nep-dcm New Economics Papers
on Discrete Choice Models
Issue of 2022‒05‒02
nine papers chosen by
Edoardo Marcucci
Università degli studi Roma Tre

  1. Bootstrap inference for fixed-effect models By Jochmans, Koen; Higgins, Ayden
  2. Nonparametric Analysis of Dynamic Random Utility Models By Nail Kashaev; Victor H. Aguiar
  3. How much choice is enough? Parental satisfaction with secondary school choice in England and Scotland By Bhattacharya, Aveek
  4. Do economics and political science scholars differ on public choice issues? Survey evidence from Brazil By Abdel-Hameed Nawar
  5. A class of dissimilarity semimetrics for preference relations By Hiroki Nishimura; Efe A. Ok
  6. Economic Benefits of Direct Current Technology for Private Households and Peer-to-Peer Trading in Germany By Liu, Xueying; Madlener, Reinhard
  7. Fast Simulation-Based Bayesian Estimation of Heterogeneous and Representative Agent Models using Normalizing Flow Neural Networks By Cameron Fen
  8. Automatic Debiased Machine Learning for Dynamic Treatment Effects By Rahul Singh; Vasilis Syrgkanis
  9. Dads and Daughters: Disentangling Altruism and Investment Motives for Spending on Children By Rebecca Dizon-Ross; Seema Jayachandran

  1. By: Jochmans, Koen; Higgins, Ayden
    Abstract: The maximum-likelihood estimator of nonlinear panel data models with fixed effects is asymptotically biased under rectangular-array asymptotics. The literature has devoted substantial effort to devising methods to correct the maximum-likelihood estimator for its bias as a means to salvage standard inferential procedures. We show that the (recursive, parametric) bootstrap replicates the distribution of the (uncorrected) maximum-likelihood estimator in large samples. This justifies the use of confidence sets constructed via conventional bootstrap methods. No adjustment for the presence of bias needs to be made.
    Keywords: Bootstrap,;fixed effects; incidental parameter problem; inference, panel data
    JEL: C23
    Date: 2022–04
  2. By: Nail Kashaev; Victor H. Aguiar
    Abstract: We study a dynamic generalization of stochastic rationality in consumer behavior, the Dynamic Random Utility Model (DRUM). Under DRUM, a consumer draws a utility function from a stochastic utility process and maximizes this utility subject to her budget constraint in each time period. Utility is random, with unrestricted correlation across time periods and unrestricted heterogeneity in a cross-section. We provide a revealed preference characterization of DRUM when we observe a panel of choices from budgets. This characterization is amenable to statistical testing. Our result unifies Afriat's (1967) theorem that works with time-series data and the static random utility framework of McFadden-Richter (1990) that works with cross-sections of choice.
    Date: 2022–04
  3. By: Bhattacharya, Aveek
    Abstract: Governments around the world have sought to promote school choice, not just in order to improve educational outcomes, but also because such choice is believed to be intrinsically valuable: parents are believed to want to choice and to feel empowered by it. This article empirically evaluates the intrinsic value of school choice, comparing the attitudes and experiences of parents in England (where expanding choice is an explicit policy goal) and Scotland (where policymakers tend to play down choice), combining an online survey with in-depth interviews. While the overwhelming majority of parents in both countries express a desire for some school choice, only a minority want choice primarily for intrinsic reasons. Rather, most believe it is necessary to avoid negative outcomes for their children. Moreover, while parents in England tend to say they have more choice than their Scottish counterparts, they are no more satisfied with the level of choice that they have. Indeed, they tend to be more cynical, fatalistic and disempowered. Based on the British experience, school choice policies have not been successful in promoting intrinsic value.
    Keywords: autonomy; intrinsic value; quasi-markets; school choice; studentship
    JEL: N0
    Date: 2021–12–06
  4. By: Abdel-Hameed Nawar (IPC-IG)
    Keywords: public choice; rationality; free riding; public goods; politics-as-exchange
    Date: 2021–10
  5. By: Hiroki Nishimura; Efe A. Ok
    Abstract: We propose a class of semimetrics for preference relations any one of which is an alternative to the classical Kemeny-Snell-Bogart metric. (We take a fairly general viewpoint about what constitutes a preference relation, allowing for any acyclic order to act as one.) These semimetrics are based solely on the implications of preferences for choice behavior, and thus appear more suitable in economic contexts and choice experiments. In our main result, we obtain a fairly simple axiomatic characterization for the class we propose. The apparently most important member of this class (at least in the case of finite alternative spaces), which we dub the top-difference semimetric, is characterized separately. We also obtain alternative formulae for it, and relative to this metric, compute the diameter of the space of complete preferences, as well as the best transitive extension of a given acyclic preference relation. Finally, we prove that our preference metric spaces cannot be isometically embedded in a Euclidean space.
    Date: 2022–03
  6. By: Liu, Xueying (E.ON Energy Research Center, Future Energy Consumer Needs and Behavior (FCN)); Madlener, Reinhard (E.ON Energy Research Center, Future Energy Consumer Needs and Behavior (FCN))
    Abstract: With the increased adoption of solar photovoltaics (PV) and batteries, and the use of electronics and appliances powered by direct current (DC), e.g. heat pumps, and electric vehicles (EVs), DC technologies offer higher energy efficiency compared to the entrenched alternating current (AC) technologies. However, the adoption of DC infrastructure is limited due to path dependency and lock-in effects of the currently dominant electric infrastructure based on AC technology. Efficiency gains in energy communities and for households may facilitate the wider-scale adoption of DC technologies. In this study, we simulate 600 household load profiles based on twelve different representative household types and estimate the possible energy cost savings of a DC architecture compared to an AC architecture under various electricity prices and feed-in-tariff levels. This is done for different combinations of battery and PV sizes, and for the case of a peer-to-peer (P2P) trading community. The results show that the DC home yields cost savings of around €90 p.a. for the median household when compared to an AC home. Moreover, we find that neither the share of DC load nor household characteristics impacts cost savings significantly, while the total load remains the most important factor influencing the cost-saving potential. In addition, while cost savings do not necessarily increase with larger PV and battery sizes, they do increase with the possibility of households to engage in P2P trading. The results yield an improved understanding regarding the cost-saving potentials of DC homes and their expected diffusion in Germany. This is especially relevant for future large-scale adoption of solar PV, batteries, and EVs in the future, thus helping both policy-makers and companies alike to better assess the market potential of DC homes.
    Keywords: DC technology; Choice of Technology; Diffusion; Industrial policy; Path dependence
    JEL: O14 O25 O33 O52
    Date: 2021–09
  7. By: Cameron Fen
    Abstract: This paper proposes a simulation-based deep learning Bayesian procedure for the estimation of macroeconomic models. This approach is able to derive posteriors even when the likelihood function is not tractable. Because the likelihood is not needed for Bayesian estimation, filtering is also not needed. This allows Bayesian estimation of HANK models with upwards of 800 latent states as well as estimation of representative agent models that are solved with methods that don't yield a likelihood--for example, projection and value function iteration approaches. I demonstrate the validity of the approach by estimating a 10 parameter HANK model solved via the Reiter method that generates 812 covariates per time step, where 810 are latent variables, showing this can handle a large latent space without model reduction. I also estimate the algorithm with an 11-parameter model solved via value function iteration, which cannot be estimated with Metropolis-Hastings or even conventional maximum likelihood estimators. In addition, I show the posteriors estimated on Smets-Wouters 2007 are higher quality and faster using simulation-based inference compared to Metropolis-Hastings. This approach helps address the computational expense of Metropolis-Hastings and allows solution methods which don't yield a tractable likelihood to be estimated.
    Date: 2022–03
  8. By: Rahul Singh; Vasilis Syrgkanis
    Abstract: We extend the idea of automated debiased machine learning to the dynamic treatment regime. We show that the multiply robust formula for the dynamic treatment regime with discrete treatments can be re-stated in terms of a recursive Riesz representer characterization of nested mean regressions. We then apply a recursive Riesz representer estimation learning algorithm that estimates de-biasing corrections without the need to characterize how the correction terms look like, such as for instance, products of inverse probability weighting terms, as is done in prior work on doubly robust estimation in the dynamic regime. Our approach defines a sequence of loss minimization problems, whose minimizers are the mulitpliers of the de-biasing correction, hence circumventing the need for solving auxiliary propensity models and directly optimizing for the mean squared error of the target de-biasing correction.
    Date: 2022–03
  9. By: Rebecca Dizon-Ross; Seema Jayachandran
    Abstract: This paper tests whether mothers and fathers differ in their spending on their daughters relative to their sons. We compare mothers’ and fathers’ willingness to pay (WTP) for specific goods for their children, diverging from the previous literature’s approach of comparing the expenditure effects of mothers’ versus fathers’ income. Our method, which we apply in Uganda, allows us to estimate gender differences and explore mechanisms with greater precision. A second innovation is that we examine why spending patterns differ between mothers and fathers, e.g., altruism, personal returns to investing in children. We find that fathers have a lower WTP for their daughters’ human capital than their sons’ human capital, whereas mothers do not. We also find evidence that altruism plays a role in the mother-father differences: fathers’ WTP for goods that simply bring joy to their daughters is lower than their WTP for such goods for their sons, but mothers’ is not.
    JEL: J13 J16
    Date: 2022–04

This nep-dcm issue is ©2022 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.