摘要: |
Americans in urban areas now have access to the services of transportation network companies (TNCs), such as Uber and Lyft. TNCs have reduced the costs of taxi use through lowered fares and apps that make the process of ordering and waiting for a taxi easier and more reliable. The result has been rapid growth in trips made by Uber and Lyft. Given expectations that TNCs will continue to reshape Americans� travel patterns, it is important to understand the impacts of these shared mobility services on road safety.
TNCs claim that shared mobility improves safety by reducing impaired driving. Reductions in driver impairment could provide large safety benefits since impairment was a factor in 29% of fatalities on American roads in 2015. A 2012 study funded by Uber found the service reduced drunk driving crashes among young adults in California. Analyses in New York City showed a 25-35% decrease in alcohol-related collisions after the introduction of Uber. However, increasing use of TNCs could have deleterious effects on road safety. Shared mobility services reduce the costs of taxi use and therefore may encourage travelers to choose TNCs rather than walking, biking, or taking transit. They may also induce new trips which otherwise would not have occurred. Economists have recently found that TNCs affected transit use. In urban areas with strong transit systems, Uber acted as a complement to the transit system effectively serving as a flexible, on-demand first-mile/last-mile service. But in less dense areas, TNCs were a substitute for transit.
This study will analyze how TNCs have influenced roadway fatalities in urban areas with results disaggregated by time period and geography. The approach will account for market penetration of TNCs in each metropolitan area, secular trends in driving, and the differential effects of TNCs on other modes such as transit. The limited existing work on the topic has failed to account for these critical factors.
Methods: (1) Data collection: The safety outcome of interest is fatalities related to roadway crashes and will be derived from the FARS data from 2003 to 2016. The FARS data contains information on crashes and drivers which will allow them to identify fatalities related to impairment. Data on TNC service will be accessed from date of entry into the market and Google Trends data which has been demonstrated to correlate with market penetration. Information on secular trends in modal patterns will be derived from the American Time Use Survey (2003-2016) and the National Transit Database. Information on the metropolitan area will come from the US Census Bureau. The research team will also analyze alcohol-related crashes to determine the patterns for this outcome variable. (2) Analysis: The team will utilize a difference-in-differences (DID) research design that will allow them to distinguish how fatalities changed in cities when shared mobility services entered relative to changes where shared mobility had not yet entered the market. (3) Project outcomes: This project will identify the conditions under which shared mobility services do or do not provide safety benefits. These findings can be used to offer policymakers guidance on how the evolution and increased market penetration may provide safety benefits and how those benefits can be maximized. |