de Souza-Jr.; Flávio Terto
  4. Econometric model of children participation in family dairy farming in the center of dairy farming, West Java Province, Indonesia By Achmad Firman; Ratna Ayu Saptati
  5. Hypothetical bias in stated choice experiments: Part II. Macro-scale analysis of literature and effectiveness of bias mitigation methods By Milad Haghani; Michiel C. J. Bliemer; John M. Rose; Harmen Oppewal; Emily Lancsar
  6. The Triple-Store Experiment: A First Simultaneous Test of Classical and Quantum Probabilities in Choice over Menus By Ismaël Rafaï; Sébastien Duchêne; Eric Guerci; Irina Basieva; Andrei Khrennikov
  7. Identification and Estimation in Many-to-one Two-sided Matching without Transfers By YingHua He; Shruti Sinha; Xiaoting Sun
  8. Labor Informality and Credit Market Accessibility By Alina Malkova; Klara Sabirianova Peter; Jan Svejnar
  9. Exploring inner-city residents’ and foreigners’ commitment to improving air pollution: Evidence from a field survey in Hanoi, Vietnam By Khuc, Quy Van; Vuong, Quan-Hoang
  10. Hypothetical bias in stated choice experiments: Part I. Integrative synthesis of empirical evidence and conceptualisation of external validity By Milad Haghani; Michiel C. J. Bliemer; John M. Rose; Harmen Oppewal; Emily Lancsar
  11. Dynamic Performance Management: An Approach for Managing the Common Goods By A. Sardi; E. Sorano
  12. Identification of Peer Effects using Panel Data By Marisa Miraldo; Carol Propper; Christiern Rose
  13. Discretizing Unobserved Heterogeneity By St\'ephane Bonhomme Thibaut Lamadon Elena Manresa
  14. Adaptive Random Bandwidth for Inference in CAViaR Models By Alain Hecq; Li Sun
  15. Testing Identifying Assumptions in Bivariate Probit Models By Acerenza, Santiago; Bartalotti, Otávio; Kedagni, Desire

  1. By: Shenhao Wang; Baichuan Mo; Stephane Hess; Jinhua Zhao
    Abstract: Researchers have compared machine learning (ML) classifiers and discrete choice models (DCMs) in predicting travel behavior, but the generalizability of the findings is limited by the specifics of data, contexts, and authors' expertise. This study seeks to provide a generalizable empirical benchmark by comparing hundreds of ML and DCM classifiers in a highly structured manner. The experiments evaluate both prediction accuracy and computational cost by spanning four hyper-dimensions, including 105 ML and DCM classifiers from 12 model families, 3 datasets, 3 sample sizes, and 3 outputs. This experimental design leads to an immense number of 6,970 experiments, which are corroborated with a meta dataset of 136 experiment points from 35 previous studies. This study is hitherto the most comprehensive and almost exhaustive comparison of the classifiers for travel behavioral prediction. We found that the ensemble methods and deep neural networks achieve the highest predictive performance, but at a relatively high computational cost. Random forests are the most computationally efficient, balancing between prediction and computation. While discrete choice models offer accuracy with only 3-4 percentage points lower than the top ML classifiers, they have much longer computational time and become computationally impossible with large sample size, high input dimensions, or simulation-based estimation. The relative ranking of the ML and DCM classifiers is highly stable, while the absolute values of the prediction accuracy and computational time have large variations. Overall, this paper suggests using deep neural networks, model ensembles, and random forests as baseline models for future travel behavior prediction. For choice modeling, the DCM community should switch more attention from fitting models to improving computational efficiency, so that the DCMs can be widely adopted in the big data context.
    Date: 2021–02
  2. By: Khai Xiang Chiong; Alfred Galichon; Matt Shum
    Abstract: Using results from convex analysis, we investigate a novel approach to identification and estimation of discrete choice models which we call the Mass Transport Approach (MTA). We show that the conditional choice probabilities and the choice-specific payoffs in these models are related in the sense of conjugate duality, and that the identification problem is a mass transport problem. Based on this, we propose a new two-step estimator for these models; interestingly, the first step of our estimator involves solving a linear program which is identical to the classic assignment (two-sided matching) game of Shapley and Shubik (1971). The application of convex-analytic tools to dynamic discrete choice models, and the connection with two-sided matching models, is new in the literature.
    Date: 2021–02
  3. By: Ary José A. de Souza-Jr.; Flávio Terto
    Abstract: Decision utility or experienced utility: which one of them helps us better understand how it will occur the process to the adaptation of behavior’s consumer regarding the impact of climate change? This paper argues how each one of the aforementioned concepts most may affect the consumer’s routine as “decision-makers”, within the context that disturbances and scarcity must narrow their available options. For this, we use the individual’s choice reported in the 8th wave of European Social Survey, based on the choice-oriented perspective, which establishes a link between well-being and the propensity to adapt within this scenario. For this, an ordered logit model is applied upon environmental and socio-economic variables. In the end, our findings are consistent with a strong presence of rationality in the decision process towards adaptation.
    Date: 2021–03
  4. By: Achmad Firman; Ratna Ayu Saptati
    Abstract: The involvement of children in the family dairy farming is pivotal point to reduce the cost of production input, especially in smallholder dairy farming. The purposes of the study are to analysis the factors that influence children's participation in working in the family dairy farm. The study was held December 2020 in the development center of dairy farming in Pangalengan subdistrict, West Java Province, Indonesia. The econometric method used in the study was the logit regression model. The results of the study determine that the there were number of respondents who participates in family farms was 52.59% of total respondents, and the rest was no participation in the family farms. There are 3 variables in the model that are very influential on children's participation in the family dairy farming, such as X1 (number of dairy farm land ownership), X2 (number of family members), and X6 (the amount of work spent on the family's dairy farm). Key words: Participation, children, family, dairy farming, logit model
    Date: 2021–02
  5. By: Milad Haghani; Michiel C. J. Bliemer; John M. Rose; Harmen Oppewal; Emily Lancsar
    Abstract: This paper reviews methods of hypothetical bias (HB) mitigation in choice experiments (CEs). It presents a bibliometric analysis and summary of empirical evidence of their effectiveness. The paper follows the review of empirical evidence on the existence of HB presented in Part I of this study. While the number of CE studies has rapidly increased since 2010, the critical issue of HB has been studied in only a small fraction of CE studies. The present review includes both ex-ante and ex-post bias mitigation methods. Ex-ante bias mitigation methods include cheap talk, real talk, consequentiality scripts, solemn oath scripts, opt-out reminders, budget reminders, honesty priming, induced truth telling, indirect questioning, time to think and pivot designs. Ex-post methods include follow-up certainty calibration scales, respondent perceived consequentiality scales, and revealed-preference-assisted estimation. It is observed that the use of mitigation methods markedly varies across different sectors of applied economics. The existing empirical evidence points to their overall effectives in reducing HB, although there is some variation. The paper further discusses how each mitigation method can counter a certain subset of HB sources. Considering the prevalence of HB in CEs and the effectiveness of bias mitigation methods, it is recommended that implementation of at least one bias mitigation method (or a suitable combination where possible) becomes standard practice in conducting CEs. Mitigation method(s) suited to the particular application should be implemented to ensure that inferences and subsequent policy decisions are as much as possible free of HB.
    Date: 2021–02
  6. By: Ismaël Rafaï (CEE-M, Univ. Montpellier, CNRS, INRAE, Institut Agro; Université Côte d'Azur, France; GREDEG CNRS); Sébastien Duchêne (CEE-M, Univ. Montpellier, CNRS, INRAE, Institut Agro); Eric Guerci (Université Côte d'Azur, France; GREDEG CNRS); Irina Basieva (Department of Psychology, City University, London, United Kingdom); Andrei Khrennikov (International Center for Mathematical Modeling, in Physics and Cognitive Science Linnaeus University, Växjö, Sweden)
    Abstract: Recently quantum probability theory started to be actively used in studies of human decision-making, in particular for the resolution of paradoxes (such as the Allais, Ellsberg, and Machina paradoxes). Previous studies were based on a cognitive metaphor of the quantum double-slit experiment - the basic quantum interference experiment. In this paper, we report on an economics experiment based on a three-slit experiment design, where the slits are menus of alternatives from which one can choose. The test of nonclassicality is based on the Sorkin equality (which was only recently tested in quantum physics). Each alternative is a voucher to buy products in one or more stores. The alternatives are obtained from all disjunctions including one, two or three stores. The participants have to reveal the amount for which they are willing to sell the chosen voucher. Interference terms are computed by comparing the willingness to sell a voucher built as a disjunction of stores and the willingness to sell the vouchers corresponding to the singleton stores. These willingness to sell amounts are used to estimate probabilities and to test both the law of total probabilities and the Born Rule. Results reject neither classical nor quantum probability. We discuss this initial experiment and our results and provide guidelines for future studies.
    Keywords: Input-output, Covid-19, Lockdown, Italy
    JEL: C63 C67 D57 E17 I18 R15
    Date: 2021–04
  7. By: YingHua He; Shruti Sinha; Xiaoting Sun
    Abstract: In a setting of many-to-one two-sided matching with non-transferable utilities, e.g., college admissions, we study conditions under which preferences of both sides are identified with data on one single market. The main challenge is that every agent's actual choice set is unobservable to the researcher. Assuming that the observed matching is stable, we show nonparametric and semiparametric identification of preferences of both sides under appropriate exclusion restrictions. Our identification arguments are constructive and thus directly provide a semiparametric estimator. In Monte Carlo simulations, the estimator can perform well but suffers from the curse of dimensionality. We thus adopt a parametric model and estimate it by a Bayesian approach with a Gibbs sampler, which works well in simulations. Finally, we apply our method to school admissions in Chile and conduct a counterfactual analysis of an affirmative action policy.
    Date: 2021–04
  8. By: Alina Malkova; Klara Sabirianova Peter; Jan Svejnar
    Abstract: The paper investigates the effects of the credit market development on the labor mobility between the informal and formal labor sectors. In the case of Russia, due to the absence of a credit score system, a formal lender may set a credit limit based on the verified amount of income. To get a loan, an informal worker must first formalize his or her income (switch to a formal job), and then apply for a loan. To show this mechanism, the RLMS data was utilized, and the empirical method is the dynamic multinomial logit model of employment. The empirical results show that a relaxation of credit constraints increases the probability of transition from an informal to a formal job, and improved CMA (by one standard deviation) increases the chances of informal sector workers to formalize by 5.4 ppt. These results are robust in different specifications of the model. Policy simulations show strong support for a reduction in informal employment in response to better CMA in credit-constrained communities.
    Date: 2021–02
  9. By: Khuc, Quy Van; Vuong, Quan-Hoang
    Abstract: Air pollution, willingness-to-pay, contingent valuation method, inner-city citizens, foreigners, Hanoi, Vietnam
    Date: 2021–04–07
  10. By: Milad Haghani; Michiel C. J. Bliemer; John M. Rose; Harmen Oppewal; Emily Lancsar
    Abstract: The notion of hypothetical bias (HB) constitutes, arguably, the most fundamental issue in relation to the use of hypothetical survey methods. Whether or to what extent choices of survey participants and subsequent inferred estimates translate to real-world settings continues to be debated. While HB has been extensively studied in the broader context of contingent valuation, it is much less understood in relation to choice experiments (CE). This paper reviews the empirical evidence for HB in CE in various fields of applied economics and presents an integrative framework for how HB relates to external validity. Results suggest mixed evidence on the prevalence, extent and direction of HB as well as considerable context and measurement dependency. While HB is found to be an undeniable issue when conducting CEs, the empirical evidence on HB does not render CEs unable to represent real-world preferences. While health-related choice experiments often find negligible degrees of HB, experiments in consumer behaviour and transport domains suggest that significant degrees of HB are ubiquitous. Assessments of bias in environmental valuation studies provide mixed evidence. Also, across these disciplines many studies display HB in their total willingness to pay estimates and opt-in rates but not in their hypothetical marginal rates of substitution (subject to scale correction). Further, recent findings in psychology and brain imaging studies suggest neurocognitive mechanisms underlying HB that may explain some of the discrepancies and unexpected findings in the mainstream CE literature. The review also observes how the variety of operational definitions of HB prohibits consistent measurement of HB in CE. The paper further identifies major sources of HB and possible moderating factors. Finally, it explains how HB represents one component of the wider concept of external validity.
    Date: 2021–02
  11. By: A. Sardi; E. Sorano
    Abstract: Public organizations need innovative approaches for managing common goods and to explain the dynamics linking the (re)generation of common goods and organizational performance. Although system dynamics is recognised as a useful approach for managing common goods, public organizations rarely adopt the system dynamics for this goal. The paper aims to review the literature on the system dynamics and its recent application, known as dynamic performance management, to highlight the state of the art and future opportunities on the management of common goods. The authors analyzed 144 documents using a systematic literature review. The results obtained outline a fair number of documents, countries and journals involving the study of system dynamics, but do not cover sufficient research on the linking between the (re)generation of common goods and organizational performance. This paper outlines academic and practical contributions. Firstly, it contributes to the theory of common goods. It provides insight for linking the management of common goods and organizational performance through the use of dynamic performance management approach. Furthermore, it shows scholars the main research opportunities. Secondly, it indicates to practitioners the documents providing useful ideas on the adoption of system dynamics for managing common goods.
    Date: 2021–02
  12. By: Marisa Miraldo (Imperial College Business School,); Carol Propper (Imperial College Business School); Christiern Rose (School of Economics, University of Queensland, Brisbane, Australia)
    Abstract: This paper provides new identification results for panel data models with contextual and endogenous peer effects. Contextual effects operate through individuals’ time-invariant unobserved heterogeneity. Identification hinges on a conditional mean restriction requiring exogenous mobility of individuals between groups over time. Some networks governing peer interactions preclude identification. For these cases we propose additional conditional variance restrictions. We conduct a Monte-Carlo experiment to evaluate the performance of our method and apply it to surgeon-hospital-year data to study take-up of minimally invasive surgery. A standard deviation increase in the average time-invariant unobserved heterogeneity of other surgeons in the same hospital leads to a 0.12 standard deviation increase in take-up. The effect is equally due to endogenous and contextual effects.
    Keywords: Peer effects, panel data, networks, identification, innovation, healthcare
    Date: 2020–12–03
  13. By: St\'ephane Bonhomme Thibaut Lamadon Elena Manresa
    Abstract: We study discrete panel data methods where unobserved heterogeneity is revealed in a first step, in environments where population heterogeneity is not discrete. We focus on two-step grouped fixed-effects (GFE) estimators, where individuals are first classified into groups using kmeans clustering, and the model is then estimated allowing for group-specific heterogeneity. Our framework relies on two key properties: heterogeneity is a function - possibly nonlinear and time-varying - of a low-dimensional continuous latent type, and informative moments are available for classification. We illustrate the method in a model of wages and labor market participation, and in a probit model with time-varying heterogeneity. We derive asymptotic expansions of two-step GFE estimators as the number of groups grows with the two dimensions of the panel. We propose a data-driven rule for the number of groups, and discuss bias reduction and inference.
    Date: 2021–02
  14. By: Alain Hecq; Li Sun
    Abstract: This paper investigates the size performance of Wald tests for CAViaR models (Engle and Manganelli, 2004). We find that the usual estimation strategy on test statistics yields inaccuracies. Indeed, we show that existing density estimation methods cannot adapt to the time-variation in the conditional probability densities of CAViaR models. Consequently, we develop a method called adaptive random bandwidth which can approximate time-varying conditional probability densities robustly for inference testing on CAViaR models based on the asymptotic normality of the model parameter estimator. This proposed method also avoids the problem of choosing an optimal bandwidth in estimating probability densities, and can be extended to multivariate quantile regressions straightforward.
    Date: 2021–02
  15. By: Acerenza, Santiago; Bartalotti, Otávio; Kedagni, Desire
    Abstract: This paper focuses on the bivariate probit model's identifyingassumptions: joint normality of errors, instrument exogeneity, and relevance conditions. First, we develop novel sharp testable equalities that can detect all possible observable violations of the assumptions. Second, we propose an easy-to-implement testing procedure for the model's validity based on feasible testable implications using existing inference methods for intersection bounds. The test achieves correct empirical size for moderately sized samples and performs well in detecting violations of the conditions in Monte Carlo simulations. Finally, we provide researchers with a road map on what to do when the bivariate probit model is rejected, including novel bounds for the average treatment effect that relax the normality assumption. Empirical examples illustrate the methodology's implementation.
    Date: 2021–03–29

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.