Abstract A great deal of research in transportation economics has been motivated by the need to solve traffic congestion problems and to diminish negative environmental effects of road transport. The question, whether the economic measures are efficient, motivates this dissertation on the value of travel time, the rules of optimal pricing and the demands for public transportation and private car use. Three concepts of the marginal value of travel time (MVT) are specified in this thesis. The first concept involves only the direct disutility of the travel time to work in addition to the utility of market goods and leisure. The second concept also includes the disutility from the time spent at work. The third concept furthermore takes into account the effect of the length of working hours, travel time, cost and income. The length of travel time, gender, family structure and flexibility of working hours have different effects on empirical MVTs, but travel costs and income affect them in a similar fashion. The pricing decisions of the firms providing bus services are analysed with and without public subsidies. The consumption externality, i.e. the quantity demanded by other users, affects the individual bus demand. The results indicate that under uniform pricing, a socially optimal subsidy equals the increase in consumer's surplus minus the fare revenue lost from current users due to lower fare. Under nonlinear pricing, the optimal pricing can be achieved when the regulator sets the subsidy so that it is inversely proportional to the network elasticity. The welfare loss due to increasing tax burden and the opportunity cost of providing cash fare service is also taken into account in the optimal pricing rule. A model of bus demand with asymmetric information on the characteristics of bus users is developed. The model allows for habit formation and network effects. The latter effect is due to the positive influence of the aggregate demand for Regional Bus Cards (RBC) on an individual's own demand. The empirical results indicate that in RBC services positive network effects are present and the elasticity of network size is less than one, which implies that the regional bus card is an impure public good. The own price elasticity of RBC in the short run is within the range of -0.3 and -1.1. The demand for RBC cards is more elastic than demand for RBC trips or passenger kilometres. The estimated price elasticity of urban bus demand is in line with that of RBC. A reasonably high cross-price elasticity of RBC trips and the ticket of 40 trips but a lower reverse elasticity were found. A weakly separable demand for car mileage from car ownership and labour supply was rejected as was the exogeneity of car ownership in the mileage model. Therefore, the price elasticity of car mileage with respect to fuel price was estimated from the two equation model of car mileage with endogenous car ownership. The estimated parameters of the Tobit model are consistent but slightly higher than those estimated from the least squares. The fuel price elasticity varies from -0.2 to -0.9 with exogenous and endogenous car ownership, respectively. The findings of this study can be applied in the analysis and implementation of different pricing and subsidy schemes for public transportation, as well as in the evaluation of the effectiveness of economic instruments for managing the growth of private car use.

Over the past decade, the popularity of regional bus services has grown in large North American cities owing to more people living in suburban areas and commuting to the Central Business District to work every day. Estimating journey time for regional buses is challenging because of the low frequencies and long commuting distances that typically characterize such services. This research project developed a mathematical model to estimate regional bus travel time using artificial neural networks (ANN). ANN outperformed other forecasting methods, namely historical average and linear regression, by an average of 35 and 26 seconds respectively. The ANN results showed, however, overestimation by 40% to 60%, which can lead to travellers missing the bus. An operational strategy is integrated into the model to minimize stakeholders’ costs when the model’s forecast time is later than the scheduled bus departure time. This operational strategy should be varied as the commuting distance decreases.

Over the past decade, the popularity of regional bus services has grown in large North American cities owing to more people living in suburban areas and commuting to the Central Business District to work every day. Estimating journey time for regional buses is challenging because of the low frequencies and long commuting distances that typically characterize such services. This research project developed a mathematical model to estimate regional bus travel time using artificial neural networks (ANN). ANN outperformed other forecasting methods, namely historical average and linear regression, by an average of 35 and 26 seconds respectively. The ANN results showed, however, overestimation by 40% to 60%, which can lead to travellers missing the bus. An operational strategy is integrated into the model to minimize stakeholders’ costs when the model’s forecast time is later than the scheduled bus departure time. This operational strategy should be varied as the commuting distance decreases.