nep-dcm New Economics Papers
on Discrete Choice Models
Issue of 2023‒03‒27
seven papers chosen by
Edoardo Marcucci
Università degli studi Roma Tre

  1. Attitudes and Latent Class Choice Models using Machine learning By Lorena Torres Lahoz; Francisco Camara Pereira; Georges Sfeir; Ioanna Arkoudi; Mayara Moraes Monteiro; Carlos Lima Azevedo
  2. Behavioral acceptance of automated vehicles: The roles of perceived safety concern and current travel behavior By Fatemeh Nazari; Mohamadhossein Noruzoliaee; Abolfazl; Mohammadian
  3. Labour market expectations and occupational choice: evidence from teaching By Fullard, Joshua
  4. Greening Vehicle Fleets: A structural analysis of scrappage programs during the financial crisis By KITANO Taiju
  5. BACE: A gretl Package for Model Averaging in Limited Dependent Variable Models By Marcin Blazejowski; Jacek Kwiatkowski
  6. Preference estimation from point allocation experiments By Marion Collewet; Paul Koster
  7. Does Machine Learning Amplify Pricing Errors in the Housing Market? -- The Economics of Machine Learning Feedback Loops By Nikhil Malik; Emaad Manzoor

  1. By: Lorena Torres Lahoz (DTU Management, Technical University of Denmark); Francisco Camara Pereira (DTU Management, Technical University of Denmark); Georges Sfeir (DTU Management, Technical University of Denmark); Ioanna Arkoudi (DTU Management, Technical University of Denmark); Mayara Moraes Monteiro (DTU Management, Technical University of Denmark); Carlos Lima Azevedo (DTU Management, Technical University of Denmark)
    Abstract: Latent Class Choice Models (LCCM) are extensions of discrete choice models (DCMs) that capture unobserved heterogeneity in the choice process by segmenting the population based on the assumption of preference similarities. We present a method of efficiently incorporating attitudinal indicators in the specification of LCCM, by introducing Artificial Neural Networks (ANN) to formulate latent variables constructs. This formulation overcomes structural equations in its capability of exploring the relationship between the attitudinal indicators and the decision choice, given the Machine Learning (ML) flexibility and power in capturing unobserved and complex behavioural features, such as attitudes and beliefs. All of this while still maintaining the consistency of the theoretical assumptions presented in the Generalized Random Utility model and the interpretability of the estimated parameters. We test our proposed framework for estimating a Car-Sharing (CS) service subscription choice with stated preference data from Copenhagen, Denmark. The results show that our proposed approach provides a complete and realistic segmentation, which helps design better policies.
    Date: 2023–02
  2. By: Fatemeh Nazari (Kouros); Mohamadhossein Noruzoliaee (Kouros); Abolfazl (Kouros); Mohammadian
    Abstract: With the prospect of next-generation automated mobility ecosystem, the realization of the contended traffic efficiency and safety benefits are contingent upon the demand landscape for automated vehicles (AVs). Focusing on the public acceptance behavior of AVs, this empirical study addresses two gaps in the plethora of travel behavior research on identifying the potential determinants thereof. First, a clear behavioral understanding is lacking as to the perceived concern about AV safety and the consequent effect on AV acceptance behavior. Second, how people appraise the benefits of enhanced automated mobility to meet their current (pre-AV era) travel behavior and needs, along with the resulting impacts on AV acceptance and perceived safety concern, remain equivocal. To fill these gaps, a recursive trivariate econometric model with ordinal-continuous outcomes is employed, which jointly estimates AV acceptance (ordinal), perceived AV safety concern (ordinal), and current annual vehicle-miles traveled (VMT) approximating the current travel behavior (continuous). Importantly, the co-estimation of the three endogenous outcomes allows to capture the true interdependencies among them, net of any correlated unobserved factors that can have common impacts on these outcomes. Besides the classical socio-economic characteristics, the outcome variables are further explained by the latent preferences for vehicle attributes (including vehicle cost, reliability, performance, and refueling) and for existing shared mobility systems. The model estimation results on a stated preference survey in the State of California provide insights into proactive policies that can popularize AVs through gearing towards the most affected population groups, particularly vehicle cost-conscious, safety-concerned, and lower-VMT (such as travel-restrictive) individuals.
    Date: 2023–02
  3. By: Fullard, Joshua
    Abstract: Using new data on teachers’ intentions to leave the profession, subjective expectations about labour market outcomes and a modified discrete-choice experiment we find that i) teachers are systematically misinformed about population earnings, and misinformation is correlated with attrition intentions; ii) non-pecuniary factors are the most cost-effective method of reducing teacher attrition; and iii) attrition intentions are more affected by reductions in workplace amenities than symmetric improvements, suggesting preventing cuts is more important that rolling out more generous benefits. Linking our survey data to teachers’ administrative records we provide the first evidence that teachers attrition intentions are strong predictors of actual behaviour.
    Date: 2023–03–03
  4. By: KITANO Taiju
    Abstract: Vehicle scrappage programs (SPs) have been a common policy tool to replace aged and/or fuel-inefficient vehicles with fuel-efficient ones, recently adopted to make national vehicle fleets greener. This study evaluates the impacts of the SPs by examining the Japanese private passenger vehicle market in which the government allocated the second-largest program expenditure during the financial crisis. The evaluation is conducted based on the structural model of oligopolistic competition in the presence of the SP, which is estimated using market-level sales, price, and attribute data for each car model from FY2006 to FY2009. To conduct the structural analysis, this study develops a simple method to estimate the demand side in the presence of the SP, which incorporates data on aggregate program outcomes such as the program expenditure in the estimation of the discrete choice models. Given the estimates of the structural model, I simulate counterfactual outcomes under alternative SP designs and discuss program designs that could cost-effectively improve the environmental quality of vehicle fleets, considering the welfare and fiscal stimulus impacts.
    Date: 2023–03
  5. By: Marcin Blazejowski (WSB University in Toruń); Jacek Kwiatkowski (Nicolaus Copernicus University in Toru´n)
    Abstract: This paper presents a software package called BACE (BayesianAveraging of Classical Estimates) which offers model-building strategy for various limited dependent variable models, including logit and probit models, ordered logit and probit models, multinomial logistic regression, Poisson regression, Tobit model, and interval regression. BACE strategy is a model selection method that incorporates both classical estimation and Bayesian techniques. It solves the problem of computation speed and model uncertainty that arise when dealing with a large number of competing advanced statistical models. Our BACE package is both fast and capable of delivering consistent results. The package also provides implementation of the latest proposals of BIC variants, and the latest measures of jointness. We use gretl, a popular, free, and open-source software for econometric analysis that features an easy-to-use graphical user interface.
    Keywords: BMA, model selection, BIC, gretl, Hansl
    Date: 2023–03
  6. By: Marion Collewet (Universiteit Leiden); Paul Koster (Vrije Universiteit Amsterdam)
    Abstract: Point allocation experiments are widely used in the social sciences. In these experiments, survey respondents distribute a fixed total number of points across a fixed number of alternatives. This paper reviews the different perspectives in the literature about what respondents do when they distribute points across options. We find three main alternative interpretations in the literature, each having different implications for empirical work. We connect these interpretations to models of utility maximization that account for point and budget constraints and investigate the role of budget constraints in more detail. We show how these constraints impact the regression specifications for point allocation experiments that are commonly used in the literature. We also show how a formulation of a taste for variety as entropy that had been previously used to analyse market shares can fruitfully be applied to choice behaviour in point allocation experiments.
    Keywords: constant-sum paired comparison, probabilistic choice, entropy, constrained optimization
    Date: 2023–03–09
  7. By: Nikhil Malik; Emaad Manzoor
    Abstract: Machine learning algorithms are increasingly employed to price or value homes for sale, properties for rent, rides for hire, and various other goods and services. Machine learning-based prices are typically generated by complex algorithms trained on historical sales data. However, displaying these prices to consumers anchors the realized sales prices, which will in turn become training samples for future iterations of the algorithms. The economic implications of this machine learning "feedback loop" - an indirect human-algorithm interaction - remain relatively unexplored. In this work, we develop an analytical model of machine learning feedback loops in the context of the housing market. We show that feedback loops lead machine learning algorithms to become overconfident in their own accuracy (by underestimating its error), and leads home sellers to over-rely on possibly erroneous algorithmic prices. As a consequence at the feedback loop equilibrium, sale prices can become entirely erratic (relative to true consumer preferences in absence of ML price interference). We then identify conditions (choice of ML models, seller characteristics and market characteristics) where the economic payoffs for home sellers at the feedback loop equilibrium is worse off than no machine learning. We also empirically validate primitive building blocks of our analytical model using housing market data from Zillow. We conclude by prescribing algorithmic corrective strategies to mitigate the effects of machine learning feedback loops, discuss the incentives for platforms to adopt these strategies, and discuss the role of policymakers in regulating the same.
    Date: 2023–02

This nep-dcm issue is ©2023 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.