nep-dcm New Economics Papers
on Discrete Choice Models
Issue of 2021‒07‒26
sixteen papers chosen by
Edoardo Marcucci
Università degli studi Roma Tre

  1. Does Omitting Downstream Water Quality Change the Economic Benefits of Nutrient Reduction? Evidence from a Discrete Choice Experiment By Shr, Yau-Huo (Jimmy); Zhang, Wendong
  2. Structural modeling of simultaneous discrete choice By Andrew Chesher; Adam Rosen
  3. The unreasonable effectiveness of optimal transport in economics By Alfred Galichon
  4. Employers’ willingness to invest in the training of temporary workers: a discrete choice experiment By Poulissen, Davey; de Grip, Andries; Fouarge, Didier; Künn, Annemarie
  5. Dynamic Ordered Panel Logit Models By Bo E. Honor\'e; Chris Muris; Martin Weidner
  6. Partial Identi?cation and Inference for Dynamic Models and Counterfactuals By Myrto Kalouptsidi; Yuichi Kitamura; Lucas Lima; Eduardo Souza-Rodrigues
  7. An Exact Method for Assortment Optimization under the Nested Logit Model By Alfandari, Laurent; Hassanzadeh, Alborz; Ljubic, Ivana
  8. Tennessee Consumer Perceptions of Milk: Purchase Considerations, Safety and Price By Eckelkamp, Liz; Zaring, Caitlin; Upendram, Sreedhar; Paskewitz, Emily A.; Sedges, Heather; Johnson, Kristen
  9. Constrained School Choice under the Immediate Acceptance Mechanism: An Experimental QRE Analysis By Jorge Alcalde-Unzu; Flip Klijn; Marc Vorsatz
  10. Inference for the proportional odds cumulative logit model with monotonicity constraints for ordinal predictors and ordinal response By Javier Espinosa; Christian Hennig
  11. Unifying Classical and Bayesian Revealed Preference By Kunal Pattanayak; Vikram Krishnamurthy
  12. Representing choice functions by a total hyper-order By Daniel Lehmann
  13. Subspace Shrinkage in Conjugate Bayesian Vector Autoregressions By Florian Huber; Gary Koop
  14. Identifying the effect of persuasion By Sung Jae Jun; Sokbae (Simon) Lee
  15. A charging infrastructure network for battery electric trucks in Europe By Sauter, Verena; Speth, Daniel; Plötz, Patrick; Signer, Tim
  16. Who Votes for Library Bonds? A Principal Component Exploration By Eric Jacobson

  1. By: Shr, Yau-Huo (Jimmy); Zhang, Wendong
    Abstract: Discrete choice experiments have been extensively used to value environmental quality; however, some important attributes may be often omitted due to design challenges. In the case of agricultural water pollution, overlooking downstream water quality benefits could lead to biased estimates and misinterpretations of local water quality attributes presented in the choice experiments. Using a split-sample design and a statewide survey of Iowa residents, we provide the first systematic evaluation of how households’ willingness-to-pay for water quality change when downstream water quality benefits, hypoxic zone reduction in our case, are omitted. We find that omitting non-local water quality attributes significantly reduces the total economic value of nutrient reduction programs but does not bias the marginal willingness-to-pay for local water quality attributes. We also find suggestive evidence showing that such omission, in line with the theoretical prediction, only changes the preferences of respondents who are aware of the downstream impacts of local water quality improvement plans. In addition, our results show that providing information on the non-local water quality benefits of nutrient reduction makes respondents less informed about the water quality issues more likely to support the water quality improvement plans.
    Date: 2021–01–01
  2. By: Andrew Chesher (Institute for Fiscal Studies and University College London); Adam Rosen (Institute for Fiscal Studies and Duke University)
    Abstract: Models of simultaneous discrete choice may be incomplete, delivering multiple values of outcomes at certain values of the latent variables and covariates, and incoherent, delivering no values. Alternative approaches to accommodating incompleteness and incoherence are considered in a unifying framework afforded by the Generalized Instrumental Variable models introduced in Chesher and Rosen (2017). Sharp identification regions for parameters and functions of interest defined by systems of conditional moment equalities and inequalities are provided. Almost all empirical analysis of simultaneous discrete choice uses models that include parametric specifications of the distribution of unobserved variables. The paper provides characterizations of identified sets and outer regions for structural functions and parameters allowing for any distribution of unobservables independent of exogenous variables. The methods are applied to the models and data of Mazzeo (2002) and Kline and Tamer (2016) in order to study the sensitivity of empirical results to restrictions on equilibrium selection and the distribution of unobservable payoff shifters, respectively. Confidence intervals for individual parameter components are provided using a recently developed inference approach from Belloni, Bugni, and Chernozhukov (2018). The relaxation of equilibrium selection and distributional restrictions in these applications is found to greatly increase the width of resulting confidence intervals, but nonetheless the models continue to sign strategic interaction parameters.
    Date: 2020–02–20
  3. By: Alfred Galichon
    Abstract: Optimal transport has become part of the standard quantitative economics toolbox. It is the framework of choice to describe models of matching with transfers, but beyond that, it allows to: extend quantile regression; identify discrete choice models; provide new algorithms for computing the random coefficient logit model; and generalize the gravity model in trade. This paper offer a brief review of the basics of the theory, its applications to economics, and some extensions.
    Date: 2021–07
  4. By: Poulissen, Davey (RS: GSBE other - not theme-related research, ROA / Health, skills and inequality); de Grip, Andries (ROA / Health, skills and inequality, RS: GSBE Theme Learning and Work, RS: SBE - MACIMIDE); Fouarge, Didier (RS: GSBE Theme Learning and Work, RS: GSBE Theme Data-Driven Decision-Making, ROA / Labour market and training); Künn, Annemarie (RS: GSBE Theme Learning and Work, ROA / Labour market and training)
    Abstract: Various studies have shown that temporary workers participate less in training than those on permanent contracts. Human resources practices are considered to be an important explanation for this difference. We develop a theoretical framework for employers’ provision of training that explicitly incorporates the costs and benefits associated with training investments in employees with different types of employment contracts. Our framework not only predicts employers to be less willing to invest in temporary workers due to the shorter time horizon associated with such an investment, but it also provides insights into how this willingness depends on characteristics of the training that are related to the expected costs and benefits of the training investment. A discrete choice experiment is used to empirically test the predictions from our theoretical framework. In line with our theoretical framework, we find that employers are less likely to invest in the training of temporary workers. This particularly holds when temporary workers do not have the prospect of a permanent contract with their current employer. Furthermore, we show that employers’ likelihood of investing in temporary workers indeed depends on aspects related to the costs and benefits of training, that is, a financial contribution to the training costs made by employees, a repayment agreement that applies when workers leave the organisation prematurely, and the transferability of the skills being trained. Our findings can be used to increase employers’ willingness to invest in temporary workers. However, similar effects are observed when looking at employers’ willingness to invest in permanent workers, suggesting that it will be difficult to decrease the gap in employers’ willingness to invest between temporary and permanent workers.
    JEL: J24 J41 J62
    Date: 2021–05–31
  5. By: Bo E. Honor\'e; Chris Muris; Martin Weidner
    Abstract: We study a dynamic ordered logit model for panel data with fixed effects. We establish the validity of a set of moment conditions that are free of the fixed effects and that can be computed using four or more periods of data. We establish sufficient conditions for these moment conditions to identify the regression coefficients, the autoregressive parameters, and the threshold parameters. The parameters can be estimated using generalized method of moments. We document the performance of this estimator using Monte Carlo simulations and an empirical illustration to self-reported health status using the British Household Panel Survey.
    Date: 2021–07
  6. By: Myrto Kalouptsidi (Institute for Fiscal Studies); Yuichi Kitamura (Institute for Fiscal Studies and Yale University); Lucas Lima (Institute for Fiscal Studies); Eduardo Souza-Rodrigues (Institute for Fiscal Studies)
    Abstract: We provide a general framework for investigating partial identi?cation of structural dynamic discrete choice models and their counterfactuals, along with uniformly valid inference procedures. In doing so, we derive sharp bounds for the model parameters, counterfactual behavior, and low-dimensional outcomes of interest, such as the average welfare e?ects of hypothetical policy interventions. We characterize the properties of the sets analytically and show that when the target outcome of interest is a scalar, its identi?ed set is an interval whose endpoints can be calculated by solving well-behaved constrained optimization problems via standard algorithms. We obtain a uniformly valid inference procedure by an appropriate application of subsampling. To illustrate the performance and computational feasibility of the method, we consider both a Monte Carlo study of ?rm entry/exit, and an empirical model of export decisions applied to plant-level data from Colombian manufacturing industries. In these applications, we demonstrate how the identi?ed sets shrink as we incorporate alternative model restrictions, providing intuition regarding the source and strength of identi?cation.
    Date: 2020–02–10
  7. By: Alfandari, Laurent (ESSEC Research Center, ESSEC Business School); Hassanzadeh, Alborz (ESSEC Research Center, ESSEC Business School); Ljubic, Ivana (ESSEC Research Center, ESSEC Business School)
    Abstract: We study the problem of finding an optimal assortment of products maximizing the expected revenue, in which customer preferences are modeled using a Nested Logit choice model. This problem is known to be polynomially solvable in a specific case and NP-hard otherwise, with only approximation algorithms existing in the literature. We provide an exact general method that embeds a tailored Branch-and-Bound algorithm into a fractional programming framework. Contrary to the existing literature, in which assumptions are imposed on either the structure of nests or the combination and characteristics of products, no assumptions on the input data are imposed. Although our approach is not polynomial in the input size, it can solve the most general problem setting for large-size instances. We show that the fractional programming scheme’s parameterized subproblem, a highly non-linear binary optimization problem, is decomposable by nests, which is the primary advantage of the approach. To solve the subproblem for each nest, we propose a two-stage approach. In the first stage, we fix a large set of variables based on the single-nest subproblem’s newly-derived structural properties. This can significantly reduce the problem size. In the second stage, we design a tailored Branch-and-Bound algorithm with problem-specific upper bounds. Numerical results show that the approach is able to solve assortment instances with five nests and with up to 5,000 products per nest. The most challenging instances for our approach are those with a mix of nests’ dissimilarity parameters, where some of them are smaller than one and others are greater than one.
    Keywords: combinatorial optimization; revenue management; assortment optimization; fractional programming; nested logit
    JEL: C00
    Date: 2020
  8. By: Eckelkamp, Liz; Zaring, Caitlin; Upendram, Sreedhar; Paskewitz, Emily A.; Sedges, Heather; Johnson, Kristen
    Abstract: The Tennessee dairy industry is facing many challenges with aging farmer populations, low milk prices and dairy farms struggling to maintain profitability. Many dairy producers have retired, sold out or lost contracts with milk handlers leading to a steady decline of dairy farms. Tennessee has declined to 179 licensed Grade ‘A’ dairy farms in January 2020 from 276 Grade ‘A’ dairy farms in January 2018 – a decrease of 97 dairies in two years (Strasser, 2021). With the loss of dairy farms, we can expect economic difficulties for businesses that provide goods and services to the dairy industry across Tennessee. Along with declining milk prices, consumer demand for fluid milk has also been decreasing (Figure 1; USDA-ERS, 2020). The trend to consume local goods and services could potentially help Tennessee producers. In 2018, a Tennessee Milk logo was created to promote milk produced and bottled in Tennessee. Theoretically, this milk could be considered premium and demand a higher price. The consumer demand for locally branded fluid milk is unknown. The goal of this publication is to provide results of a consumer survey of perceptions, preferences and purchasing considerations for local, organic and store-brand milk to dairy producers, retailers and policy makers. As part of this study, we present: • Study participants’ willingness to pay for local, organic or store-branded milk • Purchasing trends for milk • Attributes associated with local, organic or store-branded milk • Participants’ various definitions of “local” according to geographic regions and miles traveled
    Keywords: Community/Rural/Urban Development, Demand and Price Analysis, Marketing
    Date: 2021–06–18
  9. By: Jorge Alcalde-Unzu; Flip Klijn; Marc Vorsatz
    Abstract: The theoretical literature on public school choice proposes centralized mechanisms that assign children to schools on the basis of parents' preferences and the priorities children have for different schools. The related experimental literature analyzes in detail how various mechanisms fare in terms of welfare and stability of the resulting matchings, yet often provides only aggregate statistics of the individual behavior that leads to these outcomes (i.e., the degree to which subjects tell the truth in the induced simultaneous move game). In this paper, we show that the quantal response equilibrium adequately describes individual behavior and the resulting matching in a laboratory experiment on the (constrained) immediate acceptance mechanism. Specifically, the comparative statics of the quantal response equilibrium capture the directional changes of subject behavior and the prevalence of the different stable matchings when cardinal payoffs (i.e., relative preference intensities) are modified in the experiment.
    Keywords: laboratory experiment, school choice, immediate acceptance mechanism, Boston mechanism, quantal response equilibrium
    JEL: C78 C91 C92 D78 I20
    Date: 2021–07
  10. By: Javier Espinosa; Christian Hennig
    Abstract: The proportional odds cumulative logit model (POCLM) is a standard regression model for an ordinal response. Ordinality of predictors can be incorporated by monotonicity constraints for the corresponding parameters. It is shown that estimators defined by optimization, such as maximum likelihood estimators, for an unconstrained model and for parameters in the interior set of the parameter space of a constrained model are asymptotically equivalent. This is used in order to derive asymptotic confidence regions and tests for the constrained model, involving simple modifications for finite samples. The finite sample coverage probability of the confidence regions is investigated by simulation. Tests concern the effect of individual variables, monotonicity, and a specified monotonicity direction. The methodology is applied on real data related to the assessment of school performance.
    Date: 2021–07
  11. By: Kunal Pattanayak; Vikram Krishnamurthy
    Abstract: This paper establishes the equivalence between Bayesian revealed preference and classical revealed preference with non-linear budget constraints. Classical revealed preference tests for utility maximization given known budget constraints. Bayesian revealed preference tests for costly information acquisition given a utility function. Our main result shows that the key theorem in Caplin and Dean (2015) on Bayesian revealed preference is equivalent to Afriat-type feasibility inequalities for general (non-linear) budget sets. Our second result exploits this equivalence of classical and Bayesian revealed preference to construct a monotone convex information acquisition cost from decision maker's data in Bayesian revealed preference
    Date: 2021–06
  12. By: Daniel Lehmann
    Abstract: Choice functions over a set $X$ that satisfy the Outcast, a.k.a. Aizerman, property are exactly those that attach to any set its maximal subset relative to some total order of ${2}^{X}$.
    Date: 2021–07
  13. By: Florian Huber; Gary Koop
    Abstract: Macroeconomists using large datasets often face the choice of working with either a large Vector Autoregression (VAR) or a factor model. In this paper, we develop methods for combining the two using a subspace shrinkage prior. Subspace priors shrink towards a class of functions rather than directly forcing the parameters of a model towards some pre-specified location. We develop a conjugate VAR prior which shrinks towards the subspace which is defined by a factor model. Our approach allows for estimating the strength of the shrinkage as well as the number of factors. After establishing the theoretical properties of our proposed prior, we carry out simulations and apply it to US macroeconomic data. Using simulations we show that our framework successfully detects the number of factors. In a forecasting exercise involving a large macroeconomic data set we find that combining VARs with factor models using our prior can lead to forecast improvements.
    Date: 2021–07
  14. By: Sung Jae Jun (Institute for Fiscal Studies and Pennsylvania State University); Sokbae (Simon) Lee (Institute for Fiscal Studies and Columbia University and IFS)
    Abstract: We set up an econometric model of persuasion and study identification of key parameters under various scenarios of data availability. We find that a commonly used measure of persuasion does not estimate the persuasion rate of any population in general. We provide formal identification results, recommend several new parameters to estimate, and discuss their interpretation. Further, we propose methods for carrying out inference. We revisit the empirical literature on persuasion to show that the persuasive effect is highly heterogeneous. We also show that the existence of a continuous instrument opens up the possibility of point identification for the policy-relevant population.
    Date: 2019–12–09
  15. By: Sauter, Verena; Speth, Daniel; Plötz, Patrick; Signer, Tim
    Abstract: Facing climate change, The European Union has set ambitious greenhouse gas (GHG) reduction targets. Within Europe, heavy-duty vehicles (HDV) account for a quarter of greenhouse gas emissions in the transport sector and therefore plays a central role in achieving the climate targets. A potential solution to reduce GHG emissions is the use of battery electric vehicles (BEV). However, the limited range of BEV requires a European public fast-charging network to ensure widespread deployment of BEV. Here, European road freight transport flows are modelled based on the publicly available European Transport policy Information System (ETISplus) dataset. The resulting truck flows serve as input for a charging infrastructure network model. Potential charging stations are located using a coverage-oriented approach and sized according to a queuing model such that an average waiting time of five minutes is guaranteed at each location. Our results show that for a share of 15% BEV in HDV stock and a dense network with charging locations every 50 km, a total of 4,067 charging points at 1,640 locations are required by 2030. In contrast, with a share of 5% BEV and charging locations every 100 km, 1,715 charging points are needed at 812 locations. Our findings provide insights for the design of a public fastcharging network in Europe and thus supports the planning of future infrastructure projects.
    Date: 2021
  16. By: Eric Jacobson
    Abstract: Previous research has shown a relationship between voter characteristics and voter support for tax bonds. These findings, however, are difficult to interpret because of the high degree of collinearity across the measures. From 13 demographic measures of voters in a library bond election, seven independent principal components were extracted which accounted for 95 percent of the variance. Whereas the direct demographic measures showed inconsistent relationships with voting, the principal components of low SES, college experience, female and service job were related to affirmative voting, while high home value was related to negative voting.
    Date: 2021–06

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.