|
on Discrete Choice Models |
| By: | Valentino Dardanone (Università di Palermo); Carla Guerriero (Università di Napoli Federico II and CSEF) |
| Abstract: | Young generations will bear the cost of present natural capital degradation and, as the recent wave of school climate strikes for climate change proved, do not want their voices to be ignored. Discrete Choice Experiments are increasingly being used for the valuation of environmental goods, nevertheless, they have never been conducted with children. We designed and administered a discrete choice experiment to elicit children, aged 8-19 years, willingness to pay (WTP) for environmental protection projects. Our results suggest that children marginal WTP is higher for projects targeting natural protection in their own country (Italy) and that the utility of environmental protection is greater for females and for older children. Furthermore, we find that individual attitude towards environment negatively affect the probability of choosing the status quo alternative. Given recent findings on transfer of knowledge, attitudes and behaviours towards environmental protection from children to parents, these results are important to support policy makers decisions on how to deal with the issues of natural capital degradation. |
| Keywords: | Discrete Choice Experiment; Children; Natural Capital; Environmental Protection; Willingness to Pay |
| JEL: | C93 Q51 D83 |
| Date: | 2019–12–17 |
| URL: | http://d.repec.org/n?u=RePEc:sef:csefwp:550&r=all |
| By: | Eleni Aristodemou (Institute for Fiscal Studies and University College London); Adam Rosen (Institute for Fiscal Studies and Duke University) |
| Abstract: | In this paper we analyze a discrete choice model for partially ordered alternatives. The alternatives are di?erentiated along two dimensions, the ?rst an unordered “horizontal” dimension, and the second an ordered “vertical” dimension. The model can be used in circumstances in which individuals choose amongst products of di?erent brands, wherein each brand o?ers an ordered choice menu, for example by o?ering products of varying quality. The unordered-ordered nature of the discrete choice problem is used to characterize the identi?ed set of model parameters. Following an initial nonparametric analysis that relies on shape restrictions inherent in the ordered dimension of the problem, we then provide a specialized analysis for a parametric generalization of the ordered probit model. Conditions for point identi?cation are established when the distribution of unobservable heterogeneity is known, but remain elusive when the distribution is instead restricted to the multivariate normal family with parameterized variance. Rather than invoke the restriction that the distribution is known, or simply assume that model parameters are point identi?ed, we consider the use of inference methods that allow for the possibility of set identi?cation, and which are therefore robust to the possible lack of point identi?cation. A Monte Carlo analysis is provided in which inference is carried out using a method proposed by Chen, Christensen, and Tamer (2018), which is insensitive to the possible lack of point identi?cation and is found to perform adequately. An empirical illustration is then conducted using consumer purchase data in the UK to study consumers’ choice of razor blades in which each brand has product o?erings vertically di?erentiated by quality. |
| Date: | 2019–11–18 |
| URL: | http://d.repec.org/n?u=RePEc:ifs:cemmap:62/19&r=all |
| By: | Joel L. Horowitz (Institute for Fiscal Studies and Northwestern University); Lars Nesheim (Institute for Fiscal Studies and cemmap and UCL) |
| Abstract: | The multinomial logit model with random coefficients is widely used in applied research. This paper is concerned with estimating a random coefficients logit model in which the distribution of each coefficient is characterized by finitely many parameters. Some of these parameters may be zero. The paper gives conditions under which with probability approaching 1 as the sample size approaches infinity, penalized maximum likelihood (PML) estimation with the adaptive LASSO (AL) penalty function distinguishes correctly between zero and non-zero parameters in a random coefficients logit model. If one or more parameters are zero, then PML with the AL penalty function often reduces the asymptotic mean-square estimation error of any continuously differentiable function of the model’s parameters, such as a market share or an elasticity. The paper describes a method for computing the PML estimates of a random coefficients logit model. It also presents the results of Monte Carlo experiments that illustrate the numerical performance of the PML estimates. Finally, it presents the results of PML estimation of a random coefficients logit model of choice among brands of butter and margarine in the British groceries market. |
| Date: | 2019–10–15 |
| URL: | http://d.repec.org/n?u=RePEc:ifs:cemmap:50/19&r=all |
| By: | Brynjolfsson, Erik; Collis, Avinash; Eggers, Felix |
| Abstract: | GDP and derived metrics such as productivity have been central to our understanding of economic progress and well-being. In principle, changes in consumer surplus provide a superior, and more direct, measure of changes in well-being, especially for digital goods. In practice, these alternatives have been difficult to quantify. We explore the potential of massive online choice experiments to measure consumer surplus. We illustrate this technique via several empirical examples which quantify the valuations of popular digital goods and categories. Our examples include incentive compatible discrete choice experiments where online and lab participants receive monetary compensation if and only if they forgo goods for pre-defined periods. For example, the median user needed a compensation of about $48 to forgo Facebook for one month. Our overall analyses reveal that digital goods have created large gains in well-being that are not reflected in conventional measures of GDP and productivity. By periodically querying a large, representative sample of goods and services, including those which are not priced in existing markets, changes in consumer surplus and other new measures of well-being derived from these online choice experiments have the potential for providing cost-effective supplements to the existing National Income and Product Accounts. |
| Date: | 2019–04–09 |
| URL: | http://d.repec.org/n?u=RePEc:osf:osfxxx:akqhn&r=all |
| By: | Abi Adams (Institute for Fiscal Studies and University of Oxford); Alison Andrew (Institute for Fiscal Studies and Institute for Fiscal Studies) |
| Abstract: | Rajasthani women typically leave school early and marry young. We develop a novel discrete choice methodology using hypothetical vignettes to elicit average parental preferences over a daughter’s education and age of marriage, and subjective beliefs about the evolution of her marriage market prospects. We find parents have a strong preference for delaying a daughter’s marriage until eighteen but no further. Conditional on a marriage match, parents place little intrinsic value on a daughter’s education. However, they believe the probability of receiving a good marriage offer increases strongly with a daughter’s education but deteriorates quickly with her age on leaving school. |
| Keywords: | marriage markets, preferences |
| JEL: | J12 J16 I26 |
| Date: | 2019–03–07 |
| URL: | http://d.repec.org/n?u=RePEc:ifs:ifsewp:19/05&r=all |
| By: | Adam Zylbersztejn (Univ Lyon, Université Lumière Lyon 2, GATE L-SE UMR 5824, 69130 Ecully, France); Zakaria Babutsidze (SKEMA Business School, Université Côte d'Azur (GREDEG) and OFCE, Sciences Po Paris); Nobuyuki Hanaki (Institute of Social and Economic Research, Osaka University) |
| Abstract: | We experimentally investigate how much value people put in observable information about others in strategic interactions. The incentivized experimental task is to predict an unknown target player's trustworthiness in an earlier hidden action game. In Experiment 1, we vary the source of information about the target player (neutral picture, neutral video, video containing strategic content). The observed prediction accuracy rates then serve as an empirical measure of the objective value of information. In Experiment 2, we elicit the subjective value of information using the standard stated preferences method ("willingness to accept"). While the elicited subjective values are ranked in the same manner as the objective ones, subjects attach value to information which does not help predict target behavior, and exaggerate the value of helpful information. |
| Keywords: | prediction, observable information, individual characteristics, stated preferences, willingness to accept, experiment |
| JEL: | C72 D83 |
| Date: | 2019 |
| URL: | http://d.repec.org/n?u=RePEc:gat:wpaper:1936&r=all |
| By: | Aurélien Baillon (Erasmus University Rotterdam); Aleli Kraft (University of the Philippines Diliman); Owen O'Donnell (Erasmus University Rotterdam); Kim van Wilgenburg (Erasmus University Rotterdam) |
| Abstract: | Despite widespread exposure to substantial medical expenditure risk in low-income populations, health insurance enrollment is typically low. This is puzzling from the perspective of expected utility theory. To help explain it, this paper introduces a decomposition of the stated willingness to pay (WTP) for insurance into its fair price and three behavioral deviations from that price due to risk perception and risk attitude consistent with prospect theory, plus a residual. To apply this approach, we elicit WTP, subjective distributions of medical expenditures and risk attitude (utility curvature and probability weighting) from Filipino households in a nationwide survey. We find that the mean stated WTP of the uninsured is less than both the actuarially fair price and the subsidized price at which public insurance is offered. This is not explained by downwardly biased beliefs: both the mean and the median subjective expectation are greater than the subsidized price. Convex utility in the domain of losses pushes mean WTP below the fair price and the subsidized price, and the transformation of probabilities into decision weights depresses the mean further, at least using one of two specific decompositions. WTP is reduced further by factors other than risk perception and attitude. |
| Keywords: | Health insurance, willingness to pay, subjective probability, prospect theory, medical expense |
| JEL: | I13 D84 |
| Date: | 2019–11–17 |
| URL: | http://d.repec.org/n?u=RePEc:tin:wpaper:20190077&r=all |
| By: | Chunrong Ai (Institute for Fiscal Studies); Oliver Linton (Institute for Fiscal Studies and University of Cambridge); Kaiji Motegi (Institute for Fiscal Studies); Zheng Zhang (Institute for Fiscal Studies) |
| Abstract: | This paper presents a weighted optimization framework that unifies the binary, multi-valued, continuous, as well as mixture of discrete and continuous treatment, under unconfounded treatment assignment. With a general loss function, the framework includes the average, quantile and asymmetric least squares causal effect of treatment as special cases. For this general framework, we first derive the semiparametric efficiency bound for the causal effect of treatment, extending the existing bound results to a wider class of models. We then propose a generalized optimization estimator for the causal effect with weights estimated by solving an expanding set of equations. Under some sufficient conditions, we establish the consistency and asymptotic normality of the proposed estimator of the causal effect and show that the estimator attains the semiparametric efficiency bound, thereby extending the existing literature on efficient estimation of causal effect to a wider class of applications. Finally, we discuss estimation of some causal effect functionals such as the treatment effect curve and the average outcome. To evaluate the finite sample performance of the proposed procedure, we conduct a small-scale simulation study and find that the proposed estimation has practical value. To illustrate the applicability of the procedure, we revisit the literature on campaign advertising and campaign contributions. Unlike the existing procedures, which produce mixed results, we find no evidence of campaign advertising on campaign contribution. |
| Date: | 2019–11–29 |
| URL: | http://d.repec.org/n?u=RePEc:ifs:cemmap:64/19&r=all |