Browsing by Subject "Difference in differences"
Now showing 1 - 2 of 2
- Results Per Page
- Sort Options
Item Characterizing emerging urban transportation modes : models and methods(2020-12-07) Zuniga Garcia, Natalia; Machemehl, Randy B.; Scott, James G; Kockelman, Kara M; Ruiz-Juri, Natalia; Claudel, Christian GThe introduction of emerging transportation technologies, such as mobility-on-demand and shared modes, have caused disruptions in urban transportation systems. These services brought multiple challenges, including the lack of infrastructure, arbitrary pricing schemes, deficient operating rules and regulations, and safety concerns. Furthermore, the deployment of these technologies has increased the need and demand for improved management of the associated data. In particular, the volume of the collected information, the variability of data sources, the heterogeneous structure, and the inherent spatio-temporal nature highlight challenges for finding spatial and temporal relationships, dealing with computational complexity, and for the integration or fusion from various sources. This research work is based on the need for the implementation of models and methods for dealing with large-scale, diverse, and spatio-temporal datasets to adequately characterize emerging mobility technologies and their potential impacts on urban environments. Specifically, it assesses three main points: (1) the impact of emerging mobility modes on urban areas is still unknown, (2) it is not clear what is the effect of shared mobility services on public transit usage, (3) when available, the data may present several challenges. This dissertation designs and applies models and methods to evaluate emerging mobility services' impacts on different aspects of urban areas. The impacts in question are analyzed using four distinctive techniques based on advanced statistics and data analysis models and methods. These techniques are applied to several data sources describing ridesourcing (i.e., ride-hailing via transportation network companies or TNCs), microtransit (i.e., privately owned and operated shared transportation system that can have fixed or flexible routes and schedules), micromobility (e.g., bikesharing and dockless electric scooters or e-scooters), and public transit trips from Austin, Texas. The results of the analyses show that the current fare system and pricing strategies can lead to disparities in TNC driver earnings. Temporal and spatial demand variations can exacerbate search frictions, which can cause an overall market failure. The results suggest that new pricing strategies are required and that there is a need for pricing regulations. A further examination of the ridesourcing effect on the airport ground access using Intelligent Transportation Systems (ITS) showed that the average airport-accessing speed decreases in the presence of TNCs. The use of ITS data is proposed to support airport decision-making processes. Finally, this study analyzed the integration of shared modes with the public transit system. Shared modes can complement the public transportation systems (like bus, train, and air) and solve first-mile-last-mile (FMLM) accessibility issues. However, this study's results suggest that this integration is not yet happening for TNCs and microtransit modes. An analysis of the use of Public-Private Partnerships (PPPs) to introduce share modes in areas with low public transit demand suggests the service was mainly used for intrazonal trips and not for FMLM. Further analysis of the relationship between e-scooters and public transit was able to identify areas with potential e-scooter and bus interaction. The results suggest that future collaborations and PPPs should focus on integrating these mobility services into the public transit system.Item Three essays on risk perception, flooding, and housing market outcomes(2023-05-22) Plough, Julian; Olmstead, Sheila M.; Wrenn, Douglas; Stolp, Chandler; Waxman, AndrewOne of the impacts of a changing climate is the increase in frequency of extreme weather events, many associated with flooding. Flooding incidence can take various forms determined by individual contexts. These events can be highly localized nuisances or geographically ubiquitous, associated with coastal or inland storms, functions of local infrastructure or residential location, and more. Vulnerability to flooding is a product of these physical realities as well as social interactions. Wealth and federal and local administrative systems can provide resiliencies, and informational systems can mediate opportunities for individuals to assess and update their perceptions of risk. Poorer and lower lying localities and the individuals who reside in them often face different sets of constraints than wealthier and physically less-at-risk counterparts. Behavioral responses to flooding incidence are therefore responses to an amalgamation of realities, both constant and changing. Choices related to residential location provide key insight into the decision- making processes that individuals employ in these complex contexts. The policy-making context surrounding flooding and projected climate change revolves around not only projections of changes in physical incidence, but also the assumed behavioral response to that risk. It is crucial to assess now, in response to extant weather events and frequency changes, behavior that lends insight into the degree to which individuals and populations adapt and adjust. Current examples of resilience, vulnerability, and the lines upon which these determinations fall can provide instrumental inputs to guide forward-looking policy decisions. In this dissertation I focus on housing decisions in three major Gulf Coastal cities as a window into the updating and adaptation processes of individuals that scale into shared behavior. In my first paper I estimate the impact of hurricanes in Houston, TX and Tampa, FL on home transaction prices using a causal difference-in-difference strategy. In this paper I find inconsistent causal estimates of the impacts of flooding and flood risk by market and storm. In my second paper I employ a refined measure of flooding events using the National Flood Insurance Program’s redacted claims dataset and recover consistent causal impacts of localized flood damage claims in Houston, TX. In these first two paper I use transaction prices to make market-based welfare inferences, however the decision to sell or to remain in place presents a story of adaption to itself. In my final paper I employ a survival analysis of the home-sale decision post- Hurricane Katrina in New Orleans, LA where I find that income and race are linked to differential sale rates after the experience of the storm. In this dissertation I use three novel datasets, propose repeatable methods for future work, and employ several econometric and statistical lenses to lend insight into behavioral changes in flooding-related housing decisions that can inform policy.