nep-dcm New Economics Papers
on Discrete Choice Models
Issue of 2018‒07‒30
five papers chosen by
Edoardo Marcucci
Università degli studi Roma Tre

  1. Dynamic decisions under subjective expectations: a structural analysis By Yonghong An; Yingyao Hu; Ruli Xiao
  2. Specification test on mixed logit models By Jinyong Hahn; Jerry Hausman; Josh Lustig
  3. Sovereign Debt Restructurings By Dvorkin, Maximiliano; Sanchez, Juan M.; Sapriza, Horacio; Yurdagul, Emircan
  4. Sectoral choice and selectivity By Nirodha Bandara; Simon Appleton; Trudy Owens
  5. The Effect of Education on Health: Evidence from National Compulsory Schooling Reforms By Raquel Fonseca; Pierre-Carl Michaud; Yuhui Zheng

  1. By: Yonghong An (Institute for Fiscal Studies); Yingyao Hu (Institute for Fiscal Studies and Johns Hopkins University); Ruli Xiao (Institute for Fiscal Studies and Indiana University)
    Abstract: This paper studies dynamic discrete choices by relaxing the assumption of rational expectations. That is, agents' subjective expectations about the state transition are unknown and allowed to differ from their objectively estimable counterparts. We show that agents' subjective expectations and preferences can be identi ed and estimated from the observed conditional choice probabilities in both finite and infi nite horizon cases. Our identi cation of subjective expectations is nonparametric and can be expressed as a closed-form function of the observed conditional choice probabilities. We estimate the model primitives using maximum likelihood estimation and illustrate the good performance of estimators using Monte Carlo experiments. We apply our model to Panel Study of Income Dynamics (PSID) data and analyze women's labor participation. We find systematic differences between agents' subjective expectations about their income transition from those under rational expectations. A counterfactual analysis suggests that women with low and medium incomes would increase the probability of working under rational expectations, and that the probability would decrease for women with high income.
    Keywords: Dynamic discrete choice models, subjective expectations, rational expectations, nonparametric identi cation, estimation.
    Date: 2018–02–05
  2. By: Jinyong Hahn (Institute for Fiscal Studies); Jerry Hausman (Institute for Fiscal Studies and MIT); Josh Lustig (Institute for Fiscal Studies)
    Abstract: This paper proposes a specification test of the mixed logit models, by generalizing Hausman and McFadden?s (1984) test. We generalize the test even further by considering a model developed by Berry, Levinsohn and Pakes (1995).
    Date: 2017–12–12
  3. By: Dvorkin, Maximiliano (Federal Reserve Bank of St. Louis); Sanchez, Juan M. (Federal Reserve Bank of St. Louis); Sapriza, Horacio (Federal Reserve Board); Yurdagul, Emircan (Universidad Carlos III)
    Abstract: Sovereign debt crises generally involve debt restructurings characterized by a mix of face-value haircuts and debt maturity extensions. We develop a quantitative model of endogenous sovereign debt maturity choice and restructuring that captures key stylized facts of debt over the business cycle and during restructuring episodes, including the variation of haircuts, maturity extensions and default duration found in the data. We also find that policy interventions implementing minimum haircuts and redistributing losses away from holders of short term debt improve the outcome of distressed debt restructurings and reduce the frequency of debt distress events. Methodologically, the use of dynamic discrete choice solution methods allows us to smooth decision rules on default and debt portfolio choices, rendering the problem tractable.
    Keywords: Crises; Default; Sovereign Debt; Restructuring; Rescheduling; Country Risk; Maturity; International Monetary Fund; Dynamic Discrete Choice
    JEL: F34 F41 G15
    Date: 2018–06–25
  4. By: Nirodha Bandara; Simon Appleton; Trudy Owens
    Abstract: This paper attempts to examine the labour force participation decisions and earnings across employment sectors and how it varies by gender in Sri Lanka. The labour market is disaggregated into 5 sectors –public, formal private, informal private, self-employed and agriculture. Using the Labour Force Survey 2013, this paper adds to existing literature in two ways. Firstly, the paper deals with two forms of potential biases which have not been simultaneously explored for the case of Sri Lanka – sample selectivity and endogeneity of education in earnings. Secondly, it adds to the literature by including the self-employed in the analysis. The determinants of sector choice are analysed using a multinomial logit. The findings of this paper suggest that individuals with the highest levels of education get into the public and formal private sectors, whereas the least educated are likely to join the informal and agricultural sectors. The earnings functions suggest that the returns to education vary greatly across the sectors. The differences across sectors confirm the importance of disaggregating the sectors of employment to examine choices of labour force participation.
    Keywords: sample selectivity, endogeneity, multinomial logit, control function, Sri Lanka
    Date: 2018
  5. By: Raquel Fonseca; Pierre-Carl Michaud; Yuhui Zheng
    Abstract: This paper sheds light on the causal relationship between education and health outcomes. We combine three surveys (SHARE, HRS and ELSA) that include nationally representative samples of people aged 50 and over from fifteen OECD countries. We use variations in the timing of educational reforms across these countries as an instrument for education. Using IV-Probit models, we find causal evidence that more years of education lead to a lower probability of reporting poor health, less likely of having limitations in functional status (ADL and iADLs), and lower prevalence for diabetes. These effects are larger than those from the Probit that do not control for the endogeneity of education. The relationship between education and cancer is positive in both Probit and IV-Probit models. The causal impacts of education on other chronic conditions as well as functional status are not established using IV-Probit models.
    Keywords: education, health, causality, compulsory schooling laws
    JEL: I1 I2
    Date: 2018

This nep-dcm issue is ©2018 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.