nep-dcm New Economics Papers
on Discrete Choice Models
Issue of 2010‒03‒28
seven papers chosen by
Philip Yu
Hong Kong University

  1. The distinction between dictatorial and incentive policy interventions and its implication for IV estimation By Christian Belzil; J. Hansen
  2. Willingness to Pay Estimation When Protest Beliefs are not Separable from the Public Good Definition By Kimberly Rollins; M.D.R. Evans; Mimako Kobayashi; Anita Castledine
  3. Measuring the Willingness to Pay to Avoid Guilt: Estimation Using Equilibrium and Stated Belief Models By Bellemare, Charles; Sebald, Alexander; Strobel, Martin
  4. Angler Heterogeneity and the Species-Specific Demand for Marine Recreational Fishing By Timothy Haab; Robert L. Hicks; Kurt Schnier; John C. Whitehead
  5. Limited Attention and Choice By Karen Kaiser
  6. The school reentry decision on poor girls: structural estimation and policy analysis using PROGRESA database By María Nieves Valdés
  7. Taxation and the Quality of Institutions: Asymmetric Effects on FDI By Serena Fatica

  1. By: Christian Belzil (Department of Economics, Ecole Polytechnique - CNRS : UMR7176 - Polytechnique - X, ENSAE - École Nationale de la Statistique et de l'Administration Économique - ENSAE, IZA - Institute for the Study of Labor); J. Hansen (IZA - Institute for the Study of Labor, CIREQ - Centre Interuniversitaire de Recherche en Economie Quantitative, CIRANO - Montréal, Department of Economics, Concordia University - Concordia University)
    Abstract: We investigate if, and under which conditions, the distinction between dictatorial and incentive-based policy interventions, affects the capacity of Instrument Variable (IV) methods to estimate the relevant treatment effect parameter of an outcome equation. The analysis is set in a non-trivial framework, in which the right-hand side variable of interest is affected by selectivity, and the error term is driven by a sequence of unobserved life-cycle endogenous choices. We show that, for a wide class of outcome equations, incentive-based policies may be designed so to generate a sufficient degree of post-intervention randomization (a lesser degree of selection on individual endowments among the sub-population affected). This helps the instrument to fulfill the orthogonality condition. However, for a same class of outcome equation, dictatorial policies that enforce minimum consumption cannot meet this condition. We illustrate these concepts within a calibrated dynamic life cycle model of human capital accumulation, and focus on the estimation of the returns to schooling using instruments generated from mandatory schooling reforms and education subsidies. We show how the nature of the skill accumulation process (substitutability vs complementarity) may play a fundamental role in interpreting IV estimates of the returns to schooling.
    Keywords: Returns to schooling, Instrumental Variable methods, Dynamic Discrete Choice, Dynamic Programming, Local Average Treatment Effects .
    Date: 2010–03–15
  2. By: Kimberly Rollins (Department of Resource Economics, University of Nevada, Reno); M.D.R. Evans (Department of Resource Economics and Department of Sociology, University of Nevada, Reno); Mimako Kobayashi (Department of Resource Economics, University of Nevada, Reno); Anita Castledine (Department of Resource Economics, University of Nevada, Reno)
    Abstract: Public good attributes that are correlated with protest beliefs but not separable from the good's value, would affect stated preference estimates of the WTP for the public good. Survey data collected to value a program to prevent ecosystem losses on Nevada rangelands, where the majority of land is publicly owned and managed, reveal more than half of the respondents exhibiting some protest belief. Of these, about 60% voted 'yes' to some nonzero bid amount. By treating protest beliefs and opposition to the proposed program as separate concepts, we systematically analyze their determinants and impacts on WTP. In this framework, people with protest beliefs may or may not vote 'no' to all bids and people may, without being protesters, answer 'no' to all dollar amounts. Multinomial logit regression results suggest that factors motivating people to protest and/or oppose the proposed program are so diverse that a single model does not provide a good fit. We estimate nested models and conclude that different underlying processes determine WTP for "protesters" ($34.02) and "non-protesters" ($69.56).
    Keywords: Stated preferences; Willingness to pay; Protest responses; Rangelands; Valuation of ecosystem services
    JEL: Q51 Q24 Q57
    Date: 2010–02
  3. By: Bellemare, Charles (Université Laval); Sebald, Alexander (University of Copenhagen); Strobel, Martin (Maastricht University)
    Abstract: We estimate structural models of guilt aversion to measure the population level of willingness to pay (WTP) to avoid feeling guilt by letting down another player. We compare estimates of WTP under the assumption that higher-order beliefs are in equilibrium (i.e. consistent with the choice distribution) with models estimated using stated beliefs which relax the equilibrium requirement. We estimate WTP in the later case by allowing stated beliefs to be correlated with guilt aversion, thus controlling for a possible source of a consensus effect. All models are estimated using data from an experiment of proposal and response conducted with a large and representative sample of the Dutch population. Our range of estimates suggests that responders are willing to pay between 0.40 and 0.80 Euro to avoid letting down proposers by 1 Euro. Furthermore, we find that WTP estimated using stated beliefs is substantially overestimated (by a factor of two) when correlation between preferences and beliefs is not controlled for. Finally, we find no evidence that WTP is significantly related to the observable socio-economic characteristics of players.
    Keywords: guilt aversion, willingness to pay, equilibrium and stated beliefs models
    JEL: C93 D63 D84
    Date: 2010–03
  4. By: Timothy Haab; Robert L. Hicks; Kurt Schnier; John C. Whitehead
    Abstract: In this study we assess the viability of single-species recreation demand models given commonly available data sets. Using the 2000 MRFSS southeast intercept data combined with the economic add-on, we determine that the MRFSS data will support only a few species-specific recreation demand models. Considering species of management interest in the southeast, we focus on dolphin, king mackerel, red snapper and red drum. We examine single-species recreational fishing behavior using random utility models of demand. We explore mixed logit (i.e., random parameter) logit and finite mixture (i.e., latent class logit) models for dealing with angler heterogeneity. We compare these to the commonly used conditional and nested logit models in terms of the value of catching (and keeping) one additional fish. Mixed logit models illustrate that the value of catch can be highly heterogeneous and, in some cases, can include both positive and negative values. The finite mixture model generates value estimates that were some times strikingly different than conditional, nested and mixed logit models. Preference heterogeneity is significant within the MRFSS data. We find evidence that single-species models outperform multiple species models and recreational values differ. Key Words: marine recreational fishing, single-species demand, preference heterogeneity models
    JEL: Q51 Q26
    Date: 2010
  5. By: Karen Kaiser
    Abstract: This paper analyzes a boundedly rational decision maker who is uncertain about his preference and faces the following trade-off: adding a good to the choice set has a positive option value but increases the complexity of the choice problem. The increased complexity is modeled as a reduction of the information available for each good. Because of this trade-off there is an optimal number of goods that the decision maker wants to analyze before making his final choice. The choice of the optimal set can be interpreted as the choice among stores. Stores maximize profits and choose a quality, an assortment, and a price. A lower cost of providing quality implies higher price and higher quality. Assortment will be small for very high levels of quality. Better quality of information implies greater variety and higher price. Greater variety combined with good consumer service can be a signal for high quality of the store.
    Keywords: Decision making, Bounded rationality, Choice set, Stores, Quality.
    JEL: D01 D11 D21
    Date: 2010–03
  6. By: María Nieves Valdés
    Abstract: In this paper I present a dynamic structural model of girls' schooling choices and estimate it using the Mexican PROGRESA database. This structural approach allows evaluating the efectiveness of several policies to increase school reentry rates for girls in low-income households. To increase school attendance among poor children in developing countries, policy makers have implemented conditional cash transfers programs. Although transfers have been successful in keeping girls at school, they do not increase school attendance among girls who have dropped out of school. Cash transfer programs may fail because most of these poor girls leave school to stay at home helping in housework, rather than working for a salary. Results suggest that effective policies to increase school reentry rates for poor girls are free access to community nurseries and kindergartens, and increasingg the availability of secondary schools.
    Keywords: Policy evaluation, Dynamic discrete choice structural models, School choices for girls, School reentry, PROGRESA
    JEL: I21 I28 J16 O15
    Date: 2009–10
  7. By: Serena Fatica (European Commission)
    Abstract: Economic integration has intensified international competition to attract productive capital. This paper analyzes, both theoretically and empirically, the effect of tax policies and institutional quality on the allocation of FDI - two aspects that the economic literature has extensively investigated, though only in isolation. I build a simple two-country partial equilibrium model to study competition among governments vying for potential investors whose location choices are driven by both the quality of institutions and the corporate tax rate. Modeling good governance as a public good, it is shown that the jurisdiction providing better institutions is able to levy a higher tax on capital. Moreover, provided firms are sensitive enough to institutional quality, it attracts a larger share of investment than the low-quality/low-tax location. The main predictions of the model are tested on FDI stocks to 63 economies using a simple difference gravity equation derived from discrete choice theory of firms' location. Using a pair of destination countries as the unit of analysis eliminates the need to control for multilateral interdependence among receiving countries, a source of possible bias in the traditional gravity specification in the levels. The empirical evidence corroborates the claim that the sensitivity of foreign investment to the tax rate varies significantly between host countries characterized by different levels of institutional quality. The findings are robust to a number of sensitivity checks and to the use of instrumental variables to tackle endogeneity of the institutional quality variable.
    Keywords: foreign direct investment, fiscal competition, institutions, public goods
    JEL: H7 F21 F23 K00
    Date: 2010–03

This nep-dcm issue is ©2010 by Philip Yu. 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.