If there’s one thing that transit network companies (TNCs) like Lyft and Uber are consistently good at, it’s inspiring academic research and civic investigations about tech-enabled traffic congestion. Which is not their business model, unfortunately.
In the last two years alone, San Francisco has seen a pile-up of peer-reviewed results peering at the question of just what effect these companies have on traffic and other driving-related hazards.
Almost everyone agrees that ride-hailing apps create a tremendous amount of congestion in SF specifically—even the companies themselves often cop to this in the comments below, while at the same time arguing that their ultimate goal is to minimize the number of cars on the street.
But the details vary wildly, along with the methods. Here’s a brief survey of the surveys on this eternally idling question.
August 2019
Both Lyft and Uber jointly commissioned transit consulting company Fehr and Peers to study the effects of ride-hailing apps on traffic in the metro areas in and around six major cities: San Francisco, Los Angeles, Seattle, Chicago, Boston, and Washington DC.
This was the first time anybody parsing the TNC traffic question had direct access to either company’s data. Although the results specifically focus on September 2018, they’re nevertheless the most up-to-date findings available.
The results:
- In all, Fehr and Peers determined that between 12 and 14 percent of all vehicle miles traveled (VMT) in SF in September were generated by Lyft and Uber services, whereas all other vehicle activity accounted for 86 to 88 percent of total VMT.
- Overall, SF saw an estimated 259,461,000 VMT over the course of September, meaning that, between the two companies, Lyft and Uber drivers clocked 36,324,540 driving miles in San Francisco, more than 1.2 million miles per day.
- For comparison, in Washington DC ride companies accounted for six to seven percent of traffic, and in LA only two to three percent. Seattle had the lowest overall rate with just 1.5 to two percent. San Francisco ranked the highest.
- The report contextualizes the results by noting that “San Francisco has a lower rate of car ownership compared to the rest of the Bay Area, as well as a robust internal transit system,” meaning that “the higher share of VMT potentially associated with Lyft and Uber may reflect lower overall rates of driving and higher transit rates.”
- In response to the study, Lyft’s head of policy research Peter Day said, “We look forward to [...] remaking our cities designed for people, not cars.” Uber’s head of global policy for public transportation Chris Pangilinan noted, “Our goal is to reduce the need for private car ownership.”
May 2019
The University of Kentucky’s college of engineering published its research about TNC traffic in the peer-reviewed journal Science Advances, telling Curbed SF that much of the data was presented to the city in late 2018—and included in a San Francisco County Transportation Authority study—but that the results had now been subject to peer review and the two reports draw from some different data sets
- Researchers relied on “an activity-based travel demand microsimulation model” to predict what driving conditions in San Francisco probably would have looked like between 2010 and 2016 without Lyft and Uber, then compared those results to what actually happened.
- According to the model, over six years of congestion in San Francisco rose 40 percent compared to what it might have been without Lyft and Uber. Also of note, “average speed decreases from 25.6 mph in 2010 to 22.2 mph in 2016, and the vehicle hours of delay increase by 63 percent over the same period.”
- Assistant professor Gregory Erhardt added that “the current system is commonly viewed as a bridge technology that may be replaced by fleets of self-driving cars” in the future, suggesting that the industry views the current TNC system as something of a beta version of the real thing.
- A Lyft spokesperson dismissed the data as “old” and called Erhardt’s models incomplete, while Uber touted its lobbying efforts for congestion pricing as a solution to traffic woes.
January 2019
In previously published University of Kentucky research, TNCs had a measurable and decisively negative effect on public transit use:
- “Our results suggest that for each year after TNCs enter a market, heavy rail ridership can be expected to decrease by 1.3 percent and bus ridership can be expected to decrease by 1.7 percent.”
- “In a market like San Francisco, where Uber started operations in 2010, the model implies that we would expect a 12.7 percent decrease in bus ridership, all else being equal. [...] SFMTA would need to increase bus service by slightly more than 25 percent in order to offset the loss of bus ridership to TNCs.”
- The big exception, perhaps surprisingly, is light rail use. Researchers noted that “the light rail coefficient is also negative,” but is [so small that it’s] insignificant.”
October 2018
A University of Chicago paper titled “The Cost of Conveniences” pointed the finger at ride-hailing apps for spikes in traffic collisions in some major cities:
- “When we separate the accidents into those that do and do not involve a drunk driver, we find that the estimates for non-drunk accidents are similar: a 2-4 percent increase in accidents, across a variety of measures” when ride-hailing apps are in use in a city.
- While the paper acknowledges that services like Lyft and Uber can potentially decrease the number of cars on the road, in practice this effect is not observed: “Despite allowing for more utilization of carpools, and therefore potentially reducing total vehicle miles traveled, the introduction of Uber Pool and Lyft Line do not reverse the documented increase in fatal accidents.”
- Since the effect is small and the timeline short, things could change, as “we note that the documented effects may be short-term, as pooling services such as Lyft Line and Uber Pool increase ridership.”
- City observatory director Joe Cortright was skeptical about the findings, noting that variables like lower gas prices can drive up accident rates and that the study leaves out the effect of lower gas prices and increased driving on crash rates. He also points out that “rural areas–which essentially don’t have ride-hailing services–saw even bigger increases in crashes” over the same period.
Meanwhile, San Francisco City Hall’s own “TNCs and Congestion” report cited “a unique TNC trip dataset provided to the Transportation Authority by researchers from Northeastern University in late 2016” on top of “INRIX data, a commercial dataset which combines several real-time GPS monitoring sources with data from highway performance monitoring” to support its argument that Lyft and Uber drove 50 percent of recent congestion:
- “When compared to employment and population growth and network capacity shifts (such as for a bus or bicycle lane), TNCs accounted for approximately 50 percent of the change in congestion in San Francisco between 2010 and 2016.”
- “TNC passenger pick up and drop of activity may also result in increased congestion by disturbing the flow in curb lanes or traffic lanes. Out-of-service miles resulting from TNCs repositioning themselves to more optimal locations for getting new passengers, or from driving to pick up passengers who have reserved rides [...] also increases the amount of vehicular traffic.”
- Overall, since 2009 “vehicle hours of delay on the major roadways increased by 40,000 hours on a typical weekday, while vehicle miles travelled on major roadways increased by over 630,000 miles.”
- Uber blamed factors like increased tourism for running up congestion. Lyft called the study “flawed and incomplete.”
July 2018
Transit consultant Bruce Schaller’s “Automobility” study alleged that Lyft and Uber pushed traffic up 180 percent in major cities:
- “Private-ride TNC services (UberX, Lyft) put 2.8 new TNC vehicle miles on the road for each mile of personal driving removed, for an overall 180 percent increase in driving on city streets.”
- “Lyft’s recently announced goal of 50 percent of rides being shared by 2022 would produce 2.2 TNC miles being added to city streets for each personal auto mile taken off the road.”
- “Traffic could increase by approximately 50 percent if travelers favored autonomous ‘regular taxis’ that are not shared. On the other hand, the model showed a 37 percent decline in vehicle kilometers, and total elimination of congestion, under a shared-taxi scenario.”
- Zipcar CEO Robin Chase said that car culture is the real culprit—again, an unexpected bid from a car-based company. Lyft dismissed Schaller as a “taxicab consultant” and pointed out that traffic in SF was allegedly down at the time.
September 2017
The San Francisco Police Department complained to the Board of Supervisors that TNC drivers committed 64 percent of downtown traffic crimes over a two-month period:
- Commander Robert O’Sullivan told city lawmakers that on a dozen targeted days between April 1 and June 30, 2017, SFPD recorded 2,656 transit violations in SoMa, FiDi, and parts of the Mission, of which 1,723 occurred on account of TNC drivers.
- The overwhelming majority of problems stemmed from drivers straying into transit-only lanes. Of 1,715 such violations, 1,144 were TNCs. This one error accounts for more than two-thirds of TNC-related problems.
- In fairness, the identification of TNC drivers came from the Lyft and Uber stickers on their cars, meaning that any drivers committing traffic crimes on their own time ended up included in the tally as well.
Additional studies—all of them negative—go back to 2016, but the data featured in those is probably too out of date to consult again now.
Loading comments...