Just a better taxi? A survey-based comparison of taxis, transit, and ridesourcing services in San Francisco.
This study presents exploratory evidence on the use of app-based, on-demand ridesharing services in San Francisco, California. In May and June 2014, 380 intercept surveys were collected from three ridesharing "hot spots" in San Francisco: The Marina, North Beach, and the Mission. Respondents either just completed a ridesharing trip or had taken a ridesharing trip in the prior two weeks. The survey asked 18 questions regarding trip origin and destination, trip purpose, previous and alternative model choice, car ownership, and basic demographics. The survey results were compared with matched-pair taxi trip data gathered from the results of a prior taxi use survey and a survey of taxi users conducted for the San Francisco Municipal Transportation Agency. Travel times for ridesharing and taxis were also compared with those for public transit.
Despite many similarities, taxis and ridesharing differ in user characteristics, wait times, and trips served. While ridesharing trips replaced taxi trips, at least half of ridesharing trips replaced other modes, including public transit and driving alone.
When asked why they chose ridesharing, variations on speed and convenience were the main attractions to the survey respondents, another common reason was avoiding drinking and driving (20 percent). Only 2 percent said they could not get a taxi and only 6 percent said public transit was unavailable.
The alternative mode heavily influenced responses. Those choosing ridesharing over driving themselves sought to avoid parking or driving after drinking. Those choosing ridesharing over taxi called out ease of payment, short wait time, and ease of calling a car. Those choosing ridesharing over transit wanted to arrive at their destinations faster and spend less time waiting.
Ridesharing wait times are shorter than typical taxi dispatch and hail times. At off peak times only 35 percent of survey respondents indicated that if they called a taxi it would be at their home within 10 minutes, this dropped to 16 percent during peak periods. Conversely, nearly 90 percent of survey respondents had never waited more than 10 minutes for a ridesharing car and 67 percent waited five minutes or less.
When asked if they still would have made the trip had ridesharing not been available, 92 percent of respondents confirmed that they would.
Estimated total travel times, including wait and in-vehicle times, were consistently greater for public transit than ridesharing, although a few trips would have been faster by transit. The estimated average total travel time was 22 minutes for ridesharing trips, while the same trips would have taken on average 33 minutes by public transit; a typical ridesharing trip saves about 10 minutes of travel time. Overall, 66 percent of ridesharing trips would have been at least twice as long in minutes, if taken by public transit.