Four research chapters included in this dissertation analyze how changes in individuals’ travel behavior induced by information technologies affect social welfare. Chapters 2, 3 and 4 develop stylized microeconomic models which are largely set within the frameworks of transport and urban economics, and economics of industrial organization. Chapter 5 is an econometric exercise that uses an extensive dataset at the individual level. Several types of information technology are considered, encompassing both substitutes for and complements to traveling. Chapters 2 and 5 study telework-enabling technology that facilitates out-of-office work arrangements, while traffic information for drivers that supports efficient travel choices is examined in Chapter 4. First-best road tolling, varying over time of day, is the focus of Chapter 3. Given the important role that technology plays in the practical implementation of road pricing, such an analysis also provides insights into the social welfare consequences that technology-enhanced solutions to transportation problems might have. This thesis contributes to the existing literature in three ways. Firstly, it proposes novel ways of investigating, from an economic perspective, the relationship between information technologies and travel behavior. Chapter 2 incorporates into Vickrey’s (1969) microeconomic equilibrium model of dynamic road bottleneck congestion an analysis of the behavioral and welfare impacts of a teleworking technology. Chapter 3 combines Vickrey s bottleneck model and the monocentric city model, and analyzes individual travel and location behavior when travel and residential location choices are mutually dependent. Chapter 4 analyzes equilibrium and welfare in a market where the road operator and the traffic information provider are independent, and can be public or private. Secondly, Chapter 5 provides an empirical contribution to the economic literature, by suggesting an identification strategy for the estimation of a causal long-run effect of teleworking technology adoption on commuting distances and, indirectly, on negative travel externalities. We combine data from a large labor force survey from a period before information technology became widespread and a period when Internet is ubiquitous. This offers an interesting way to addressing a potential reverse causality problem of endogenous technology access. Thirdly, the findings reported in this thesis may help policy makers to promote theoretically sound measures to tackle long standing transportation problems. The results of the thesis help to be aware of the economic consequences, both desirable and not, of novel transport policies. A brief overview is as follows. Chapter 2 studies the potential equilibrium and welfare effects of teleworking for congested road commuting. Teleworking is broadly defined as an out-of-office work arrangement where an employee can perform some of the job tasks from home. Unlike the majority of the existing studies that analyze the economic effects of whole-day teleworking, this chapter focuses on the travel impacts of - empirically relevant - part-day teleworking. This means that an employee might work from home for a few hours in the morning to avoid peak period congestion. The analysis is cast in the framework of Vickrey s (1969) dynamic bottleneck model, which allows analyzing the individuals’ departure time choices as an outcome of trade-offs between the opportunity costs of spending time at home, work and in a vehicle. The possibility to work from home due to the teleworking-enabling technology, such as Internet and computers, translates in a higher marginal utility of spending additional time at home, and thus changes the travel trade-off that a person faces. It turns out that an individual who is equipped with teleworking-enabling technology tends to postpone the commute, and to depart from home later than she would otherwise do. While it is beneficial for the teleworker, it might also be welfare improving for unequipped drivers who face weaker congestion. However, the larger the number of drivers who telework and postpone their commute in the same manner, the more the congestion peak period shifts to a later time period. Social benefits from a dispersion of preferences may then decrease in the number of teleworkers. This result shows that, in absence of optimal congestion pricing, there is an optimal level of technology penetration above which teleworking technology, even if it is free-of-charge, might be socially detrimental. Due to this, private monopolistic supply of the technology, albeit restrictive, might yield a higher social welfare than perfectly competitive supply. Chapter 3 analyses the longer run economic interactions between commuting, including the scheduling of commuter trips, and urban development. This chapter attempts to merge, for the first time, Vickrey’s bottleneck model and the monocentric city model in which both scheduling and location choices influence each other. The new model considers a monocentric city where a traffic bottleneck is located at the entrance of the central business district. The commuters’ departure times from home, residential locations, and lot sizes, are all endogenous. The results show that the elimination of queuing time under optimal road pricing induces individuals to spend more time at home and therefore to have larger houses, thus causing urban sprawl in the long run. This is opposite to the typical results of urban models with static congestion, which predict cities to become denser with road pricing. To reach this conclusion the model considers how the amount of time that drivers spend at home and in the car changes due to the introduction of congestion pricing. Because drivers would no longer spend time waiting in the car, the amount of time spent at home increases. An important assumption that the model introduces is that, if everything else is constant, the larger the house one lives in, the more benefit one derives from spending additional time in it. Thus, when a driver spends more time at home in the morning, he or she has a stronger demand for a larger house. While this effect might be relatively small at an individual level, over the long term for the entire city it might be noticeable. Despite the difference with findings in static models, the result is in another sense in line with previous studies that show that improved urban transportation, such as availability of extended road capacity or cheaper modes of transportation, causes urban sprawl as well. Chapter 4 considers welfare effects of traffic information provided to drivers. While both private and public firms might inform drivers about traffic conditions, it is not clear to what extent these arrangements are socially beneficial in a market where a separate firm operates a road network. This chapter uses a simple microeconomic model of elastic travel demand and stochastic travel cost to derive an endogenous demand for traffic information. Profit-maximizing pricing strategies of the information provider and road operator depend on each other as both road toll and price of traffic information affect the number of drivers making a trip, and thus the market size that can generate profit. It appears that the distortive welfare effect of monopolistic information provision is relatively small. The main pricing strategy of the information provider is to appropriate the (constant) consumer surplus of drivers who travel even when the travel costs are high. The monopolistic mark-up redistributes surplus from the consumers to the information monopolist, but does not crowd many drivers out of the market. The results show that a cooperating road operator and traffic information provider offer a lower joint price compared to the case where these companies operate separately, because their profits are interrelated. The mechanism behind this is closely related to the regular argument in models of double marginalization, albeit that now both firms sell directly to consumers and the two goods are not strictly complementary: one could travel without having the information. There appears little reason to prevent private road operators from offering information on traffic conditions on their roads. Chapter 5 returns to the topic of teleworking and examines its spatial consequences empirically. Despite strong policy support for teleworking, economists raise concerns that if commuting becomes less expensive due to teleworking, an employee might change the residential location to a (cheaper) place further away from the workplace. Additional kilometers traveled, although less frequently, could potentially offset the positive effects of teleworking on, for example, congestion and air pollution. This chapter checks whether within professions where many people telework, average commuting distance is affected by the technology. The identification strategy of chapter 5 tests this assertion while taking care of the reverse causality problem of endogenous technology availability, by applying a difference-in-differences method in combination with propensity score matching. Cross-sectional Dutch labor force survey data from 1996, when technology was barely present, and 2010, when Internet was pervasive, provides a way to estimate the long-run casual effect of the adoption of teleworking practices on the length of commuting distances. The results indicate that average commuting distances have increased over time by 2 km for professions where substantial share of employees telework. The same increase took place over time in the professions where no one teleworks. This implies that the adoption of teleworking technologies does not affect average commuting distances. The results also suggest that there is a divergent effect of technology on teleworkers and non-teleworkers within teleworking professions. While teleworkers would still have a longer commute, on average it is counterweighted by the reduction of commuting distances for non-teleworkers, as compared to non-teleworkers in non-teleworking professions. This puzzle warrants an explanation which appears difficult to pinpoint. (Author/publisher)
Abstract