Travel demand forecasting is essential for many decisions, such as infrastructure investments and policy measures. Traditionally travel demand modelling has considered trip frequency, mode, destination and route choice. This thesis considers two other choice dimensions, hypothesised to have implications for travel demand forecasting. The first part investigates how the increased possibilities to overcome space that ICT (information and communication technology) provides, can be integrated in travel demand forecasting models. We find that possibilities of modelling substitution effects are limited, irrespective of data source and modelling approach. Telecommuting explains, however, a very small part of variation in work trip frequency. It is therefore not urgent to include effects from telecommuting in travel demand forecasting. The results indicate that telecommuting is a privilege for certain groups of employees, and we therefore expect that negative attitudes from management, job suitability and lack of equipment are important obstacles. We find also that company benefits can be obtained from telecommuting. No evidences that telecommuting gives rise to urban sprawl is, however, found. Hence, there is ground for promoting telecommuting from a societal, individual and company perspective. The second part develops a departure time choice model in a mixed logit framework. This model explains how travellers trade-off travel time, travel time variability, monetary and scheduling costs, when choosing departure time. We explicitly account for correlation in unobserved heterogeneity over repeated SP choices, which was fundamental for accurate estimation of the substitution pattern. Temporal constraints at destination are found to mainly restrict late arrival. Constraints at origin mainly restrict early departure. Sensitivity to travel time uncertainty depends on trip type and intended arrival time. Given appropriate input data and a calibrated dynamic assignment model, the model can be applied to forecast peak-spreading effects in congested networks. Combined stated preference (SP) and revealed preference (RP) data is used, which has provided an opportunity to compare observed and stated behaviour. Such analysis has previously not been carried out and indicates that there are systematic differences in RP and SP data. / QC 20100825

La naturaleza potencialmente anticompetitiva de las prácticas impartidas desde las organizaciones de profesionales ha sido racionalizada por la literatura económica y ha ocupado un lugar no menor en la agenda de los organismos de defensa de la competencia. Sin embargo, la economía empírica carece de estudios sobre el ejercicio profesional autorregulado. Este trabajo relaciona los mercados de servicios profesionales con los modelos de diferenciación vertical y emplea un modelo Logit Mixto para evaluar la conducta del desempeño médico en Argentina. La evidencia hallada sugiere la existencia de algún acuerdo de precios. / The potentially anticompetitive nature of some practices driven by professional organizations has been approached in economic literature and appears as an important issue in the antitrust organism's agenda. However, empirical economics lacks of a self-regulated professionals analysis. This work relates the market for professional services with vertical product differentiation models and uses a Mixed Logit model to assess the medical profession behavior in Argentina. The evidence suggests the existence of a price arrangement. / Tesis de la Maestría en Economía, bajo la dirección de Fernando Navajas, de la Facultad de Ciencias Económicas de la Universidad Nacional de La Plata.

New Zealand is internationally renowned for having some of the finest and most challenging trout fishing in the world. However, due to continuing development and angling pressure many fishing sites are showing signs of environmental degradation and over fishing. This trend is almost certain to continue into the future given continued population and economic growth. Understanding the determinants of site choice, preference heterogeneity and anglers’ substitution patterns is fundamentally important to fishery managers who have the difficult task of maintaining quality angling experiences on a number of fishing sites, managing angling pressure and maintaining license sales. Recent advances in simulation techniques and computational power have improved the capability of discrete choice models to reveal preference heterogeneity and complex substitution patterns among individuals. This thesis applies and evaluates a number of state-of-the-art discrete choice models to study angler site choice in New Zealand. Recreation specialisation theory is integrated into the analysis to enhance the behavioural representation of the statistical models. A suite of models is presented throughout the empirical portion of this thesis. These models demonstrate different ways and degrees of explaining preference heterogeneity as well as identifying anglers’ substitution patterns. The results show that North Canterbury anglers’ preferences vary considerably. Resource disturbances such as riparian margin erosion, reduced water visibility and declines in catch rates can cause significant declines in angler use of affected sites, and at the same time non-proportional increases in the use of unaffected sites. Recreation specialisation is found to be closely related to the types of fishing site conditions, experiences and regulations preferred by anglers. Anglers’ preference intensities for fishing site attributes, such as catch rates, vary across different types of fishing sites. This location specific preference heterogeneity is found to be related to specialisation. Overall, the empirical findings indicate that conventional approaches to modelling angler site choice which do not incorporate a strong understanding of angler preference heterogeneity can lead to poorly representative models and suboptimal management and policy outcomes.

Submitted by Bernardo Dantas Pereira Coelho (bernardo.coelho27@gmail.com) on 2018-05-02T21:35:04Z No. of bitstreams: 1 Tese Doutorado - Bernardo Coelho.pdf: 2101693 bytes, checksum: 84ef02769381cf1afb6f413b79ebde05 (MD5) / Approved for entry into archive by GILSON ROCHA MIRANDA (gilson.miranda@fgv.br) on 2018-05-28T16:24:46Z (GMT) No. of bitstreams: 1 Tese Doutorado - Bernardo Coelho.pdf: 2101693 bytes, checksum: 84ef02769381cf1afb6f413b79ebde05 (MD5) / Made available in DSpace on 2018-06-14T20:31:22Z (GMT). No. of bitstreams: 1 Tese Doutorado - Bernardo Coelho.pdf: 2101693 bytes, checksum: 84ef02769381cf1afb6f413b79ebde05 (MD5) Previous issue date: 2017-08-18 / Essa tese contém três capítulos. O primeiro capítulo estuda a relação entre o programa brasileiro de transferência condicional de renda Bolsa Família e os resultados das eleições de 2010. Nós procuramos estimar esse efeito utilizando uma abordagem estrutural, identificando características individuais que afetam o impacto eleitoral do programa. Fazemos isso utilizando um modelo mixed logit, um modelo de escolha discreta que considera tanto a distribuição paramétrica de variáveis não observadas quanto a distribuição não-paramétrica de variáveis conhecidas. Resultados indicam que o caráter redistributivo do programa possui um impacto eleitoral nos eleitores maior do que os ganhos individuais de renda dos beneficiários. O efeito marginal de ser um beneficiário do programa na decisão de voto é equivalente a um aumento de 81 reais na renda mensal do trabalho, menos do que o valor médio recebido por beneficiário que é de 90 reais. Nosso exercício contrafactual aponta que, sem o programa Bolsa Família, a incumbente, Sra. Rousseff, perderia 5,6% do total de votos, deixando o resultado da eleição inconclusivo. O segundo capítulo estuda a participação feminina na política, que aumentou na última década tanto em países ricos como em desenvolvimento. Não é claro, no entanto, se isso é parte de uma tendência ou apenas um crescimento reversível. A literatura apresenta argumentos teóricos tanto para um efeito de reforço quanto para um negativo da exposição a uma liderança negativa na probabilidade de apoio futuro a uma candidata mulher. Usando dados eleitorais e do Censo para o Brasil, testamos se o efeito da presença de uma prefeita mulher numa cidade impacta o apoio futuro a candidatas mulheres para Deputada Federal e não encontramos evidência de efeito significativo. Além disso, mostramos que apenas o uso de estatísticas agregadas, como médias demográficas, levaria a concluir equivocadamente que eleitores expostos ao governo de uma prefeita mulher teriam uma menor probabilidade de votar numa candidata mulher. O último capítulo investiga os determinantes para a queda de desigualdade de renda entre municípios brasileiros entre 2000 e 2010. Usando dados censitários, mostramos que a desigualdade caiu mais rápido em municípios com um maior nível de desigualdade em 2000 – sugerindo -convergência. Nós então, utilizamos a decomposição dinâmica (Shorrocks, 1982) para identificar a contribuição de mudanças nas condições do mercado de trabalho, como aumento do salário mínimo, formalização e melhoria na educação na convergência de desigualdade regional. Encontramos que a queda na desigualdade de renda no emprego formal foi o principal contribuinte para a redução de desigualdade de renda entre municípios no período. / This thesis contains three chapters. The first chapter studies the relationship between the Brazilian CCT program Bolsa Família and the outcome of the 2010 elections. We seek to estimate this effect using a structural approach, identifying individual characteristics that affect the electoral impact of the program. We do so by using a mixed logit model, a discrete choice model that considers both a parametrical distribution of unobserved variables and a non-parametrical distribution of known variables. Results indicate that the redistributive character of the program has a larger electoral impact on voters than the individual income gains of the beneficiaries. The marginal effect of being a beneficiary of the program on voting decision is equivalent to 81 Reais increase in monthly labor income, less than the average value received by a beneficiary, which is 90 reais. Our counterfactual exercise points that, without Bolsa Família, the incumbent, Mrs. Rousseff, would have lost 5.6% of the votes, making the election results unclear. The second chapter studies female participation in politics has increased in the last decade in both rich and developing countries. It is not clear, however, if this is part of a trend or just a reversible growth. Literature presents theoretical arguments for both a reinforcing force and a negative effect of the exposure to a female leadership on the probability of supporting a future female candidate. Using electoral and Census data for Brazil, we test the effect that the presence of a female mayor in a municipality has on future the support for a female candidate for Federal Deputy and find no evidence of a significant effect. Furthermore, we show that the use of aggregate statistics alone, as demographic averages, would mislead us to conclude that voters exposed to a female mayor have a smaller probability to support a female candidate. The last chapter investigates the determinants of the decline of income inequality across municipalities in Brazil between 2000 and 2010. Using censuses data, we show that inequality fell faster in municipalities with higher inequality levels in 2000 – suggesting - convergence. We, then, employ a dynamic decomposition (Shorrocks, 1982) to assess the contribution of changes in private labor market conditions as the increase in minimum wage, formalization and increase in education levels on the regional inequality convergence. We find that the fall in wage inequality in the private formal sector was the main driver of the reduction in income inequality across municipalities in the period.

Travel demand forecasting is essential for many decisions, such as infrastructure investments and policy measures. Traditionally travel demand modelling has considered trip frequency, mode, destination and route choice. This thesis considers two other choice dimensions, hypothesised to have implications for travel demand forecasting. The first part investigates how the increased possibilities to overcome space that ICT (information and communication technology) provides, can be integrated in travel demand forecasting models. We find that possibilities of modelling substitution effects are limited, irrespective of data source and modelling approach. Telecommuting explains, however, a very small part of variation in work trip frequency. It is therefore not urgent to include effects from telecommuting in travel demand forecasting. The results indicate that telecommuting is a privilege for certain groups of employees, and we therefore expect that negative attitudes from management, job suitability and lack of equipment are important obstacles. We find also that company benefits can be obtained from telecommuting. No evidences that telecommuting gives rise to urban sprawl is, however, found. Hence, there is ground for promoting telecommuting from a societal, individual and company perspective. The second part develops a departure time choice model in a mixed logit framework. This model explains how travellers trade-off travel time, travel time variability, monetary and scheduling costs, when choosing departure time. We explicitly account for correlation in unobserved heterogeneity over repeated SP choices, which was fundamental for accurate estimation of the substitution pattern. Temporal constraints at destination are found to mainly restrict late arrival. Constraints at origin mainly restrict early departure. Sensitivity to travel time uncertainty depends on trip type and intended arrival time. Given appropriate input data and a calibrated dynamic assignment model, the model can be applied to forecast peak-spreading effects in congested networks. Combined stated preference (SP) and revealed preference (RP) data is used, which has provided an opportunity to compare observed and stated behaviour. Such analysis has previously not been carried out and indicates that there are systematic differences in RP and SP data. / QC 20100825

Managed lane refers to the application of various operational and design strategies on highway facilities to improve system efficiency and mobility by proactively allocating traffic capacity to different lanes. One of the key elements to understand the behavior changes and underlying causalities in user responses to managed lanes is to examine the value of time (VOT) and value of reliability (VOR). The breadth of this dissertation encompasses two major dimensions of VOT and VOR estimation – distributions or variations across different users and under different circumstances; and influences of unobserved attitudinal characteristics on roadway pricing valuation. To understand travelers’ choice behavior regarding the usage of managed lanes, combined revealed preference (RP) and stated preference (SP) data were used in this study. Mixed logit modeling was applied as the state of the art methodology to capture heterogeneity in users’ choice behavior. The model revealed an average value of $10.68 per hour for VOT and $13.91 per hour for VOR, which are reasonable considering the average household income in the region, and are well within the ranges found in the literature. In terms of user heterogeneity, the mixed logit model was further enhanced by adding interaction effects of variables, which helped recognize and quantify potential sources of heterogeneity in user sensitivities to time, reliability, and cost. The findings indicated that travelers were likely to exhibit higher willingness to pay when they were female, younger (years), older (>54 years), had higher income (> 50 K), driving alone, and traveled on weekdays. Attitudinal aspects are rarely incorporated into roadway pricing analysis. The study herein presents an effort to explore the role of attitudinal factors in drivers’ propensity toward using managed lanes. Model results boded for a significant contribution of attitudinal parameters in the model, both in terms of coefficients and model performance. This study provides a robust approach to quantify user heterogeneity in VOT and VOR and capture the impacts of attitudinal attributes in pricing valuation. The results of this study contribute to a better understanding on what attributes lead to higher or lower VOT and VOR and to what extent.

Artificial Intelligence in form of Machine Learning classifiers is increasingly applied for travel choice modeling issues and therefore constitutes a promising, competitive alternative towards conventional discrete choice models like the Logit approach. In comparison to traditional theory-based models, data-driven Machine Learning generally shows powerful predictive performance, but often lacks in model interpretability, i.e., the provision of comprehensible explanations of individual decision behavior. Consequently, the question about which approach is superior remains unanswered. Thus, this paper performs an in-depth comparison between benchmark Logit models and Artificial Neural Networks and Decision Trees representing two popular algorithms of Artificial Intelligence. The primary focus of the analysis is on the models’ prediction performance and its ability to provide reasonable economic behavioral information such as the value of travel time and demand elasticities. For this purpose, I use crossvalidation and extract behavioral indicators numerically from Machine Learning models by means of post-hoc sensitivity analysis. All models are specified and estimated on synthetic and empirical data. As the results show, Neural Networks provide plausible aggregate value of time and elasticity measures, even though their values are in different regions as those of the Logit models. The simple Classification Tree algorithm, however, appears unsuitable for the applied computation procedure of these indicators, although it provides reasonable interpretable decision rules for travel choice behavior. Consistent with the literature, both Machine Learning methods achieve strong overall predictive performance and therefore outperform the Logit models in this regard. Finally, there is no clear indication of which approach is superior. Rather, there seems to be a methodological tradeoff between Artificial Intelligence and discrete choice models depending on the underlying modeling objective.

Movements of cyclists and m.otor vehicles at intersections involve a wide variety of potential conflicting interactions. In Australia, the high numbers of motor vehicles, particularly light passenger vehicles, mixed with cyclists results in many bicycle-light passengervehicle (LPV) crashes (3,135 crashes during 2002-2014). About 68% of cyclist deaths at Australian intersections in 2016 were due to crashes between bicycles and LPVs (DITRLDG, 2016). The high number ofLPV crashes among fatalities among cyclists is an increasing safety concem. When an LPV collides with a cyclist, the resulting impact forces in.tluence the probability of cyclist injury severity outcom.e. Therefore, the goa1 at intersections should be to understand whether and which particular crash patterns are more injurious, in order to better inform approaches to reduce the impact forces to levels that do not result in severe injury outcomes. To examine how crash pattem (or mechanism) influences the injury severity of cyclists in bicycle-motor vehicle crashes at intersections, researchers typically describe the crash mechanism in terms of crash types, such as angle crashes, head--on crashes, rear-end crashes, and sideswipe crashes (e.g., Kim et al., 2007; Pai, 2011 ). While crash types explain crash mechanisms to some extent, this study hypothesiz.es that the trajectories of the crash involved vehicles may provide additional information because they better capture the movements of the vehicles prior to collision. Furthermore, it is argued that injury pattem might be in.tluenced by vehicle travel direction and manoeuvre (Isaksson-Hellman and Wemeke, 2017). For example, when a car is moving straight ahead it is likely to have a higher speed than when it is turning, and if cyclists are struck at a higher impact speed, they tend to sustain more severe injury (Badea-Romero and Lenard, 2013). While many studies have evaluated the association between cyclist injwy severity and crash types, the factors that might influence cyclist injury severity related to trajectory types (vehicle movement and travel direction) have not yet been thoroughly investigated. This study aims to examine the factors associated with cyclists' injury severity for 'trajectory types• compared with the typically used 'crash types' at intersections.

個體選擇模式(discrete choice model)廣泛應用於國外的交通運輸及行銷領域，而國內交通運輸領域，也長期以此模式分析個體的運具選擇行為。反觀國內的行銷領域，因較難取得消費者的商品品牌購買紀錄，而鮮少應用個體選擇模式分析消費者的選擇行為。有鑒於此，本研究嘗試以問卷收集消費者對三個洗髮精品牌的選擇行為，以個體選擇模式中的多項邏輯模式(multinomial logit model)、巢狀邏輯模式(nested multinomial logit model)、混合多項邏輯模式(mixed logit model)進行分析，檢驗問卷設計中的促銷活動、消費者特性對選擇行為的影響性。 實證分析的結果發現，洗髮精的原價格及促銷折扣、贈品容量、加量不加價等促銷活動，皆對消費者的選擇行為有顯著的影響力，其中促銷折扣與贈品容量影響的程度較大，是較具有效果的促銷活動。而消費者的性別、年齡、職業及品牌更換的頻率，皆影響洗髮精的選擇行為。此外，消費者若固定選擇自己最常購買的洗髮精，此類型的消費者與其他人的品牌選擇行為，也有顯著的不同。 此外，根據本研究樣本，我們也發現海倫仙度絲與潘婷間的替代、互補性較強。 / Discrete choice model has been demonstrated to be a useful tool for analyzing consumers’ choice behavior data in the area of transportation and marketing research. However, since a complete data set containing consumers’ history of purchase behavior was rarely available to public, the model was less popular in the marketing research area than in the transportation research in Taiwan. Based on limited survey data on consumers’ choice among three different brands of shampoo, we applied multinomial logit model、nested multinomial logit model、mixed logit model in this study to understand promotion program’s effect on consumers’ choice behavior , the result showed that shampoos’ original price、discount、volume of hair conditioner bestowal、more volume with the same price all had significant impacts on consumers’ choice behavior, among them, discount and volume of hair conditioner bestowel influenced more .In addition, consumers’ gender、age、occupation and frequency of changing brands also affected consumers on choosing brands of shampoos. The study also found that a consumer who chose the same brand regularly behaved notably differently.

Le problème de tarification qui nous intéresse ici consiste à maximiser le revenu généré par les usagers d'un réseau de transport. Pour se rendre à leurs destinations, les usagers font un choix de route et utilisent des arcs sur lesquels nous imposons des tarifs. Chaque route est caractérisée (aux yeux de l'usager) par sa "désutilité", une mesure de longueur généralisée tenant compte à la fois des tarifs et des autres coûts associés à son utilisation. Ce problème a surtout été abordé sous une modélisation déterministe de la demande selon laquelle seules des routes de désutilité minimale se voient attribuer une mesure positive de flot. Le modèle déterministe se prête bien à une résolution globale, mais pèche par manque de réalisme. Nous considérons ici une extension probabiliste de ce modèle, selon laquelle les usagers d'un réseau sont alloués aux routes d'après un modèle de choix discret logit. Bien que le problème de tarification qui en résulte est non linéaire et non convexe, il conserve néanmoins une forte composante combinatoire que nous exploitons à des fins algorithmiques. Notre contribution se répartit en trois articles. Dans le premier, nous abordons le problème d'un point de vue théorique pour le cas avec une paire origine-destination. Nous développons une analyse de premier ordre qui exploite les propriétés analytiques de l'affectation logit et démontrons la validité de règles de simplification de la topologie du réseau qui permettent de réduire la dimension du problème sans en modifier la solution. Nous établissons ensuite l'unimodalité du problème pour une vaste gamme de topologies et nous généralisons certains de nos résultats au problème de la tarification d'une ligne de produits. Dans le deuxième article, nous abordons le problème d'un point de vue numérique pour le cas avec plusieurs paires origine-destination. Nous développons des algorithmes qui exploitent l'information locale et la parenté des formulations probabilistes et déterministes. Un des résultats de notre analyse est l'obtention de bornes sur l'erreur commise par les modèles combinatoires dans l'approximation du revenu logit. Nos essais numériques montrent qu'une approximation combinatoire rudimentaire permet souvent d'identifier des solutions quasi-optimales. Dans le troisième article, nous considérons l'extension du problème à une demande hétérogène. L'affectation de la demande y est donnée par un modèle de choix discret logit mixte où la sensibilité au prix d'un usager est aléatoire. Sous cette modélisation, l'expression du revenu n'est pas analytique et ne peut être évaluée de façon exacte. Cependant, nous démontrons que l'utilisation d'approximations non linéaires et combinatoires permet d'identifier des solutions quasi-optimales. Finalement, nous en profitons pour illustrer la richesse du modèle, par le biais d'une interprétation économique, et examinons plus particulièrement la contribution au revenu des différents groupes d'usagers. / The network pricing problem consists in finding tolls to set on a subset of a network's arcs, so to maximize a revenue expression. A fixed demand of commuters, going from their origins to their destinations, is assumed. Each commuter chooses a path of minimal "disutility", a measure of discomfort associated with the use of a path and which takes into account fixed costs and tolls. A deterministic modelling of commuter behaviour is mostly found in the literature, according to which positive flow is only assigned to \og shortest\fg\: paths. Even though the determinist pricing model is amenable to global optimization by the use of enumeration techniques, it has often been criticized for its lack of realism. In this thesis, we consider a probabilistic extension of this model involving a logit dicrete choice model. This more realistic model is non-linear and non-concave, but still possesses strong combinatorial features. Our analysis spans three separate articles. In the first we tackle the problem from a theoretical perspective for the case of a single origin-destination pair and develop a first order analysis that exploits the logit assignment analytical properties. We show the validity of simplification rules to the network topology which yield a reduction in the problem dimensionality. This enables us to establish the problem's unimodality for a wide class of topologies. We also establish a parallel with the product-line pricing problem, for which we generalize some of our results. In our second article, we address the problem from a numerical point of view for the case where multiple origin-destination pairs are present. We work out algorithms that exploit both local information and the pricing problem specific combinatorial features. We provide theoretical results which put in perspective the deterministic and probabilistic models, as well as numerical evidence according to which a very simple combinatorial approximation can lead to the best solutions. Also, our experiments clearly indicate that under any reasonable setting, the logit pricing problem is much smoother, and admits less optima then its deterministic counterpart. The third article is concerned with an extension to an heterogeneous demand resulting from a mixed-logit discrete choice model. Commuter price sensitivity is assumed random and the corresponding revenue expression admits no closed form expression. We devise nonlinear and combinatorial approximation schemes for its evaluation and optimization, which allow us to obtain quasi-optimal solutions. Numerical experiments here indicate that the most realistic model yields the best solution, independently of how well the model can actually be solved. We finally illustrate how the output of the model can be used for economic purposes by evaluating the contributions to the revenue of various commuter groups.