nep-dcm New Economics Papers
on Discrete Choice Models
Issue of 2007‒04‒21
twelve papers chosen by
Philip Yu
Hong Kong University

  1. El mercado de trabajo de los economistas: análisis de los factores determinantes de la duración del primer desempleo. By Cristina Borra Marcos; Francisco Gómez García; Manuel Salas Velasco
  2. Location Choices of the Pharmaceutical Industry in Europe after 1992 By Frances Ruane; Xiaoheng Zhang
  3. Agency in Health-Care: Are Medical Care-Givers Perfect Agents? By Einat Neuman; Shoshana Neuman
  4. Mums and Their Sons, Dads and Their Daughters: Panel Data Evidence of Interdependent Marginal Utilities across 14 EU Countries By José Alberto Molina; María Navarro; Ian Walker
  5. Maximum Fee vs. Child Benefit: A Welfare Analysis of Swedish Child-Care Fee Reform By Anna Brink; Katarina Nordblom; Roger Wahlberg
  6. A Trade-by-Trade Surprise Measure and Its Relation to Observed Spreadson the NYSE By Valeri Voev
  7. "Misclassification of the Dependent Variable in Binary Choice Models: Evidence from Five Latin American Countries" By Evangelos M. Falaris
  8. Is Fairness in the Eye of the Beholder? An Impartial Spectator Analysis of Justice By Konow, James
  9. The Single-mindedness theory: empirical evidence from the U.K. By Emanuele, Canegrati
  10. The Socioeconomic Determinants of Individual Environmental Concern: Evidence from Shanghai Data By Junyi Shen; Tatsuyoshi Saijo
  11. Deriving welfare measures from stated preference discrete choice modelling experiments, CHERE Discussion Paper No 48 By Emily Lancsar
  12. Men?s preferences for treatment of early stage prostate cancer: Results from a discrete choice experiment, CHERE Working Paper 2006/14 By Madeleine King; Rosalie Viney; Ishrat Hossain; David Smith; Sandra Fowler; Elizabeth Savage; Bruce Armstrong

  1. By: Cristina Borra Marcos (Universidad de Sevilla); Francisco Gómez García (Universidad de Sevilla); Manuel Salas Velasco (Universidad de Sevilla)
    Abstract: Este trabajo examina los determinantes de la duración del primer desempleo de una cohorte de jóvenes graduados: los que obtuvieron su titulación en el curso 2001/2002 en las carreras del área económica de la Universidad de Sevilla. Tras estimar un logit ordenado para datos de duración se obtiene la conclusión de que la nota media del egresado es un favor que coadyuva a la inserción laboral. Además, haber trabajado durante los estudios acelera el proceso de inserción laboral, al igual que haber hecho prácticas durante la carrera. Se tarda más en conseguir el primer empleo cuanto menor es el nivel educativo de la madre. Además, no se observan diferencias estadísticamente significativas en el acceso al primer empleo por género o según el tipo de estudios finalizados en el área económica.
    Keywords: duración del desempleo, transición universidad-empleo, microeconometría,modelo logit ordenado, unemployment, duration, university-to-work transition, microeconometrics,ordered logit model.
    JEL: J64 J44
    Date: 2007
  2. By: Frances Ruane; Xiaoheng Zhang
    Abstract: Differences in regulations,technical standards and national medical cultures across EU member states created a highly segmented pharmaceutical market in Europe prior to the implementation of the Single Market Programme. The subsequent reduction in non-tariff barriers to trade would be expected to have an impact on where pharmaceutical multinationals locate production within the EU.Using discrete choice models, we study separately the determinants of multinational location choices in terms of expanded production at existing facilities and location of start-up firms.Our results support the findings of models which predict reduced rather than increased agglomeration in the face of trade-cost reductions.
    Date: 2007–04–17
  3. By: Einat Neuman (Academic College of Tel-Aviv-Yaffo); Shoshana Neuman (Bar-Ilan University, CEPR and IZA)
    Abstract: It has been suggested in the literature that a source of incompleteness in the agency relationship between the doctor and the patient is that the provider may respond to an incomplete or biased perception of the patient’s interests. However, this has not been shown empirically. This paper is novel in presenting an empirical test of the fundamental assumption of the agency model that health care professionals understand what their patients want. Discrete Choice Experiments (DCEs) are conducted simultaneously within samples of patients (women who gave birth) and care-givers (doctors and nurses), to elicit and contrast patients’ authentic preferences (for five maternity ward attributes) with what care-givers believe them to be. Conclusion: agents have a biased perception of principals’ preferences, and therefore a complete agency relationship does not exist. Our findings add a novel empirical contribution to the agency relationship literature. Moreover, parallel preference patterns of patients and care-givers are certainly of much interest to the field of health economics: Informing the unaware medical care-givers about the patients' preferences, will improve treatment and patients' satisfaction.
    Keywords: principal-agent relationship, health-care, maternity wards, discrete choice experiment, preferences
    JEL: I1
    Date: 2007–04
  4. By: José Alberto Molina (University of Zaragoza and IZA); María Navarro (FEDEA, Madrid); Ian Walker (University of Warwick, Princeton University and IZA)
    Abstract: We study how fathers and mothers income satisfaction correlates with the income satisfaction of their sons and daughters, as well as with other economic and sociodemographic variables. We estimate these correlations using data on parents and children in households surveyed in the eight waves of the European Community Household Panel- ECHP (1994-2001) for 14 EU countries. To assess the robustness of these correlations, we use siblings in the Panel and we investigate the sensitivity of the estimates with the inclusion of other control variables. We also adopt a multi-level random effects ordered probit specification, that uses step-parents in the data, to allow us to distinguish nature effects from nurture effects. Our main results show evidence of strong altruism effects, but these estimated effects differ across countries, differ between mothers and fathers, and differ between sons and daughters.
    Keywords: parents and children, income satisfaction, interdependent marginal utilities, altruism, Europe
    JEL: D13 D60 D64 C33
    Date: 2007–04
  5. By: Anna Brink (Ministry of Finance, Sweden); Katarina Nordblom (Göteborg University); Roger Wahlberg (Göteborg University and IZA)
    Abstract: The effects of a recent Swedish child-care fee reform are compared with those of an alternative reform, increased child benefits. The fee reform implied considerably decreased fees and was intended to increase both labor supply among parents and their economic wellbeing. We estimate labor supply effects using a discrete choice labor supply model, and simulate behavioral responses to the changes. We find positive, but small, effects on labor supply from reduced fees, while increased child benefits would make single mothers decrease their labor supply. On the other hand, increased child benefits would make income distribution more equal. We make a social welfare comparison and conclude that for plausible values of inequality aversion, the alternative reform would have been preferable to the implemented fee reform.
    Keywords: labor supply, redistribution, reform, child care, fees, child benefit
    JEL: H31 I38 J22
    Date: 2007–04
  6. By: Valeri Voev (University of Konstanz)
    Abstract: We analyze the relationship between spreads and an indicator for information based transactions on trade-by-trade data. Classifying trades on the NYSE in six categories with respect to their volume relative to the quoted depth, we employ an ordered probit model to predict the category of a trade given the current market conditions. This approach allows us to test certain market microstructure hypothesis on the determinants of the buy-sell pressure. The difference between the predicted and the actual trade category (the surprise) is found to have explanatory power for the observed spreads beyond raw volume, volume relative to the quoted depth, and previous trading volume. The positive effect of the previous surprise on the observed spreads confirms the hypothesis that market-makers react to the increased probability of having traded with an informed trader by widening the spread.
    Date: 2006–03–14
  7. By: Evangelos M. Falaris (Department of Economics,University of Delaware)
    Abstract: Misclassification of the dependent variable in binary choice models can result in inconsistency of the parameter estimates. I estimate probit models that treat misclassification probabilities as estimable parameters for three labor market outcomes: formal sector employment, pension contribution and job change. I use Living Standards Measurement Study data from Nicaragua, Peru, Brazil, Guatemala, and Panama. I find that there is significant misclassification in eleven of the sixteen cases that I investigate. If misclassification is present, but is ignored, estimates of the probit parameters and their standard errors are biased toward zero. In most cases, predicted probabilities of the outcomes are significantly affected by misclassification of the dependent variable. Even a moderate degree of misclassification can have substantial effects on the estimated parameters and on many of the predictions.
    Keywords: Data Quality; Misclassification; Formal Sector; Pension Contributor; Job Change; Nicaragua; Peru; Brazil; Guatemala; Panama
    JEL: C81 C25 O17 J26 J62
  8. By: Konow, James
    Abstract: A popular sentiment is that fairness is inexorably subjective and incapable of being determined by objective standards. This study, on the other hand, seeks to establish evidence on unbiased justice and to propose and demonstrate a general approach for measuring impartial views empirically. Most normative justice theories associate impartiality with limited information and with consensus, i.e., a high level of agreement about what is right. In both the normative and positive literature, information is usually seen as the raw material for self-serving bias and disagreement. In contrast, this paper proposes a type of impartiality that is associated with a high level of information. The crucial distinction is the emphasis here on the views of impartial spectators, rather than implicated stakeholders. I describe the quasi-spectator method, i.e., an empirical means to approximate the views of impartial spectators that is based on a direct relationship between information and consensus, whereby consensus refers to the level of agreement among actual evaluators of real world situations. Results of surveys provide evidence on quasi-spectator views and support this approach as a means to elicit moral preferences. By establishing a relationship between consensus and impartiality, this paper seeks to help lay an empirical foundation for welfare analysis, social choice theory and practical policy applications.
    Keywords: Justice; fairness; impartial spectator
    JEL: D61 A12 D63
    Date: 2006–12
  9. By: Emanuele, Canegrati
    Abstract: In this paper I will exploit answers coming from the British Election Study in order to assess the validity of the Single Mindedness Theory. In particular, I will evaluate whether political preferences of voters for political candidates depend on their age and some other characteristics such as gender, education, religion, social and economic conditions. Performing LOGIT and PROBIT regression I will demonstrate that variable age is statistically significant, demonstrating that Single Mindedness Theory assumptions hold in the UK political environment.
    Keywords: Single-mindedness; political survey; electorate preferences; Logit; Probit
    JEL: H31 D72 D83 H56 H51 C35 C25 D78 J14
    Date: 2007–04
  10. By: Junyi Shen (Osaka School of International Public Policy, Osaka University); Tatsuyoshi Saijo (ISER, Osaka University)
    Abstract: This study examines the influence of socioeconomic characteristics on eleven measures of environmental concern by applying a pooled sample of 1200 individuals in Shanghai, China. Previous studies, which made efforts to explain environmental concern as a function of social structure, suggest that there are traditionally five hypotheses (the age, gender, social class, residence, and political hypotheses) for socioeconomic determinants, which are associated with individual environmental concerns. Unlike those methodologies adopted in many previous studies, we apply an ordered probit model to test three hypotheses (the age, gender, and social class hypotheses) in this study. As a result, high income and high education level are found to be positively related to environmental concern as expected. However, we find that in contrast to most of the existing studies, the marginal effect of age on the probability of being environmentally concerned is positive in several measures, implying that the older are more concerned about the environment than the younger. In addition, weak evidences indicate that women are less concerned about the environment than men. Other socioeconomic characteristics such as employment status and household size are not significant in most of the environmental concern measures we defined.
    Keywords: Socioeconomic determinants, Environmental Concern, Ordered Probit Model, Chinese
    JEL: C25 C81 Q59
    Date: 2007–04
  11. By: Emily Lancsar (University of Newcastle Upon Tyne)
    Abstract: The use of Stated Preference Discrete Choice Modelling (SPDCM) is gaining currency in the health economics field as a method of eliciting: preferences for goods and services; the rate at which individuals are prepared to trade off different attributes of a good or service; and the willingness to pay for goods and services. The purpose of this paper is to develop welfare measures from SPDCM data that are consistent with microeconomic welfare theory. The theory of welfare measurement using discrete data and links to the more well known literature using continuous data are presented. The estimation of welfare measures obtained from SPDCM and conjoint analysis experiments reported in the health economics literature to date are discussed, focusing on whether commonly adopted measures are consistent with microeconomic welfare theory. Finally, the Hicksian compensating variation is calculated from discrete data collected from a SPDCM experiment designed to elicit patient preferences for preventive asthma medications.
    Keywords: Discrete Choice Modelling, conjoint analysis
    JEL: I11
  12. By: Madeleine King (CHERE, University of Technology, Sydney); Rosalie Viney (CHERE, University of Technology, Sydney); Ishrat Hossain (CHERE, University of Technology, Sydney); David Smith (Cancer Council, NSW); Sandra Fowler (CHERE, University of Technology, Sydney); Elizabeth Savage (CHERE, University of Technology, Sydney); Bruce Armstrong (University of Sydney)
    Abstract: Prostate cancer is the most common cancer in men in Australia; each year over 10,000 Australians are diagnosed with this disease. There are a number of treatment options for early stage prostate cancer (ESPC); radical prostatectomy, external beam radiotherapy, brachytherapy, hormonal therapy and combined therapy. Treatment can cause serious side-effects, including severe sexual and urinary dysfunction, bowel symptoms and fatigue. Furthermore, there is no evidence as yet to demonstrate that any of these treatments confers a survival gain over active surveillance (watchful waiting). While patient preferences should be important determinants in the type of treatment offered, little is known about patients? views of the relative tolerability of side effects and of the survival gains needed to justify these. To investigate this, a discrete choice experiment (DCE) was conducted in a sample of 357 men who had been treated for ESPC and 65 age-matched controls. The sample was stratified by treatment, with approximately equal numbers in each treatment group. The DCE included nine attributes: seven side-effects and two survival attributes (duration and uncertainty). An orthogonal fractional set of 108 scenarios from the full factorial was used to generate three versions of the questionnaire, with 18 scenarios per respondent. Multinomial logit (MNL) and mixed logit (MXL) models were estimated. A random intercept MXL model provided a significantly better fit to the data than the simple MNL model, and adding random coefficients for all attributes dramatically improved model fit. Each side-effect had a statistically significant mean effect on choice, as did survival duration. Most attributes had significant variance parameters, suggesting considerable heterogeneity among respondents in their preferences. To model this heterogeneity, we included men?s health-related quality of life scores following treatment as covariates to see whether their preferences were influenced by their previous treatment experience. This study demonstrate how DCEs can be used to quantify the trade-offs patients make between side-effects and survival gains. The results provide useful insights for clinicians who manage patients with ESPC, highlighting the importance of patient preferences in treatment decisions.
    Keywords: Prostate cancer, discrete choice experiment, preferences, quality of life
    JEL: I10

This nep-dcm issue is ©2007 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.