The purpose of this research dissertation is to compare the effects of using small, medium, and large cordon designs in road tolling on residential and commercial neighborhoods in Portland, Oregon. Changes in land use patterns are assessed by comparing the projected output of each cordon scenario to a "no toll" alternative in 2035. The performance of each cordon design is tested using two different prices ($1.65 and $8) and compared to a default scenario 25 years after the initial implementation in MetroScope's year 0, 2010. The following areas embedded within the cordon perimeter were considered in determining changes in land use: all the zones closest to the boundaries, the entire city of Portland, and the neighborhood towns surrounding Portland. Understanding the impact of choosing the "right" cordon size on economic development and residential location choices can be of utmost interest to lawmakers when they assess economic development policies. Unanswered questions remain regarding the impact a cordon scheme has on economic development and business location decisions, as well as its effect on the spatial pattern in the city. While studies on optimal toll pricing are abundant, there are very few studies that determine the optimal cordon location and size for a particular network. Accordingly, a critical question is whether cordon pricing will influence the centralization or decentralization of land use and affect jobs, population, and economic activities. The implementation of a cordon scheme is expected to affect areas both inside and outside of the designated perimeter and is further expected to contribute to changes that will affect land use. Existing studies have ignored land use effects and, instead, assume a monocentric city model. What sets this study apart is that instead of using a monocentric model to test the hypothetical cordon scenarios, the MetroScope model is used to predict changes in economy, demographics, and land use. The MetroScope model is one of only a few models that can assist in forecasting changes in both land use and prices. This study found primary evidence that the implementation of diverse sizes of cordon designs differently affect residential and non-residential land use patterns and trends.

In this thesis we study the problem of road taxation. This problem consists in finding the toll on the roads belonging to the government or a private company in order to maximize the revenue. An optimal taxation policy consists in determining level of tolls low enough to favor the use of toll arcs, and high enough to get important revenues. Since there are twolevels of decision, the problem is formulated as a bilevel bilinear program. / Doctorat en sciences, Orientation recherche opérationnelle / info:eu-repo/semantics/nonPublished