Skip to main content

Lyft, Uber have mixed impact on San Fran mobility

The extent to which ride-hailing has become a real force in the mobility landscape of San Francisco is great for consumers – but there are downsides, a report finds. Andrew Stone takes a look. Uber and Lyft, the two major ride-hailing platforms in San Francisco, are out-competing local cab firms in many ways - and are firmly established as a significant part of the daily mobility mix there, a recent study reveals. Researchers mined publicly-available data derived from the application programming interface
May 14, 2018 Read time: 7 mins

The extent to which ride-hailing has become a real force in the mobility landscape of San Francisco is great for consumers – but there are downsides, a report finds. Andrew Stone takes a look.

8336 Uber and 8789 Lyft, the two major ride-hailing platforms in San Francisco, are out-competing local cab firms in many ways - and are firmly established as a significant part of the daily mobility mix there, a recent study reveals. Researchers mined publicly-available data derived from the application programming interfaces (APIs) of the two transportation network companies (TNCs) and found they are meeting a strong demand among the travelling public.

The report, called TNCs Today: A Profile of San Francisco Transportation Network Company Activity, was commissioned by the 1798 San Francisco County Transportation Authority to study ride-hailing services’ impact on congestion, public transit ridership and other urban mobility metrics.

“The success of TNCs in attracting rides in San Francisco reflects the high unmet demand for premium services and the extensive benefits they provide to users who can afford their services,” says the report.

Congestion downside

There may be downsides to their popularity, however. Beyond the scope of the study but posing topics for further research, its findings suggest that ride-hailing platforms are likely to be adding to congestion as well as local and global pollution. They may also be impeding other transport modes, particularly at peak times and in central downtown areas, owing to their popularity and ubiquity. The study looked at both daily vehicle and person trips as measures of the reach and impact of TNCs in San Francisco’s streets and found that approximately 170,000 TNC vehicle trips happen during a typical weekday, which accounts for approximately 15% of all weekday vehicle trips. This amounts to around 12 times as many TNC trips as taxi trips during a typical weekday.

“The number of TNC vehicle trips is important because more vehicle trips generally lead to increased congestion and conflict with other street users, while more person trips may indicate enhanced mobility,” says the report. “Person trips are different from vehicle trips because person trips include walking and biking trips, and also because private vehicles, public transit vehicles and taxis may carry more than one person.”

The report estimates that approximately 290,000 TNC person trips happen within San Francisco during a typical weekday – which is approximately 9% of all weekday person trips in the city. Ride-hailing trips increase as the week progresses, reaching their peak volume on Friday (222,500 trips) and hit their lowest volume on Sundays (129,000 trips).

Monday to Friday peak

Highest usage of TNC services coincides with existing morning and afternoon peak periods. “The timing of TNC trips is important because trips that occur during peak periods and weekdays are more likely to exacerbate congestion and delay on roads, affecting both general traffic, surface public transit as well as conflicts with bicycles and pedestrians,” says the report.

As well as being concentrated in peak times, the report also found that the vast majority of TNC trips occur in San Francisco’s north-east quadrant, containing the most congested areas of the city that are already well served by public transit, cycling and walking.

This includes the downtown and north-eastern core of the city, where many of the most significant employment and public transit hubs are located. At peak periods, the report estimates that TNCs are estimated to comprise 25% of vehicle trips in South of Market. TNC trips are concentrated on the busiest arterials, yet also operate extensively on neighbourhood streets, including along major public transit lines.

San Francisco ride-hailing in numbers

•  45,000 estimated Uber and Lyft drivers operating in San Francisco

•  5,700+ Ride-hailing vehicles on San Francisco streets at workday peaks

•  170,000 daily ride-hailing vehicle trips made within San Francisco on weekdays

•  12-15 times daily ride-hailing versus taxi trips

•  15% estimated share of all intra-San Francisco vehicle trips
9% estimated share of all San Francisco person trips

“The location of TNC trips is important because trips that occur where there is already significant traffic are more likely to exacerbate congestion and conflicts with other road users,” says the report, “while trips that occur in less congested areas may reflect lower transportation impacts.”


The study also found that TNCs provide a broader service across the city than taxis, particularly in the western neighbourhoods but that they provide fewer trips relative to population and employment in the southern and south-eastern areas of the city.

“This may reflect either a lack of vehicles available in this area, or may reflect inability of residents of these areas to use TNCs, or some combination of these or other factors. Additional data is required to better understand this pattern,” suggests the report.

Vehicle miles travelled

The researchers measured vehicle miles travelled (VMT) by the ride-hailing vehicles, both in and out of service (essentially with or without passengers).

The study estimated that TNCs are conservatively covering 570,000 VMT on a typical weekday, accounting for as much as 20% on weekday VMT for intra-SF vehicle trips, at least 6.5% of total weekday VMT in San Francisco and approximately 10% of VMT on Saturdays.

“The vehicle miles travelled by TNCs is important because VMT is a fundamental measure of transportation system performance,” says the report. “Higher levels of VMT are associated with greater levels of emissions of greenhouse gases such as CO2 as well as other pollutants. In addition, higher levels of VMT are also associated with greater roadway congestion and conflicts.”

The study also found that TNCs generate more than 10 times as many VMT as taxis on a typical weekday, while generating 12 times as many trips.

TNCs vs. taxis

The TNCs appear to be more efficient than taxis by some metrics. Approximately 20% of total TNC VMT are out-of-service miles, significantly lower than the more than 40% of taxi VMT that are out-of-service miles.

“The greater efficiencies of TNCs, as reflected in a lower share of out-of-service miles, are likely primarily a reflection of the larger fleets of TNC drivers circling at any given time, enabling shorter distances to pick-up locations. In addition, TNCs’ routing software may be more efficient than the taxi dispatch systems.”

The report notes that TNCs have an advantage over taxis not only owing to their better exploitation of technology but also due to the relatively light regulatory requirements under which they operate, compared to taxis and other for-hire vehicles.

“In San Francisco, the taxi medallion system limits the number of taxi vehicles that can serve the city, taxis are subject to price controls, must provide access to all areas of the city, must provide service to people with disabilities, have greater insurance requirements, and are subject to driver background checks and vehicle inspections.”

In the light of the report, the transportation authority plans further study of the range of questions the report raises - including the scope of regulation of TNCs, their impact on street safety, transit demand and operation, congestion, access for disabled people, access for different socio-economic and geographic communities, as well as impacts on land use and kerb management.

Methodology and data sources

The report is an estimate of local TNC usage (trips made entirely within San Francisco) from mid-November to mid-December 2016, based on data gathered by researchers at Northeastern University from the Uber and Lyft public-facing APIs.

The researchers collected information on TNC vehicle locations every five seconds for approximately six weeks. The team cleaned the location data by removing unnecessary, anomalous or redundant information - then it identified trips and imputed missing attributes. For example, individual trips were imputed by linking the pick-up and trip drop-off locations.
Deriving data from the APIs has its limitations, the report admits. For example, pooled services such as UberPOOL and LyftLine, which are designed to encourage users to share rides, may not show up in the datastream.

The methodology also fails to estimate actual vehicle occupancy or traveller demographics. The study’s dataset did not include trips beginning or ending outside the city, so some of the given estimates will be conservative as a result.

The transportation authority website makes available aggregate travel analysis zone (TAZ) level summaries of pick-ups and drop-offs by hour of day (%$Linker: 2 External <?xml version="1.0" encoding="utf-16"?><dictionary /> 0 0 0 link-external www.sfcta.org/tncstoday sfcta website link false http://www.sfcta.org/tncstoday false false%>) and an interactive visualisation of the TAZ-level data at %$Linker: 2 External <?xml version="1.0" encoding="utf-16"?><dictionary /> 0 0 0 link-external http://tncstoday.sfcta.org tncstoday website link false http://tncstoday.sfcta.org/ false false%>.

Related Content

  • October 17, 2019
    Waymo may operate AVs in Phoenix ‘without safety driver’
    Ride-hailing company Waymo may be about to start operating fully-autonomous vehicles (AVs) to pick people up - without a safety driver. An email sent to users, which appeared on Reddit, said people in Phoenix, Arizona, who were matched with an AV will see a notification in the app that confirms the car will not have a trained driver. Users can tap a ‘What to Expect’ button within the app to learn more about the AVs. They can also communicate with a rider support agent at any part of the trip via the app o
  • October 12, 2017
    ITF study: shared mobility can cut congestion and CO2 emissions
    A new ITF mobility study in Helsinki confirms that a combination of 6-seater shared taxis and taxi buses can carry out all of today’s car journeys in the city area with just 4% of the current number of privately owned vehicles. These shared mobility platforms also mean fewer changes, less waiting and shorter travel times compared to traditional public transport. In addition, the results confirm improved access to jobs and public services, most notably for citizens in areas with few such offers.
  • August 2, 2018
    UK government gets future mobility challenge underway
    The UK government has unveiled plans under its Future of Mobility Grand Challenge which could change how people, goods and services move around the country. These initiatives have been outlined in the Last Mile and Future of mobility call for evidence, which provide an insight into how technology could make transport safer, more accessible and greener. Under the plans, electric cargo bikes, vans, quadricycles and micro vehicles could replace vans in UK cities as part of a strategy to change last-mile
  • May 4, 2018
    Atlanta conference drives on-demand transport agenda
    The US city of Atlanta is the latest major urban area to consider how integrated on-demand transportation services could help ease congestion and reduce pollution – and boost the regional economy. The subject will top the agenda at next week’s conference on Mobility as a Service (MaaS) which takes place on May 9 and 10 in the city. The conference, called MaaS Market - Concept to Delivery, has attracted leading international experts from Europe and across the US and is supported by the City of Atlan