2009 Student Speakers


Evaluation of Pricing Effects for Water Using Discrete-Continuous Choice (DCC) with Multilevel Modeling

David Hsu
dhsu2@u.washington.edu

Slides (pdf)

Public water utilities have increasingly turned to increasing block rate price structures to reduce water consumption and signal the high environmental costs of water supply. This paper evaluates the impact of a new and substantially higher price block added to the existing block rate price structure in Seattle -- often referred to as a `shock rate' -- which only affected those who consume very high quantities of water. The public water utility in Seattle added such a rate to its existing price structure in 2001, and has subsequently seen significant decreases in per capita water demand.

Rigorous evaluation of policies such as the shock rate is often constrained by the limited availability of appropriate data that describes individual consumption decisions. Much of the previous literature of water demand relies upon aggregated data which often results in theoretically implausible results such as price elasticities with the wrong signs. As a result of these empirical data limitations, there has also been limited application of theoretically appropriate models such as DCC models within the literature of demand for water and other resources. Furthermore, data limitations often make it difficult to elucidate how broadly-applied environmental policies might affect different neighborhoods and groups differently.

This paper applies a DCC model to a new, rich source of observational micro-data to evaluate changes in water consumption in Seattle as a result of the new pricing structure. A comprehensive billing database of water consumption for individual households in the period from 1991 to 2007 was obtained from Seattle Public Utilities. The DCC model was combined with multilevel modeling in order to develop a theoretically appropriate model to describe realistically the effect of inclined block rate price structures on water consumption, and to identify group-level effects on water consumption.

Regional Growth Centers and Their Impact on Travel Behavior: The Case of Puget Sound

Alon Bassok
abassok@u.washington.edu

Slides (pdf)

Under the Washington State Growth Management Act (1990), the Seattle metropolitan area introduced Urban Growth Areas (UGAs) and 21 urban centers to accommodate population, housing, and job growth. The primary purpose is to accomplish jobs and housing balance and reduce the needs for vehicular travel by promoting public and non-motorized transportation. Furthermore, the region's transportation plan adopted in 2002, Destination 2030, received a national award from the American Planning Association as "America's Best Plan." Yet, the Seattle metropolitan area is one of the most congested metropolitan areas in the U.S. as reported in several Texas Transportation Institute (TTI) congestion reports. What has influenced this dichotomy? This paper investigates the effectiveness of the growth-center strategy in relation to transportation mode shifts in comparison with prior to- and post-UGA policy. A counterfactual planning method is introduced, and U.S. Census block groups within urban centers are matched to ones outside of them for comparison. It is shown that urban centers have a positive but modest impact on the increase of transit usage.

Re-examining the influence of work and non-work accessibility on residential location choices with a micro-analytic framework

Brian Lee
bhylee@u.washington.edu

Paper (pdf), Slides (pdf)

The concept of accessibility has long been theorized as a principal determinant of residential choice behavior. Research on this influence is extensive but the empirical results have been mixed, with some research suggesting that accessibility is becoming a relatively insignificant influence on housing choices. Further, the measurement of accessibility must contend with complications arising from the increasing prevalence of trip-chains, non-work activities, and multi-worker households, and reconcile person-specific travel needs with household residential decisions. This paper contributes to the literature by addressing the gap framed by these issues and presents a novel residential choice model with three main elements of innovation. First, it operationalized a time-space prism (TSP) accessibility measure, which the authors believe to be the first application of its kind in a residential choice model. Second, it represented the choice sets in a building-level framework, the lowest level of spatial disaggregation available for modeling residential choices. Third, it explicitly examined the influence of nonwork accessibility at both the local- and person-level. This residential choice model was applied in the central Puget Sound region using a 2006 household activity survey. The model estimation results confirmed that accessibility remain an important influence, with individual-specific work accessibility as the most critical consideration. By using the TSP approach, it was established that non-work accessibility in a trip-chaining context does contribute to the residential choice decision, even after accounting for work accessibility. Empirical tests also revealed a useful aggregation method to incorporate individual-specific accessibility measures into a household-level choice model.

Where is the pedestrian collision zone? Using a uniform grid to assess the risk of pedestrian collisions in King County, Washington

Junfeng Jiao
jialan@u.washington.edu

Slides (pdf)

A growing body of literature has investigated the relationship between built environment, road design and pedestrian collisions over the last two decades. Different variables such as residential population density, the presence of crosswalks, traffic signals, the facility’s number of lanes, speed limit, average daily traffic volume, and the presence of retail uses, restaurants, alcohol stores…had been identified as key variables that affected the occurrence and severity of pedestrian collisions. Researchers have also tried to identify the dangerous collision areas based on predefined geographic zones, such as traffic analysis zone, census tract, and block group, however due to their uneven sizes and shapes, more questions were raised than solved. In order to fill this knowledge gap and develop an area-based risk assessment method for pedestrian collisions, this research employed a uniform grid structure as the spatial unit to analyze the pedestrian collisions that occurred in the King County urban growth boundary from 2001-2004. 

In terms of detailed research design, this was an area based collision analysis, which tried to identify the dangerous collision zones based on aggregated collision and built environment/road design data. The cell size was defined as 500m. This size better fit the Seattle urban structure and provided statistical advantages over the other two different cells (305m and 1000m). The research area covered the whole King County Urban Growth Boundary (KC-UGB), which included over 90% of pedestrian collisions that happened in King County from 2001-2004. The collision data was developed from the original police officer records. Based on the dataset, 2952 collisions in the period were successfully geo-coded, 2944 of them were within the King County urban growth boundary and were included in the analysis. 

After decomposing KC-UGB area by the uniform (500m) grid, 5506 cells were generated. The count of collisions within each cell was used as the dependent variable, which followed a quasi-Poisson distribution. For independent variables, 20 different built environment and road design characteristics, such as the number of bus stops, restaurant, retail, bus ridership, total length of street network, sidewalk, average speed limit, average ADT, average home value, total office, retail area, were measured within each cell. Distances from cell centroids to urban centers and neighborhood centers were also included in the analysis.

Finally, a Poisson model and a Negative Binomial model were used to analyze the data and estimate the occurrence of collisions within each cell. Then the results were compared and a surface map of the risk of collisions within KC-UGB was generated based on the model predictions. Hopefully this research could establish a method to assess the risk of pedestrian collisions within different urban environments and identify the possible collision zones in US