nep-dcm New Economics Papers
on Discrete Choice Models
Issue of 2010‒10‒30
eleven papers chosen by
Philip Yu
Hong Kong University

  1. Limits on Individual Choice By Eytan Sheshinski
  2. Estimating ordered categorical variables using panel data: a generalized ordered probit model with an autofit procedure By Christian Pfarr; Andreas Schmid; Udo Schneider
  3. Neighbourhood Choice and Neighbourhood Reproduction By Bergström, Lina; van Ham, Maarten; Manley, David
  4. Why Junior Doctors Don’t Want to Become General Practitioners: A Discrete Choice Experiment from the MABEL Longitudinal Study of Doctors By Peter Sivey; Anthony Scott; Anthony Scott; Julia Witt; Catherine Joyce; John Humphreys
  5. Custo de acessibilidade entre residência e trabalho: um enfoque das características individuais, familiares e locais By Admir Antonio Betarelli Junior
  6. Assessing The Effects of Trade Liberalization on Wage Inequalities in Egypt: A Microsimulation Analysis By Rana Hendy; Chahir Zaki
  7. Equilibrium policy simulations with random utility models of labour supply By Ugo Colombino
  8. Off-Season Tourists and the Cultural Offer of a Mass-Tourism Destination: The Case of Rimini By Paolo Figini; Laura Vici
  9. Labour Supply, Work Effort and Contract Choice: Theory and Evidence on Physicians By Fortin, Bernard; Jacquemet, Nicolas; Shearer, Bruce
  10. An Ordered Probit Model Analysis of Transaction Costs and Market Participation by Small-Holder Cassava Farmers in South-Eastern Nigeria. By Okoye, B.C; Onyenweaku, C.E; Ukoha, O.O
  11. Is health status of elderly worsening in India:A comparison of successive rounds of National Sample Survey data By Husain, Zakir; Ghosh, Saswata

  1. By: Eytan Sheshinski
    Abstract: Individuals behave with choice probabilities defined by a multinomial logit (MNL) probability distribution over a finite number of alternatives which includes utilities as parameters. The salient feature of the model is that probabilities depend on the choice-set, or domain. Expanding the choice-set decreases the probabilities of alternatives included in the original set, providing positive probabilities to the added alternatives. The wider probability 'spread' causes some individuals to fur- ther deviate from their higher valued alternatives, while others find the added alternatives highly valuable. For a population with diverse preferences, there ex- ists a subset of alternatives, called the optimum choice-set, which balances these considerations to maximize social welfare. The paper analyses the dependence of the optimum choice-set on a parameter which specifies the precision of individuals' choice ('degree of rationality'). It is proved that for high values of this parame- ter the optimum choice-set includes all alternatives, while for low values it is a singleton. Numerical examples demonstrate that for intermediate values, the size and possible nesting of the optimum choice-sets is complex. Governments have various means (defaults, tax/subsidy) to directly a¤ect choice probabilities. This is modelled by 'probability weight'parameters. The paper analyses the structure of the optimum weights, focusing on the possible exclusion of alternatives. A binary example explores the level of 'type one'and 'type two'errors which justify the imposition of early eligibility for retirement benefits, common to social security systems. Finally, the e¤ects of heterogeneous degrees of rationality among individuals are briefly discussed.
    Date: 2010–06
  2. By: Christian Pfarr; Andreas Schmid; Udo Schneider
    Abstract: Estimation procedures for ordered categories usually assume that the estimated coefficients of independent variables do not vary between the categories (parallel-lines assumption). This view neglects possible heterogeneous effects of some explaining factors. This paper describes the use of an autofit option for identifying variables that meet the parallel-lines assumption when estimating a random effects generalized ordered probit model. We combine the test procedure developed by Richard Williams (gologit2) with the random effects estimation command regoprob by Stefan Boes.
    Keywords: Generalized ordered probit; panel data; autofit, self-assessed health.
    JEL: C23 C25 C87 I10
    Date: 2010–10–23
  3. By: Bergström, Lina (Uppsala University); van Ham, Maarten (University of St. Andrews); Manley, David (University of St. Andrews)
    Abstract: Although we know a lot about why households choose certain dwellings, we know relatively little about the mechanisms behind neighbourhood choice. Most studies of neighbourhood choice only focus on one or two dimensions of neighbourhoods: typically poverty and ethnicity. This paper argues that neighbourhoods have multiple dimensions and that models of neighbourhood choice should take these dimensions into account. We propose the use of a conditional logit model. From this approach we can gain insight into the interaction between individual and neighbourhood characteristics which lead to the choice of a particular neighbourhood over alternative destinations. We use Swedish register data to model neighbourhood choice for all households which moved to a neighbourhood in the city of Uppsala between 1997 and 2006. Our results show that neighbourhood sorting is a highly structured process where households are very likely to choose neighbourhoods where the neighbourhood population matches their own characteristics.
    Keywords: neighbourhood, housing choice, sorting, residential mobility, conditional logit, Sweden
    JEL: I30 J60 R23
    Date: 2010–10
  4. By: Peter Sivey (Melbourne Institute of Applied Economic and Social Research, The University of Melbourne); Anthony Scott (Melbourne Institute of Applied Economic and Social Research, The University of Melbourne); Anthony Scott (Melbourne Institute of Applied Economic and Social Research, The University of Melbourne); Julia Witt (Department of Economics, University of Manitoba); Catherine Joyce (Department of Epidemiology and Preventive Medicine, Monash University); John Humphreys (School of Rural Health, Monash University)
    Abstract: A number of studies suggest there is an over-supply of specialists and an undersupply of GPs in many developed countries. Previous econometric studies of specialty choice from the US suggest that a number of factors play a role, including expected future earnings, educational debt, and having predictable working hours. Given endogeneity issues in revealed preference studies, a stated-preference approach is warranted. This paper presents results from a discrete-choice experiment completed by a sample of 532 junior doctors in 2008 before they choose a specialty training program. This was conducted as part of the first wave of the MABEL (Medicine in Australia: Balancing Employment and Life) longitudinal survey of doctors. We include key job attributes such as future earnings and hours worked, but also allow the choice to be influenced by academic research opportunities, continuity of care and the amount of procedural work. Interactions of attributes with doctor characteristics, including gender, educational debt, and personality traits are also examined. We find the income/working hours trade-offs estimated from our discrete choice model are close to the actual wages of senior specialists, but much higher than those of senior GPs. In a policy simulation we find that increasing GPs’ earnings by $50,000, increasing opportunities for procedural or academic work can increase the number of junior doctors choosing General Practice by between 8 and 16 percentage points, approximately 212 to 376 junior doctors per year. The results can inform policymakers looking to address unbalanced supply of doctors across specialties.
    Keywords: Junior doctors, discrete choice experiment, specialty choice
    JEL: C9 I11 J24
    Date: 2010–10
  5. By: Admir Antonio Betarelli Junior (Cedeplar-UFMG)
    Abstract: The main aim of this paper is to analyze the probability of busy individuals to have a greater or lesser cost of accessibility in the displacement of their residence to the workplace. The global analysis is based on individual, family and local characteristics. To do so, we assign to the cost of accessibility as the average time from home to work and this variable is estimated by maximum likelihood with the Generalized Ordered Logit Model (MLOG). The main results show that: (a) households without children promote a negative effect on the probability of the cost of accessibility; (b) increasing age decreases the chances of individuals have a higher cost of accessibility; and (c) when individuals receive transport help, your chances increases to spend up to 30 minutes of residence to the workplace.
    Keywords: Cost of accessibility; Generalized Ordered Logit Model
    JEL: C25 R40
    Date: 2010–10
  6. By: Rana Hendy (Université Paris 1 Panthéon-Sorbonne & Paris School of Economics); Chahir Zaki
    Abstract: This paper develops a microsimulation analysis to evaluate the impact of trade liberalization policies in Egypt on income redistribution. Our analysis aims at identifying the effects of those measures on redistribution aspects. For this, we rely on a macro - micro approach integrating results obtained from a discrete choice model of labor supply in a Computable General Equilibrium model (CGE). In the empirical work, we use the Egyptian Labor Market and Panel Survey (ELMPS) of 1998 and 2006 as well as the Social Accounting Matrix (SAM) of 2001. This assessment allows us to find out to what extent such macroeconomic policies affect, on the microeconomic level, females poverty, wages and employment opportunities.
    Date: 2010–10
  7. By: Ugo Colombino
    Abstract: Many microeconometric models of discrete labour supply include alternative-specific constants meant to account for (possibly besides other factors) the density or accessibility of particular types of jobs (e.g. parttime jobs vs. full-time jobs). The most common use of these models is the simulation of tax-transfer reforms. The simulation is usually interpreted as a comparative static exercise, i.e. the comparison of different equilibria induced by different policy regimes. The simulation procedure, however, typically keeps fixed the estimated alternative-specific constants. In this note we argue that this procedure is not consistent with the comparative statics interpretation. Equilibrium means that the number of people willing to work on the various job types must be equal to the number of available jobs. Since the constants reflect the number of jobs and since the number of people willing to work change as a response to the change in tax-transfer regime, it follows that the constants should also change. A structural interpretation of the alternative-specific constants leads to the development of a simulation procedure consistent with the comparative static interpretation. The procedure is illustrated with an empirical example.
    Keywords: Random Utility; Discrete Choice; Labour Supply; Simulation of tax reforms; Alternative-specific constants; Equilibrium simulation
    JEL: C35 C53 H31 J22
    Date: 2010
  8. By: Paolo Figini (Department of Economics, University of Bologna, Italy; The Rimini Centre for Economic Analysis (RCEA)); Laura Vici (Department of Economics, University of Bologna, Italy; The Rimini Centre for Economic Analysis (RCEA))
    Abstract: This paper assesses the potential implications on off-season tourism of enhancing the cultural offer of Rimini, a popular Italian seaside holiday destination. Rimini, a city of about 130,000 people hosts a total of around 12 million overnight stays, 10 million of which are concentrated in the summer months. In the last twenty years or so, Rimini has been undergoing a policy of deseasoning, which mainly pivots around business tourism (a new fair quarter and important conference venues have been built) and cultural tourism (the city has been investing on both its cultural heritage and art exhibitions). This assessment is carried out through discrete choice experiments submitted to a sample of about 800 off-season tourists, that is, tourists who visited Rimini outside the summer months. Since tourism can be viewed as a composite good, which overall utility depends on the arrangement of the component characteristics, the choice experiments allow to disentangle the importance and the willingness to pay of tourists for different levels of the holiday's characteristics. The choice model incorporates as attributes a number of possible changes to actual tourism features (which are also the subject of public debate), including them in hypothetical alternative "holiday packages". The conditional logit analysis of the choice experiments can highlight the potential synergies and trade-offs between cultural and business tourism. Moreover, the methodology and the structure of the questionnaire allow a partial comparison of our findings with results stemming from two previous studies carried out in Rimini, respectively on summer tourists and on residents. Such comparison highlights synergies and trade-offs between off-season tourists, summer tourists, and residents.
    Keywords: tourism demand; cultural tourism; business tourism, conditional logit; urban planning; choice experiments
    Date: 2010–01
  9. By: Fortin, Bernard; Jacquemet, Nicolas; Shearer, Bruce
    Abstract: We develop and estimate a generalized labour supply model that incorporates work effort into the standard consumption-leisure trade-off. We allow workers a choice between two contracts: a piece rate contract, wherein he is paid per unit of service provided, and a mixed contract, wherein he receives an hourly wage and a reduced piece rate. This setting gives rise to a nonconvex budget set and an efficient budget constraint (the upper envelope of contract-specific budget sets). We apply our model to data collected on specialist physicians working in the Province of Quebec (Canada). Our data set contains information on each physician’s labour supply and their work effort (clinical services provided per hour worked). It also covers a period of policy reform under which physicians could choose between two compensation systems: the traditional fee-for-service, under which physicians receive a fee for each service provided, and mixed remuneration, under which physicians receive a per diem as well as a reduced fee-for-service. We estimate the model using a discrete choice approach. We use our estimates to simulate elasticities and the effects of ex ante reforms on physician contracts. Our results show that physician services and effort are much more sensitive to contractual changes than is their time spent at work. Our results also suggest that a mandatory reform, forcing all physicians to adopt the mixed remuneration system, would have had substantially larger effects on physician behaviour than those observed under the voluntary reform.
    Keywords: labour supply, effort, contracts, practice patterns of physicians, discrete choice econometric models, mixed logit
    JEL: C25 J22 J33 I10 J44
    Date: 2010–10–21
  10. By: Okoye, B.C; Onyenweaku, C.E; Ukoha, O.O
    Abstract: The Ordered Probit model analysis procedure was applied to determine the factors (related to fixed and variable transaction costs) influencing the decision to participate in cassava markets by a sample of 360 smallholder farmers in South-Eastern Nigeria. Participation decisions revealed that membership of cooperatives or social organizations, farming experience and marketing experience had a positive relationship with decision to be autarkic other than buyer and seller other than autarkic and significant at 1.0% level of probability. The coefficients for frequency of extension contacts, age, native of community, road conditions to the nearest town and yield were also positive and significantly related to decision to be autarkic other than buyer and seller other than autarkic at 5% level of probability. The coefficient for access to communication facilities was positive and significantly related to decision to remain autarkic other than buyer and seller other than autarkic. The coefficients for education, distance to the nearest town, distance from the farm to the market and crop transportation were negative and significantly related with the decision to remain autarkic other than a seller and buyer other than autarkic at 1% level of probability. The coefficient for gender was positive and significantly related to decision by female farmers to be autarkic other than buyer and seller other than autarkic. These decisions to participate as a buyer, seller or remain autarkic were as a result of fixed and proportional transaction costs associated with participating in the market.
    Keywords: Ordered Probit; Transaction Costs; Market Participation and Cassava
    JEL: D23 D21 D2
    Date: 2010–08–18
  11. By: Husain, Zakir; Ghosh, Saswata
    Abstract: The increasing graying of India’s population raises concerns about the welfare and health status of the aged. One important source of information of health status of the elderly is the National Sample Survey Rounds on Morbidity and Health Care Expenditure. Using unit level data for 1995-96 and 2004 this paper has examined changes in reported health status of elderly in India and analyzed their relationship with living arrangements and extent of economic dependency. It appears that even after controlling for factors like caste, education, age, economic status and place of residence, there has been deterioration in self-perceived current health status of elderly. This paper argues that, although there have been changes in the economic condition and traditional living arrangements - with a decline in co-residential arrangements - this is not enough to explain the decline in reported health status and calls for a closer look at narratives of neglect being voiced in developing countries.
    Keywords: economic independence; co-residence; elderly; health status; ordered logit model; India
    JEL: J14 C35
    Date: 2010–09–05

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.