Search

The effect of corruption on airport productive efficiency is analyzed using an unbalanced panel data of major European airports from 2003 to 2009. We first compute the residual (or net) variable factor productivity using the multilateral index number method and then apply robust cluster random effects model in order to evaluate the importance of corruption. We find strong evidence that corruption has negative impacts on airport operating efficiency; and the effects depend on the ownership form of the airport. The results suggest that airports under mixed public–private ownership with private majority achieve lower levels of efficiency when located in more corrupt countries. They even operate less efficiently than fully and/or majority government owned airports in high corruption environment. We control for economic regulation, competition level and other airports’ characteristics. Our empirical results survive several robustness checks including different control variables, three alternative corruption measures: International Country Risk Guide (ICRG) corruption index, Corruption Perception Index (CPI) and Control of Corruption Index (CCI). The empirical findings have important policy implications for management and ownership structuring of airports operating in countries that suffer from higher levels of corruption.

In this paper, we present a model that can be viewed as an extension of the traditional Tobit model. As opposed to that specific model, ours also accounts for the the fixed costs of car ownership. That extension is needed since being carless is an option for many households in societies that have a good system of public transportation, the main reason being that carless households wish to save the fixed costs of car ownership. So far, no existing model can adequately map the impact of these fixed costs on car ownership. The Multiple Discrete-Continuous Extreme Value Model (MDCEV) with fixed costs fills this gap. In fact, this model can evaluate the effect of policies intended to influence household behaviour with respect to car ownership, which can be of great interest to policy makers. Our model makes it possible to compute the effect of policies such as taxes on fuel or on car ownership on both the share of carless households and the average driving distance. We calibrated the model using data on Swiss private households in order to forecast were then able to forecast responses to policies. One result of particular interest that cannot be produced by other models is the evaluation of the impact of a tax on car ownership. Our results show that a tax on car ownership has a much lower impact on aggregate driving demand – per unit of tax revenues – than a tax on fuel.

This paper presents a methodological disaggregated approach to analyze the impact of interventions on road safety. The model aims to describe the accident rates of an individual using mileage as a measure of risk exposure. The model is formulated as a system of equations that takes into account interactions between the mileage of a given individual and the other drivers. Once estimated, the model acts as a simulator allowing us to measure the performance of policy interventions to increase road safety.

In this paper we discuss the specification, covariance structure, estimation, identification, and point-estimate analysis of a logit model with endogenous latent attributes that avoids problems of inconsistency. We show first that the total error term induced by the stochastic latent attributes is heteroskedastic and nonindependent. In addition, we show that the exact identification conditions support the two-stage analysis found in much current work. Second, we set up a Monte Carlo experiment where we compare the finite-sample performance of the point estimates of two alternative methods of estimation, namely frequentist full information maximum simulated likelihood and Bayesian Metropolis Hastings-within-Gibbs sampling. The Monte Carlo study represents a virtual case of travel mode choice. Even though the two estimation methods we analyze are based on different philosophies, both the frequentist and Bayesian methods provide estimators that are asymptotically equivalent. Our results show that both estimators are feasible and offer comparable results with a large enough sample size. However, the Bayesian point estimates outperform maximum likelihood in terms of accuracy, statistical significance, and efficiency when the sample size is low.

In this article we develop, implement and apply a Markov chain Monte Carlo (MCMC) Gibbs sampler for Bayesian estimation of a hybrid choice model (HCM), using stated data on both vehicle purchase decisions and environmental concerns. Our study has two main contributions. The first is the feasibility of the Bayesian estimator we derive. Whereas classical estimation of HCMs is fairly complex, we show that the Bayesian approach for HCMs is methodologically easier to implement than simulated maximum likelihood because the inclusion of latent variables translates into adding independent ordinary regressions; we also find that, using the Bayesian estimates, forecasting and deriving confidence intervals for willingness to pay measures is straightforward. The second is the capacity of HCMs to adapt to practical situations. Our empirical results coincide with a priori expectations, namely that environmentally-conscious consumers are willing to pay more for low-emission vehicles. The model outperforms standard discr...

We study the dynamic behaviour of household electricity consumption on the basis of four large independent surveys conducted in the province of Quebec from 1989 to 2002. The latter region displays some rather unique features such as the very extensive use of electricity for space heating in a cold climate and the wide range of energy sources used to meet space heating requirements. We adopt Deaton (1985) approach to create 25 cohorts of households that form a pseudo-panel. The cohorts have on average 131 households. The model error terms allow for group heteroskedasticity and serial correlation. Short-run and long-run own and cross-price elasticities are statistically significant. Electricity and natural gas are estimated to be substitutes while electricity and fuel oil are complements, as it may occur in the Quebec context. The estimate of the income elasticity is not significant. Comparisons with related studies are provided.

While transport modelers in developed countries are accustomed to working with relatively rich datasets including transport networks and land use data, such databases are rarely available in developing countries. However, developing countries such as China with its immense rate of economic growth are, arguably, most in need of demand models. The research addressed in this paper is how to develop mode choice models for planning and policy analysis when level of service data are not available and resources are limited. The research makes use of a 1,001 household travel and activity survey from Chengdu collected in 2005. Chengdu has an urban population of over 3 million and a GDP growth rate of over 20% per year. By coding transportation networks, course estimates of level of service by mode are first developed. As these measures are assumed to have accuracy issues, particularly for transit, level of service is treated as a latent (i.e., unobservable) variable. Measurement equations (from the structu...

We study the problem of building confidence sets for ratios of parameters, from an identification robust perspective. In particular, we address the simultaneous confidence set estimation of a finite number of ratios. Results apply to a wide class of models suitable for estimation by consistent asymptotically normal procedures. Conventional methods (e.g. the delta method) derived by excluding the parameter discontinuity regions entailed by the ratio functions and which typically yield bounded confidence limits, break down even if the sample size is large ( Dufour, 1997 ). One solution to this problem, which we take in this paper, is to use variants of Fieller ’s ( 1940 , 1954 ) method. By inverting a joint test that does not require identifying the ratios, Fieller-based confidence regions are formed for the full set of ratios. Simultaneous confidence sets for individual ratios are then derived by applying projection techniques, which allow for possibly unbounded outcomes. In this paper, we provide simple explicit closed-form analytical solutions for projection-based simultaneous confidence sets, in the case of linear transformations of ratios. Our solution further provides a formal proof for the expressions in Zerbe et al. (1982) pertaining to individual ratios. We apply the geometry of quadrics as introduced by Dufour and Taamouti, 2005 , Dufour and Taamouti, 2007 , in a different although related context. The confidence sets so obtained are exact if the inverted test statistic admits a tractable exact distribution, for instance in the normal linear regression context. The proposed procedures are applied and assessed via illustrative Monte Carlo and empirical examples, with a focus on discrete choice models estimated by exact or simulation-based maximum likelihood. Our results underscore the superiority of Fieller-based methods.

The area of discrete choice modelling has developed rapidly in recent years. In particular, continuing refinements of the Generalised Extreme Value (GEV) model family have permitted the representation of increasingly complex patterns of substitution and parallel advances in estimation capability have led to the increased use of model forms requiring simulation in estimation and application. One model form especially, namely the Mixed Multinomial Logit (MMNL) model, is being used ever more widely. Aside from allowing for random variations in tastes across decision-makers in a Random Coefficients Logit (RCL) framework, this model additionally allows for the representation of inter-alternative correlation as well as heteroscedasticity in an Error Components Logit (ECL) framework, enabling the model to approximate any Random Utility model arbitrarily closely. While the various developments discussed above have led to gradual gains in modelling flexibility, little effort has gone into the development of model forms allowing for a representation of heterogeneity across respondents in the correlation structure in place between alternatives. Such correlation heterogeneity is however possibly a crucial factor in the variation of choice-making behaviour across decision-makers, given the potential presence of individual-specific terms in the unobserved part of utility of multiple alternatives. To the authors' knowledge, there has so far only been one application of a model allowing for such heterogeneity, by Bhat (1997). In this Covariance NL model, the logsum parameters themselves are a function of socio-demographic attributes of the decision-makers, such that the correlation heterogeneity is explained with the help of these attributes. While the results by Bhat show the presence of statistically significant levels of covariance heterogeneity, the improvements in terms of model performance are almost negligible. While it is possible to interpret this as a lack of covariance heterogeneity in the data, another explanation is possible. It is clearly imaginable that a major part of the covariance heterogeneity cannot be explained in a deterministic fashion, either due to data limitations, or because of the presence of actual random variation, in a situation analogous to the case of random taste heterogeneity that cannot be explained in a deterministic fashion. In this paper, we propose two different ways of modelling such random variations in the correlation structure across individuals. The first approach is based on the use of an underlying GEV structure, while the second approach consists of an extension of the ECL model. In the former approach, the choice probabilities are given by integration of underlying GEV choice probabilities, such as Nested Logit, over the assumed distribution of the structural parameters. In the most basic specification, the structural parameters are specified as simple random variables, where appropriate choices of statistical distributions and/or mathematical transforms guarantee that the resulting structural parameters fall into the permissible range of values. Several extensions are then discussed in the paper that allow for a mixture of random and deterministic variations in the correlation structure. In an ECL model, correlation across alternatives is introduced with the help of normally distributed error-terms with a mean of zero that are shared by alternatives that are closer substitutes for each other, with the extent of correlation being determined by the estimates of the standard deviations of the error-components. The extension of this model to a structure allowing for random covariance heterogeneity is again divided into two parts. In the first approach, correlation is assumed to vary purely randomly; this is obtained through simple integration over the distribution of the standard deviations of the error-terms, superseding the integration over the distribution of the error-components with a specific draw for the standard deviations. The second extension is similar to the one used in the GEV case, with the standard deviations being composed of a deterministic term and a random term, either as a pure deviation, or in the form of random coefficients in the parameterisation of the distribution of the standard deviations. We next show that our Covariance GEV (CGEV) model generalises all existing GEV model structures, while the Covariance ECL (CECL) model can theoretically approximate all RUM models arbitrarily closely. Although this also means that the CECL model can closely replicate the behaviour of the CGEV model, there are some differences between the two models, which can be related to the differences in the underlying error-structure of the base models (GEV vs ECL). The CECL model has the advantage of implicitly allowing for heteroscedasticity, although this is also possible with the CGEV model, by adding appropriate error-components, leading to an EC-CGEV model. In terms of estimation, the CECL model has a run-time advantage for basic nesting structures, when the number of error-components, and hence dimensions of integration, is low enough not to counter-act the gains made by being based on a more straightforward integrand (MNL vs advanced GEV). However, in more complicated structures, this advantage disappears, in a situation that is analogous to the case of Mixed GEV models compared to ECL models. A final disadvantage of the CECL model structure comes in the form of an additional set of identification conditions. The paper presents applications of these model structures to both cross-sectional and panel datasets from the field of travel behaviour analysis. The applications illustrate the gains in model performance that can be obtained with our proposed structures when compared to models governed by a homogeneous covariance structure assumption. As expected, the gains in performance are more important in the case of data with repeated observations for the same individual, where the notion of individual-specific substitution patterns applies more directly. The applications also confirm the slight differences between the CGEV and CECL models discussed above. The paper concludes with a discussion of how the two structures can be extended to allow for random taste heterogeneity. The resulting models thus allow for random variations in choice behaviour both in the evaluation of measured attributes C as well as the correlation across alternatives in the unobserved utility terms. This further increases the flexibility of the two model structures, and their potential for analysing complex behaviour in transport and other areas of research.

We develop an econometric model to estimate the impact of Electronic Vehicle Management Systems (EVMS) on the load factor (LF) of heavy trucks. This technology is supposed to improve capacity utilization. The model is estimated on the Quebec subsample of the 1999 National Roadside Survey. The LF is explained as a function of truck, trip, and carrier characteristics. We show that the use of EVMS results in an increase of 16 percentage points of LF on backhaul trips. However, we also find that there is a rebound effect on fronthaul movements, with a reduction of LF by about 7.6 percentage points. Nous estimons un modele econometrique pour evaluer l'impact des systemes de gestion electronique des vehicules (SGEV) sur le taux de chargement (TC) des camions lourds. Cette technologie est censee ameliorer l'utilisation de la capacite. Le modele est estime sur le sous-echantillon quebecois des donnees de l'enquete nationale routiere en bord de route de 1999. Le TC est explique en fonction des caracteristiques du camion, du voyage et de l'entreprise de transport. Nous montrons que l'utilisation de SGEV accroit le TC sur le retour d'environ 16 points de pourcentage. Par contre, nous trouvons egalement un effet rebond sur l'aller avec une reduction de TC d'environ 7.6 points de pourcentage.

The formulation of hybrid discrete choice models, including both observable alternative attributes and latent variables associated with attitudes and perceptions, has become a topic of discussion once more. To estimate models integrating both kinds of variables, two methods have been proposed: the sequential approach, in which the latent variables are built before their integration with the traditional explanatory variables in the choice model and the simultaneous approach, in which both processes are done together, albeit with a sophisticated but fairly complex treatment. Here both approaches are applied to estimate hybrid choice models by using two data sets: one from the Santiago Panel (an urban mode choice context with many alternatives) and another consisting of synthetic data. Differences between both approaches were found as well as similarities not found in earlier studies. Even when both approaches result in unbiased estimators, problems arise when valuations are obtained such as the value of time for forecasting and policy evaluation.

Dans cette étude, nous proposons une généralisation de la formulation à composantes d’erreurs qui permet de représenter différents effets explicatifs de la présence de corrélation dans les erreurs de modèles de régression avec données de flux. Selon la formulation proposée, le terme d’erreur se décompose en une somme d’une erreur relative à la zone d’origine, une erreur relative à la zone de destination et une erreur associée au flux. Chaque composante d’erreur est issue d’un processus générateur auto-régressif spatial d’ordre 1. L’estimation des paramètres du modèle est basée sur la méthode du maximum de vraisemblance. La méthodologie proposée a l’avantage de demeurer applicable même dans le contexte d’échantillons de grande taille., In this study, we propose a generalization of the error component formulation to model the correlation among the errors of a regression based on travel flow data. The error term is broken down into a sum of one component related to the origin zone, one component related to the destination zone and a remainder. Each component is assumed to result from a first-order spatial autoregressive generating process. An efficient estimation approach based on maximum likelihood which addresses the practical implementation of such a model with a large sample size is suggested.

Dans le présent article, nous décrivons une méthodologie générale à information complète pour analyser la dangerosité des sites routiers. La technique proposée, de type bayésienne, permet de traiter simultanément les problèmes d’hétérogénéité déterministe et aléatoire ainsi que celui de la corrélation spatiale attribuable à la proximité ou l’environnement similaire caractérisant les sites à l’étude. Notre cadre méthodologique englobe des approches bayésiennes de pratique courante qui mettent l’accent sur l’analyse des fréquences d’accidents et d’autres du même type qui étudient les proportions d’accidents impliquant une caractéristique donnée. Les propriétés et l’intérêt de la nouvelle méthode sont démontrés à l’aide d’un exemple concret basé sur des données de la région de Québec., In this paper, we describe a general full information Bayesian methodology to analyze road accident sites. The technique allows for the presence of deterministic and random heterogeneity together with spatial autocorrelation among neighboring sites. The suggested framework contains as subcases the Bayesian approaches currently used to study accidents frequencies and those intended for the analysis of accidents proportions of accidents with a given characteristic. To demonstrate the feasibility and the usefulness of the suggested approach, we apply it on accidents data taken from the Quebec city database.

In the past decade, a new trend in discrete choice modeling has emerged: psychological factors are explicitly incorporated to enhance the behavioral representation of the choice process. In this context, hybrid models expand on standard choice models by including attitudes and perceptions as latent variables. The complete model is composed of a group of structural equations describing the latent variables in terms of observable exogenous variables and a group of measurement relationships linking latent variables to certain observable indicators. Although the estimation of hybrid models requires the evaluation of complex multidimensional integrals, simulated maximum likelihood is implemented to solve the integrated multiple-equation model. This study empirically evaluates the application of hybrid choice modeling to data from a survey conducted by the Energy and Materials Research Group (Simon Fraser University, 2002 and 2003) of the virtual personal vehicle choices made by Canadian consumers when they are faced with technological innovations. The survey also includes a complete list of indicators that allows the application of a hybrid choice model formulation. It is concluded that the hybrid choice model is genuinely capable of adapting to practical situations by including latent variables among the set of explanatory variables. The incorporation of perceptions and attitudes in this way leads to more realistic models and gives a better description of the profile of consumers and their adoption of new automobile technologies.

Although the basic structure of logit-mixture models is well understood, important identification and normalization issues often get overlooked. This paper addresses issues related to the identification of parameters in logit-mixture models containing normally distributed error components associated with alternatives or nests of alternatives (normal error component logit mixture, or NECLM, models). NECLM models include special cases such as unrestricted, fixed covariance matrices; alternative-specific variances; nesting and cross-nesting structures; and some applications to panel data. A general framework is presented for determining which parameters are identified as well as what normalization to impose when specifying NECLM models. It is generally necessary to specify and estimate NECLM models at the levels, or structural, form. This precludes working with utility differences, which would otherwise greatly simplify the identification and normalization process. Our results show that identification is not always intuitive; for example, normalization issues present in logit-mixture models are not present in analogous probit models. To identify and properly normalize the NECLM, we introduce the ‘equality condition’, an addition to the standard order and rank conditions. The identifying conditions are worked through for a number of special cases, and our findings are demonstrated with empirical examples using both synthetic and real data. Copyright © 2007 John Wiley & Sons, Ltd.

We extend choice models where the deterministic part of the utility is subject to a deformation filter that affects how the choice probabilities react to changes in explanatory variables. We focus on extending the multinomial logit. Other generalizations are considered.

The capital costs of natural gas distribution pipelines account for 70% of the rate base of SCGM, the main local distribution utility in the province of Quebec. In allocating these capital costs to consumer groups, the regulatory process divides them into an access fee which is supposed to represent the costs of making available natural gas, and a per unit fee reflecting the costs of the capacity which they use. We estimate a cost function to probe whether distribution capital costs can be broken down into these two parts by applying regression analysis to a unique set of 131 observations related to natural gas extension projects in four Quebec regions. It is found that capital costs are not separable into linearly additive constant and variable components, that the cost elasticity with respect to maximum daily demand is small, and that the cost elasticity with respect to pipe length is slightly less than one. On the basis of these findings we conclude that the regulatory process assigns too large a share of distribution capital costs to use relative to access fee.

We discuss the development of predictive choice models that go beyond the random utility model in its narrowest formulation. Such approaches incorporate several elements of cognitive process that have been identified as important to the choice process, including strong dependence on history and context, perception formation, and latent constraints. A flexible and practical hybrid choice model is presented that integrates many types of discrete choice modeling methods, draws on different types of data, and allows for flexible disturbances and explicit modeling of latent psychological explanatory variables, heterogeneity, and latent segmentation. Both progress and challenges related to the development of the hybrid choice model are presented.

The Multinomial Probit (MNP) formulation provides a very general framework to allow for inter-dependent alternatives in discrete choice analysis. Up until recently, its use was rather limited, mainly because of the computational difficulties associated with the evaluation of the choice probabilities which are multidimensional normal integrals. In recent years, the econometric estimation of Multinomial Probit models has greatly been focused on. Alternative simulation based approaches have been suggested and compared. Most approaches exploit a conventional estimation technique where easy to compute simulators replace the choice probabilities. For situations such as in transportation demand modelling where samples and choice sets are large, the existing literature clearly suggests the use of a maximum simulated likelihood (MSL) framework combined with a Geweke–Hajivassiliou–Keane (GHK) choice probability simulator. The present paper gives the computational details regarding the implementation of this practical estimation approach where the scores are computed analytically. This represents a contribution of the paper, because usually, numerical derivatives are used. The approach is tested on a 9-mode transportation choice model estimated with disaggregate data from Santiago, Chile.

We review the case against the standard model of rational behavior and discuss the consequences of various ‘anomalies’ of preference elicitation. A general theoretical framework that attempts to disentangle the various psychological elements in the decision-making process is presented. We then present a rigorous and general methodology to model the theoretical framework, explicitly incorporating psychological factors and their influences on choices. This theme has long been deemed necessary by behavioral researchers, but is often ignored in demand models. The methodology requires the estimation of an integrated multi-equation model consisting of a discrete choice model and the latent variable model system. We conclude with a research agenda to bring the theoretical framework into fruition.

The paper compares the empirical performance of two recently suggested techniques for estimating Multinomial Probit (MNP) models. The application concerns the choice of the first practice location of general practitioners in Quebec (Canada). Regional similarities are accounted for by modeling interdependent choice decisions. One technique is a simulated maximum likelihood based approach that relies on a Geweke, Hajivassiliou, and Keane (GHK) choice probability simulator, and the other one exploits the Gibbs sampler with data augmentation. The results indicate that both estimation techniques give similar results. Compared to its competitor, the Gibbs approach is much simpler to implement both conceptually and computationally.

This paper introduces new forms, sampling and estimation approaches fordiscrete choice models. The new models include behavioral specifications oflatent class choice models, multinomial probit, hybrid logit, andnon-parametric methods. Recent contributions also include new specializedchoice based sample designs that permit greater efficiency in datacollection. Finally, the paper describes recent developments in the use ofsimulation methods for model estimation. These developments are designed toallow the applications of discrete choice models to a wider variety ofdiscrete choice problems.

This paper considers the problem of estimating a nonlinear statistical model subject to stochastic linear constraints among unknown parameters. These constraints represent prior information which originates from a previous estimation of the same model using an alternative database. One feature of this specification allows for the disign matrix of stochastic linear restrictions to be estimated. The mixed regression technique and the maximum likelihood approach are used to derive the estimator for both the model coefficients and the unknown elements of this design matrix. The proposed estimator whose asymptotic properties are studied, contains as a special case the conventional mixed regression estimator based on a fixed design matrix. A new test of compatibility between prior and sample information is also introduced. Thesuggested estimator is tested empirically with both simulated and actual marketing data.

In this article, the authors estimate a spatial autoregressive multinomial probit model of the choice of location by general practitioners for establishing their initial practice. This model allows them to account for potential interdependencies among location choices. The model is used to assess the effect of various incentive measures introduced in Quebec to influence the geographical distribution of physicians across eighteen regions. The authors' data set covers subperiods before and after the introduction of these measures. Incentive policies are captured through price and income effects. The authors' results provide evidence that these measures had a significant effect on location choices. Copyright 1996 by University of Chicago Press.

In Japan, demands for broadband Internet access and IP telephony have increased dramatically in recent years. According to official sources, as of September 2009, there are 30.9 million users of broadband Internet access and 20.9 million of IP telephony. This study evaluates and estimates the market share of fiber-optic Internet connection, which is becoming the major player in broadband services, paying specific attention to IP telephony. A comprehensive combined choice model of Internet access line, IP telephony, and awareness of IP telephony is presented. The most suitable parameters for the model were determined by using an original market research survey conducted in Japan during 2004 with stated-preference choice experiments of both Internet access and IP telephony. The results indicate that increasing awareness has the potential to dramatically increase the penetration of IP telephony. The results also indicate a great variability in price sensitivity across income groups for the choices of Internet access line and IP telephony. The fiber optic share is shown to change with a change in its own monthly usage charge, indicating that market share gains are possible through reduced usage fees.

The idea of transferability is to employ in model estimation, fitted model parameters computed from a different data set. Thecombined estimator approach to the transferability problem is expressed as a linear combination of the unbiased direct estimators on the two data sets. The major gain is in variance reduction. The combined estimator is shown to have superior accuracy, in a Mean Square Error sense, to a unbiased direct estimator whenever the transfer bias is relatively small. A test that indicates if the combined estimator is superior to the direct estimator is provided. Variances of the direct estimators are assumed to be known. Monte Carlo experiments are performed to assess the quality of the approximations. The results show that the approximations used are highly conservative. An empirical example of the combined estimator applied to a discrete choice problem is presented.

In recent years, major advances have taken place in three areas of random utility modeling: (1) semiparametric estimation, (2) computational methods for multinomial probit models, and (3) computational methods for Bayesian estimation. This paper summarizes these developments and discusses their implications for practice.

The search for flexible models has led the simple multinomial logit model to evolve into the powerful but computationally very demanding mixed multinomial logit (MMNL) model. That flexibility search lead to discrete choice hybrid choice models (HCMs) formulations that explicitly incorporate psychological factors affecting decision making in order to enhance the behavioral representation of the choice process. It expands on standard choice models by including attitudes, opinions, and perceptions as psychometric latent variables. In this paper we describe the classical estimation technique for a simulated maximum likelihood (SML) solution of the HCM. To show its feasibility, we apply it to data of stated personal vehicle choices made by Canadian consumers when faced with technological innovations. We then go beyond classical methods, and estimate the HCM using a hierarchical Bayesian approach that exploits HCM Gibbs sampling considering both a probit and a MMNL discrete choice kernel. We then carry out a Monte Carlo experiment to test how the HCM Gibbs sampler works in practice. To our knowledge, this is the first practical application of HCM Bayesian estimation. We show that although HCM joint estimation requires the evaluation of complex multi-dimensional integrals, SML can be successfully implemented. The HCM framework not only proves to be capable of introducing latent variables, but also makes it possible to tackle the problem of measurement errors in variables in a very natural way. We also show that working with Bayesian methods has the potential to break down the complexity of classical estimation.

An empirical study has been developed for the Greek Aegean island area. Data were collected from 900 HHs in Greece contacted via telephone. The HHs were presented hypothetical scenarios involving policy variables, where 2010 was the reference year. ICLV binary logit (BL) and mixed binary logit (MBL) relocation choice models were estimated sequentially. Findings suggest that MBL models are superior to BL models, while both the policy and the latent variables significantly affect the relocation decision and improve considerably the models' goodness of fit. Sample enumeration method is finally used to aggregate the results over the Greek population.

In this study, we propose a generalization of the error components formulation to model the correlation among the errors of a regression based on travel flow data. The error term is broken down into a sum of one component related to the origin zones, one component related to the destination zones and a remainder. The inter-dependences among the errors are assumed to result from applying a first-order spatial autoregressive generating process to each component. An efficient estimation approach based on maximum likelihood is suggested to address the practical implementation of such a model with a large sample size.

In discrete choice analysis, the multinational probit (MNP) provides the most flexible framework to allow for general interdependencies among the alternatives. These interdependencies are usually modeled through the correlation structure of the error term. This framework suffers from two serious impediments, however. The first and major one is computational and is related to the evaluation of the response probabilities, which are multidimensional normal integrals. In the past, this has restricted its utilization to studies involving less than five alternatives where using numerical integration remains practical. A recent solution to the dimensionality problem consists in replacing the choice probabilities with easy to compute efficient simulators. The second impediment arises in models with large choice sets when a fully unconstrained error correlation structure is postulated. In that case, the large number of nuisance parameters to estimate in the error covariance matrix becomes a problematic issue that can exacerbate the estimation process. To tackle that problem, a first-order generalized autoregressive [GAR(1)] error approach is suggested. The approach enables one to approximate general correlation structures with parsimonious parametric specifications. The key feature of the approach is that the number of nuisance parameters grows linearly with the number of alternatives considered. The methodology is most useful in models with large choice sets where the estimation also requires to use probability simulators. The paper focuses on the GAR(1) solution to the error covariance matrix estimation problem. The issue of identification of the nuisance parameters is examined in detail and a rank condition is suggested. Some theoretical and numerical examples based on synthetic data are presented.

In this paper, we develop an econometric model to estimate the impacts of Electronic Vehicle Management Systems (EVMS) on the load factor (LF) of heavy trucks using data at the operational level. This technology is supposed to improve capacity utilization by reducing coordination costs between demand and supply. The model is estimated on a subsample of the 1999 National Roadside Survey, covering heavy trucks travelling in the province of Quebec. The LF is explained as a function of truck, trip and carrier characteristics. We show that the use of EVMS results in a 16 percentage points increase of LF on backhaul trips. However, we also find that the LF of equipped trucks is reduced by about 7.6 percentage points on fronthaul movements. This last effect could be explained by a rebound effect: higher expected LF on the returns lead carriers to accept shipments with lower fronthaul LF. Overall, we find that this technology has increased the tonne-kilometers transported of equipped trucks by 6.3% and their fuel efficiency by 5%.

We derive a general ordinal probit model by specifying the break points or bounds as random functions of explanatory variables. The proposed technique makes it possible to reduce classification errors caused by imperfections in the data collection process. The standard model with constant break points or bounds is expressed as a special case of the proposed model.

The problem of constructing confidence set estimates for parameter ratios arises in a variety of econometrics contexts; these include value-of-time estimation in transportation research and inference on elasticities given several model specifications. Even when the model under consideration is identifiable, parameter ratios involve a possibly discontinuous parameter transformation that becomes ill-behaved as the denominator parameter approaches zero. More precisely, the parameter ratio is not identified over the whole parameter space: it is locally almost unidentified or (equivalently) weakly identified over a subset of the parameter space. It is well known that such situations can strongly affect the distributions of estimators and test statistics, leading to the failure of standard asymptotic approximations, as shown by Dufour. Here, we provide explicit solutions for projection-based simultaneous confidence sets for ratios of parameters when the joint confidence set is obtained through a generalized Fieller approach. A simulation study for a ratio of slope parameters in a simple binary probit model shows that the coverage rate of the Fieller's confidence interval is immune to weak identification whereas the confidence interval based on the delta-method performs poorly, even when the sample size is large. The procedures are examined in illustrative empirical models, with a focus on choice models

In this paper, we applied the generalized mixed estimation approach to the problem of estimating the Quebec residential electricity demand for space and water heating. A multinomial discrete-continuous choice model is used and estimated in two stages. The discrete choice is modelled as a multinomial probit model, while the continuous choice is estimated from a reduced form approach which corrects for the simultaneity biases. The results indicate that the GM estimator which combines prior and sample information dominates the classical ML estimator of the MNP models and hence, provides better prevision for electricity consumption. Evidence also shows that heating-system capital and operating costs, households characteristics, and energy prices have a significant impact on the choice of heating systems and electricity use. In particular, price substitution effects are well predicted.

An important aspect in road safety research concerns the development of analytical tools to identify road sites with high risk. Within a context of optimization subject to financial constraints, decisions have to be taken as to which sites should be considered for treatment or safety improvement. The most economically reasonable selection criterion is to select those sites which had the highest accident rate in the preceding year. This is a bad procedure because of the well known regression to the mean problem. Even if no remedial treatment is made, the number of accidents recorded at the same site in the following year will naturally decrease toward its temporal mean. In other word, very high accident rates should be viewed as outliers.

In this paper we estimate three different discrete choice models of provider choice using data from the rural District of Ouidah in Benin. These three model are: Multinomial Logit (ML); (2) Independent Multinomial Probit (IMP); (3) Multinomial Probit (MP). A comparison of IMP and MP allows us to reject the independence assumption between providers. Furthermore, the cross-price elasticities computed from the restrictive specifications (ML and IMP) are dramatically different from those computed from the more general one (MP). These results cast some doubt on the validity of the previous findings and policy recommendations that are typically based on the ML specification.

In this chapter we use empirical examples to demonstrate the usefulness of the generalized error component framework suggested in Bolduc et al. (1992) for dealing with the problem of correlation among the errors of a regression based on travel flow data. This methodology augments Standard error component decompositions with first-order spatial autoregressive processes, i.e., SAR(l), with the purpose of allowing for the different sources of misspecification generally associated with this type of model. The error component approach splits the error term into a sum of one component related to the zones in origin, one component associated with the zones in destination and a remainder. The interdependencies among the errors are modeled with the help of SAR(l) processes. This decompositional approach extends the previous works by Brandsma and Ketellapper (1979) and Bolduc et al. (1989) which also relied on spatial autoregressive processes to model the error correlation.

Several models have been proposed to express customer preference variation. In this paper, we express it not by customer segmentation but by latent class. Not only decision-making factors but also decision-making processes differ according to the class. We propose a service-choice behaviour modelling constructed with multiple decision-making processes using an estimation method based on the expectation-maximisation algorithm. As an empirical experiment, we constructed a model of telephone service choice with optional services. We assume that there are two typical decision-making processes for this choice. One is to choose an optional service type first. The other is to choose a telephone company first. A nested structure model for the classified customer segment is constructed for each process. Namely, the model has two types of nested models which correspond to two decision-making processes. In a comparison with conventional models, we showed that the proposed model can improve the accuracy of the model.

Hydro-Quebec vient de reporter a plus tard le debut des travaux du projet Grande-Baleine, La societe d'Etat appuie sa decision sur la prevision d'une demande quebecoise d'electricite plus faible a plus long terme. En cela, elle imite d'autres services d'electricite au Canada et aux Etats-Unis. Les raisons invoquees sont la recession economique de 1991 et les economies d'energie de la part des consommateurs. Cependant, lorsque nous considerons les facteurs sous-jacents a la demande d'electricite, a savoir l'evolution de la population, de l'economie, des prix relatifs des sources d'energie et des programmes d'economie d'energie, il est difficile de partager entierement le pessimisme d'Hydro-Quebec quant a l'evolution future de la demande d'electricite au Quebec. Ce sont ces facteurs qui sont analyses dans le contexte quebecois. Selon nos estimations, la demande d'electricite pourrait etre superieure de 11 TWh en l'an 2006, soit 6% par rapport a la prevision d'Hydro-Quebec parue dans le plan de developpement de mars 1990.

This paper provides evidence that workers' compensation insurance (WC) affects not only the occurrence but also the composition of reported injuries. In our theoretical approach, WC is the source of two interrelated moral hazard problems: underprovision of accident-preventing efforts by the insured worker (ex ante moral hazard) and false reporting of injuries (ex post moral hazard). Our model predicts that, under certain assumptions, the impact of WC benefits is stronger on the probability of reporting a difficult-to-diagnose injury than on the probability of reporting an injury that is easy to diagnose. Panel data on 9,800 workers in the Quebec construction industry over each month of the period 1977-86, combining administrative data from the Quebec Construction Board with data from the Quebec Workers' Compensation Board, are used for the estimates. The parameters of the model are estimated using a threealternative logit kernel [hybrid multinomial probit (MNP)] framework with individual random effects. Our results confirm our theoretical pre

In this paper we use a micro approach to model Quebec residential demand for electricity. The model incorporates a joint continuous/discrete decision framework which allows for interrelationships between decisions on electricity-related durable holdings and those on usage. In the spirit of recent studies, which used a continuous/discrete frame- work, the model parameters are estimated using a two-stage approach. At the first stage the decisions regarding space and water heating systems are modelled with a very flexible Multinomial Probit (MNP) framework. Then, at the second stage, the demand for electricity conditional on the chosen heating system is estimated using ordinary least squares, and a correction is applied in order to eliminate a potential estimation bias. The estimated short-run and long-run price and income elasticities have the correct signs and are rather small, as is expected under such circumstances.

In this study, we use a first-order spatial autoregressive formulation to model the correlation among the errors of a linear demand equation that explains origin-destination flows. The process splits the error term for each observation into a weighted sum of all the other errors and a purely random noise. The weights are new parametric functional forms defined to measure the proximity between origins and destinations of flows. The parameters of these weights, along with the other parameters of the model, are estimated by the method of maximum likelihood. We apply the technique to an aggregate binary logit share model that explains peak A.M. trips to work in Winnipeg, Canada.

Parce qu’il admet des structures très générales d’interdépendance entre les modalités, le probit polytomique (MNP) fournit une des formes les plus intéressantes pour modéliser les choix discrets qui découlent d’une maximisation d’utilité aléatoire. L’obstacle majeur et bien connu dans l’estimation de ce type de modèle tient à la complexité que prennent les calculs lorsque le nombre de modalités considérées est élevé. Cette situation est due essentiellement à la présence d’intégrales normales multidimensionnelles qui définissent les probabilités de sélection. Au cours des deux dernières décennies, de nombreux efforts ont été effectués visant à produire des méthodes qui permettent de contourner les difficultés de calcul liées à l’estimation des modèles probit polytomiques. L’objectif de ce texte consiste à produire un survol critique des principales méthodes mises de l’avant jusqu’à maintenant pour rendre opérationnel le cadre MNP. Nous espérons qu’il éclairera les praticiens de ces modèles quant au choix de technique d’estimation à favoriser au cours des prochaines années., The Multinomial Probit (MNP) model provides the most general framework to allow for interdependent alternatives in discrete choice analysis. The primary impediment to this methodology is related to the dimensionality of the response probabilities which are multifold normal integrals of about the size of the choice set. During the last two decades, numerous researches have been devoted to develop practical methodologies to replace these hard to compute choice probabilities in the estimation process. The main objective of this paper is to survey the major and the most important of these techniques.

In the presence of choice-based sampling strategies for data collection, the property of multinomial logit (MNL) models, that consistent estimates of all parameters but the constants can be obtained from an exogenous sample maximum likelihood (ESML) estimation, does not hold in general for generalized extreme value (GEV) models. We propose a consistent ESML estimator for GEV models in this context. We first identify a specific class of GEV models with the desired property that, similarly to MNL, the constants absorb the potential bias. We then propose a new and simple weighted conditional maximum likelihood (WCML) estimator for the more general case. Contrarily to the weighted exogenous sample maximum likelihood (WESML) estimator by Manski and Lerman [Manski, C., Lerman, S., 1977. The estimation of choice probabilities from choice-based samples. Econometrica 45, 1977–1988], the new WCML estimator does not require an external knowledge of the market shares. We show that this applies also to the case where alternatives are sampled from a large choice set, and we illustrate the use of the estimator on synthetic and real data.