nep-dcm New Economics Papers
on Discrete Choice Models
Issue of 2007‒01‒13
thirteen papers chosen by
Philip Yu
Hong Kong University

  1. Efficient Compromising By Peter Postl
  2. Statistical Properties of Consideration Sets By Richard Carson; Jordan Louviere
  3. Therapeutic non adherence: a rational behaviour revealing patient preferences? By Karine Lamiraud; Pierre-Yves Geoffard
  4. Academic journals as two-sided platforms : empirical evidence from data on french libraries By Dubois, P.; Hernandez Perez, A.; Ivaldi, M.
  5. Dynamic choices of hyperbolic consumers: the continuous time case. By Ivar Ekeland; Lazrak Ali
  6. Designing Optimal Taxes with a Microeconometric Model of Household Labour Supply By Rolf Aaberge and Ugo Colombino
  7. A Framework for Analyzing Rank Ordered Panel Data with Application to Automobile Demand By John K. Dagsvik and Gang Liu
  8. Labor Supply as a Choice among Latent Job Opportunities. A Practical Empirical Approach By John K. Dagsvik and Zhiyang Jia
  9. Estimation of Earnings- and Schooling Choice Relations: A Likelihood Approach By John K. Dagsvik, Torbjørn Hægeland and Arvid Raknerud
  10. Simulating labor supply behavior when workers have preferences for job opportunities and face nonlinear budget constraints By John K. Dagsvik, Marilena Locatelli and Steinar Strøm
  11. "Coping with Damages caused by the Great Hanshin-Awaji Earthquake"(in Japanese) By Yasuyuki Sawada; Satoshi Shimizutani
  12. Utility in WTP Space: A Tool to Address Confounding Random Scale Effects in Destination Choice to the Alps By Ricardo Scarpa; Mara Thiene; Kenneth Train
  13. A Relational Account of the Causes of Spatial Firm Mobility By Joris Knoben

  1. By: Peter Postl
    Abstract: Two agents have to choose one of three alternatives. Their ordinal rankings of these alternatives are commonly known among them. The rankings are diametrically opposed to each other. Ex-ante efficiency requires that they reach a compromise, that is choose the alternative which they both rank sec- ond, if and only if the sum of their von Neumann Morgenstern utilities from this alternative exceeds the sum of utilities when either agent's favorite al- ternative is chosen. We assume that the von Neumann Morgenstern utilities of the middle ranked alternative are independent and identically distributed, privately observed random variables, and ask whether there are incentive compatible decision rules which elicit utilities and implement efficient deci- sions. We show that no such decision rules exist if the distribution of agents' types has a density with full support. We also study the problem of finding second-best decision rules in our set-up, and explain how this problem differs from more familiar second best problems. Finally, we give some numerical insights into the nature of second-best rules. For a variety of distributions of types, second-best rules involve very little inefficiency.
    Keywords: arbitration, mechanism design without transferrable utility
    JEL: C72 D70 D80
  2. By: Richard Carson; Jordan Louviere (School of Marketing, University of Technology, Sydney)
    Abstract: Consideration sets have become a central concept in the study of consumer behavior. Frequently, consumers are asked to split choice alternatives into those that that they would consider and those that they would not. Information on alternatives not in the consideration set is then typically not used in subsequent analysis. This practice is shown to lead to biased estimates of preference parameters. The reason for this is shown to be a form of sample selection bias.
    Keywords: choice models, random utility, sample selection bias,
    Date: 2006–07–01
  3. By: Karine Lamiraud; Pierre-Yves Geoffard
    Abstract: This paper offers an indirect measure of patient welfare based on whether patients comply with the prescription they receive. Adherence behavior is supposed to reveal patients' subjective valuations of particular therapies. We write a simple theoretical model of patient adherence behavior, that reflects the trade-off between perceived costs and observed regimen efficacy. A discrete choice framework is then used for the estimation, ie the comparison of the incremental benefit of drug intake between two regimens. Consequently, the empirical analysis is based on the identification of patient and drug characteristics associated with adherence. The econometric approach is implemented through a bivariate panel two-equation simultaneous system studying jointly the factors associated with adherence and response to treatment. The data come from a randomized clinical trial conducted in France between 1999 and 2001 and comparing the efficacy of 2 tritherapy strategies in HIV disease. Both the theoretical and empirical results suggest that, for comparable clinical efficacy and toxicity levels, a higher adherence level is associated with higher patient welfare, thus adding valuable information to conclusions drawn by a mere biostatistical analysis. Therefore, from the perspective of the patient, the adherence-enhancing drug must be favored. Our results based on panel data also stress that unobserved patient characteristics account substantially for drug valuation and that the assessment evolves during the course of the treatment. Furthermore, we provide a new framework for the analysis of adherence data. The microeconometric framework highlights that non adherence is an endogenous behavior, thus suggesting new ways for improving adherence.
    Date: 2006
  4. By: Dubois, P.; Hernandez Perez, A.; Ivaldi, M.
    Abstract: This paper analyzes the demand and cost structure of the french market of academic journals, taking into account its intermediary role between researchers, who are both producers and consumers of knowledge. This two sidedness feature will echoes similar problems already observed in electronic markets-payment card systems, video games console, etc-such as the chicken and egg problem, where readers won't buy a journal if they do not expect its articles to be academically relevant and researchers, that live under the mantra "publish or perish", will not submit to a journal with either limited public reach or weak reputation. After the merging of several databases, we estimate the aggregated nested logit demand system combined simultaneously with a cost function. We identify the structural parameters of this market and find that price elasticities of demand are quite large and margins relatively low, indicating that this industry experiences competitive constraints. ...French Abstract : Cet article analyse la structure de la demande et des coûts du marché français des revues scientifiques. Nous estimons un "nested logit" pour le modèle de demande et identifions les paramètres structurels de ce marché. Nous trouvons que les élasticités prix de la demande sont assez grandes et les marges relativement faibles ce qui indique que cette industrie est relativement concurrentielle.
    Date: 2006
  5. By: Ivar Ekeland; Lazrak Ali (Sauder school of business university of british columbia)
    Abstract: This paper characterizes the set of differentiable subgame perfect equilibria in a continuous time intertemporal decision optimization problem with non-constant discounting. The idea of an infinitesimal self is formalized and the equilibrium characterization takes the form of an integral equation (IE) which is reminiscent of the Hamilton-Jacobi-Bellman equation. Beginning with a local existence proof of IE, we analyze some equilibria of the consumption saving problem. We then use the equation IE to suggest a critical indeterminacy in the Ramsey growth model with non-constant discounting
    Keywords: Hyperbolic discount, consumption saving decision, growth theory
    JEL: D91
    Date: 2006–12–03
  6. By: Rolf Aaberge and Ugo Colombino (Statistics Norway)
    Abstract: The purpose of this paper is to present an exercise where we identify optimal income tax rules under the constraint of fixed tax revenue. To this end, we estimate a microeconomic model with 78 parameters that capture heterogeneity in consumption-leisure preferences for singles and couples as well as in job opportunities across individuals based on Norwegian household data for 1994. The estimated model is for a given tax rule used to simulate the choices made by single individuals and couples. Those choices are therefore generated by preferences and opportunities that vary across the decision units. Differently from what is common in the literature, we do not rely on a priori theoretical optimal taxation results, but instead we identify optimal tax rules – within a class of 6-parameter piece-wise linear rules - by iteratively running the model until a given social welfare function attains its maximum under the constraint of keeping constant the total net tax revenue. We explore a variety of social welfare functions with differing degree of inequality aversion and also two alternative social welfare principles, namely equality of outcome and equality of opportunity. All the social welfare functions turn out to imply an average tax rate lower than the current 1994 one. Moreover, all the optimal rules imply – with respect to the current rule – lower marginal rates on low and/or average income levels and higher marginal rates on sufficiently high income levels. These results are partially at odds with the tax reforms that took place in many countries during the last decades. While those reforms embodied the idea of lowering average tax rates, the way to implement it has typically consisted in reducing the top marginal rates. Our results instead suggest to lower average tax rates by reducing marginal rates on low and average income levels and increasing marginal rates on very high income levels.
    Keywords: Labour supply; optimal taxation; random utility model; microsimulation
    JEL: H21 H31 J22
    Date: 2006–09
  7. By: John K. Dagsvik and Gang Liu (Statistics Norway)
    Abstract: In this paper we develop a framework for analyzing panel data with observations on rank ordered alternatives that allows for correlated random taste shifters across time and across alternatives. As a special case we obtain a nested logit model type for rank ordered alternatives. We have applied this framework to estimate several model versions for household demand for conventional and alternative fuel automobiles in Shanghai based on rank ordered data obtained from a stated preference survey. The preferred model is then used to calculate demand probabilities and elasticities and the willingness-to-pay for alternative fuel vehicles.
    Keywords: Random utility models; Nested rank ordered logit models; Automobile demand; Alternative fuel vehicles
    JEL: C25 C33 L92
    Date: 2006–10
  8. By: John K. Dagsvik and Zhiyang Jia (Statistics Norway)
    Abstract: In this paper, we discuss aspects of a particular framework for modeling labor supply and the application of this approach in practical policy simulation experiments. This modeling framework differs from the standard models of labor supply in that the notion of job choice is fundamental. Specifically, the worker is assumed to have preferences over a latent worker-specific choice set of jobs from which he or she chooses his or her preferred job. A job is characterized with fixed (job-specific) working hours and other non-pecuniary attributes. As a result, observed hours of work are interpreted as the job-specific (fixed) hours of work that is associated with the chosen job. We then show that our framework is practical with respect to applications in empirical analysis and simulation experiments, and is able to produce satisfactory out-of-sample predictions by estimating the model on Norwegian microdata from 1997 and predicting the corresponding microdata from 2003.
    Keywords: Labor supply; non-pecuniary job attributes; non-convex budget sets; latent choice sets; random utility models.
    JEL: J22 C51
    Date: 2006–10
  9. By: John K. Dagsvik, Torbjørn Hægeland and Arvid Raknerud (Statistics Norway)
    Abstract: In this paper we develop a full information maximum likelihood method for the estimation of a joint model for the choice of length of schooling and the corresponding earnings equation. The model for schooling is assumed to be an ordered probit model, whereas the earnings equation is allowed to be very general with explanatory variables that are flexible transformations of schooling and experience. The coefficients associated with length of schooling and experience are allowed to be random and all the random terms of the model may be correlated. Under normality assumptions, we show that the joint probability distribution for schooling and earnings can be expressed on a closed form that is tractable for empirical analysis.
    Keywords: Schooling choice; earnings equation; treatment effects; self-selection; ordered probit; random coefficients; full information maximum likelihood
    JEL: C31 I20 J30
    Date: 2006–11
  10. By: John K. Dagsvik, Marilena Locatelli and Steinar Strøm (Statistics Norway)
    Abstract: This paper analyzes the properties of a particular sectoral labor supply model developed in Dagsvik and Strøm (2006). The model is estimated on labor supply data for married women in Norway 1994. In this model, workers have preferences over sectors and latent job attributes. Moreover, the model allows for a representation of the individual choice sets of feasible jobs in the economy. The properties of the model are explored by calculating elasticities and through simulations of the effects of particular tax reforms. The overall wage elasticities are rather small, but these small elasticities shadow for much stronger sectoral responses. An overall wage increase and, of course, a wage increase in the private sector only, gives women an incentive to shift their labor supply from the public to the private sector. Marginal tax rates were cut considerably in the 1992 tax reform. We find that the impact on overall labor supply is rather modest, but again these modest changes shadow for stronger sectoral changes. The tax reform stimulated the women to shift their labor from the public to the private sector and to work longer hours. A calculation of mean compensated variation shows that the richest households benefited far more from the 1992 tax reform than did the poorest households.
    Keywords: Labor supply; married females; structural model; sectoral choice; wage elasticities; evaluation of tax reforms
    JEL: J22 C51
    Date: 2006–10
  11. By: Yasuyuki Sawada (Faculty of Economics, University of Tokyo); Satoshi Shimizutani (Institute of Economic Research, Hitotsubashi University)
    Abstract: Applying the ordered probit model to household data, we investigate people's behavior against unexpected losses caused by the Great Hanshin-Awaji (Kobe) earthquake in 1995. We found that credit market played an important role in providing an effective risk-coping measure. In the process of preparing well-designed social safety nets against future natural disasters, there exist two policy implications based on our analyses: first, in its attempt to provide ex post public support in the event of a natural disaster, the government may create a moral hazard problem by encouraging people to expose themselves to greater risks than required. Our empirical results suggest that providing subsidized loans to victims can be a good example of facilitating risk-coping behavior; such interventions are less likely to create serious moral hazard problems. Second, it would be imperative to design ex ante risk-management policies against the earthquake. For example, development of markets for earthquake insurance would lead to the efficient pricing of insurance premium and efficient land market prices reflective of the amount of risk involved with lands. This development would generate proper incentives to invest in mitigations such as investments in earthquake-proof constructions against future earthquakes. These ex ante measures would significantly reduce the overall social loss caused by the earthquake.
    Date: 2005–09
  12. By: Ricardo Scarpa (University of Waikato); Mara Thiene (University of Padua Viale dell’Universita`); Kenneth Train (University of California at Berkeley)
    Abstract: Destination choice models with individual-specific taste variation have become the presumptive analytical approach in applied nonmarket valuation. Continuous mixtures of taste distributions are often modeled using computationally convenient distributions based on the multivariate normal. Though conceptually appealing, empirically these often imply results with untenable distributions of willingness-to-pay in the population. Furthermore, interpersonal variation in the scale of the error may confound variation in taste intensities thereby producing biased WTP estimates. We compare estimates from random utility models that use normal and log-normal distributions first for taste intensities of destination attributes and then for WTPs. Estimates from simulated maximum likelihood and hierarchical Bayes approaches are compared. The results indicate that specifications in WTP space produce more reasonable features of implied WTP distributions for the population. This approach to specification of utility is hence deemed promising in applied nonmarket valuation.
    Keywords: mixed logit random utility parameters; random willingness to pay; travel cost method; destination choice modeling
    JEL: C15 C25 Q26
    Date: 2006–12–15
  13. By: Joris Knoben
    Abstract: It is accepted in the literature that exchanges within networks have an ongoing social structure that both enables and constrains the behavior of its members (Pfeffer and Nowak 1976; Uzzi 1996). However, most research in inter-organizational settings has focused on the enabling effects of networks and network structures only, even though some noteworthy exceptions exist (e.g. Romo and Schwartz 1995; Singh and Mitchell 1996). A possible constraining effect of network participation is spatial lock-in, also known as spatial inertia, of a firm. Following Resource Dependence Theory (Pfeffer and Salancik 1978), it can be argued that a firm that makes extensive use of knowledge resources possessed or controlled by external actors for its innovative processes can become dependent on these actors. By themselves, the relationships in which these dependencies exist are non-spatial. However, since geographical proximity is assumed to facilitate the successful exchange of (especially tacit) knowledge through inter-organizational relationships (IORs) (Bretschger 1999), dependency on other firms located in the same region can also lead to dependency on a certain geographical location, and thus to spatial lock-in (Stam 2003). The IORs that are enabling for the firm in terms of its innovative processes act, at the same time, as constraining factors for the spatial behavior of the firm. Similar reasonings can be found in the literature on Territorial Innovation Models (Moulaert and Sekia 2003), which indicates that economic embeddedness in a region can be beneficial for the performance of firms. However, this embeddedness can also lead to dependence on localized inputs and production factors. Due to these dependencies, a firm can become very unlikely to relocate, even if doing so is beneficial from a cost perspective. As Romo and Schwartz state: “Firms are usually too dependent on the material, political and social resources available in the local production culture to risk departure, even when production costs might be substantially reduced (Romo and Schwartz 1995:874).†There currently is, however, only weak empirical evidence for the proposed relationship between the level of (local) embeddedness and a firm’s propensity to relocate. Moreover, several authors even propose that geographical distance in IORs is becoming irrelevant since it effects can be replicated by ICT (Morgan 2004), or high levels of organizational or technological proximity (Kirat and Lung 1999). If this is indeed the case, then participation in localized innovative IORs will have no effect on the spatial behavior of firms, since a firm can operate exactly the same on a different geographical location. The main goal of this research is to provide empirical insights into the effects of a firm’s level of participation in innovative (localized) inter-organizational relationships (IORs) on its propensity to relocate. Based on the above, the following research question has been formulated is “To what extent is the level of embeddedness of a firm in (localized) innovative inter-organizational relationships of influence on its propensity to relocate?†Answering this research question adds to the insights about the constraining effects of networks by focusing on the spatially constraining effect of inter-organizational relationships. This research question will be answered based on a data from a survey among Dutch automation service firms in 2006. In line with earlier research (c.f. Van Dijk and Pellenbarg 2000; Brouwer et al. 2004) an ordinal logit model will be used to relate the relocation propensity of a firm to that firm’s participation in localized innovative IORs, the strength of these IORs, and the level of geographical, organizational and technological proximity. It also provides insight into the question whether or not high levels of technological and organizational proximity can negate the need for geographical proximity in inter-organizational collaboration (Boschma 2005). References: Boschma, R. A. (2005). "Proximity and innovation: A critical assessment." Regional Studies 39(1): 61-74 Bretschger, L. (1999). "Knowledge diffusion and the development of regions." Annals of Regional Science 33(3): 251-268 Brouwer, A. E., I. Mariotti and J. N. van Ommeren (2004). "The firm relocation decision: An empirical investigation." Annals of Regional Science 38(2): 335-347 Van Dijk, J. and P. H. Pellenbarg (2000). "Firm relocation decisions in The Netherlands: An ordered logit approach." Papers in Regional Science 79(1): 191-219 Kirat, T. and Y. Lung (1999). "Innovation and proximity - Territories as loci of collective learning processes." European Urban and Regional Studies 6(1): 27-38 Morgan, K. (2004). "The exaggerated death of geography: Learning, proximity and territorial innovation systems." Journal of Economic Geography 89(1): 3-21 Moulaert, F. and F. Sekia (2003). "Territorial innovation models: A critical review." Regional Studies 37(3): 289-302 Pfeffer, J. and P. Nowak (1976). "Joint-ventures and interorganizational interdependence." Administrative Science Quarterly 21(3): 398-418 Pfeffer, J. and G. R. Salancik (1978). The external control of organizations: A resource dependency perspective. New York, Harper and Row Romo, F. P. and M. Schwartz (1995). "The structural embeddedness of business decisions: The migration of manufacturing plants in New York state, 1960 to 1985." American Sociological Review 60(1): 874-907 Singh, K. and W. Mitchell (1996). "Precarious collaboration: Business survival after partners shut down or form new partnerships." Strategic management journal 17(2): 99-116 Stam, F. C. (2003). Why butterflies don't leave: Locational evolution of evolving enterprises. Utrecht, Utrecht University Uzzi, B. (1996). "The sources and consequences of embeddedness for the economic performance of organizations: The network effect." American Sociological Review 61(4): 674-698
    Date: 2006–08

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.