Study: Uber and Lyft Caused U.S. Transit Decline

Photo:  Khol Vinh
Photo: Khol Vinh

Uber and Lyft are still crushing transit across the U.S, according to new study examining the effects in 22 cities.

When Uber and Lyft enter a city, the app-based taxis decrease rail ridership by 1.29 percent per year and decrease bus ridership by 1.7 percent, the study by three University of Kentucky researchers found [PDF].

Worse, the effect is cumulative. Authors Michael Graehler, Richard Mucci and Gregory Erhardt estimate that Uber and Lyft for example, have reduced bus ridership in San Francisco a staggering 12.7 percent since they entered the market in 2010.

This isn’t the first study to make such findings, but it is one of the broadest, helping to explain why transit ridership has declined in almost every U.S. city over recent years. These declines could not be explained by service reductions or by maintenance issues — although those have been an issue in Washington and New York City.

Increased service could counter the trend, but it would not be enough to make up for the damage. Graehler and his co-authors estimated that San Francisco would have to increase bus service by 25 percent to offset the effects of Uber and Lyft in depressing ridership.

The study comes as transit agencies like L.A. Metro are grappling with unexplained ridership drops. The system’s ridership declined 3.4 percent last year, despite major investments in the rail network.

But the declines have occurred in almost every major city — save Seattle — and they track closely with the rise of Uber and Lyft. In New York City, Graehler points out, daily Uber and Lyft trips grew from 60,000 to 600,000 from 2015 to 2018. That’s almost identical to the decline in daily transit boardings in New York: 580,000.

“This makes sense because [Uber/Lyft] trips are short and concentrated in city centers, which are the exact same places bus ridership is highest,” Erhardt told Streetsblog. “What appears to happen is that travelers divert from transit to [taxis], increasing congestion for everyone, including the buses.”

The study also found bike share has an effect on transit ridership — although the relationship was a little more complex. The study found bike share increased rail ridership, but decreased bus ridership by 1.7 percent.

The study did find that Uber and Lyft increased ridership slightly on one kind of transit: Commuter rail, but the effect was insignificant.

“Our results suggest that the recent decline in transit ridership in major U.S. cities cannot be attributed to transit service cuts alone,” said Erhardt. “The ridership decline is steepest from 2015 onwards, and correlates with the introduction of Uber into a market.”

Transit agencies may not have the resources to increase service at the scale needed to offset the effects of Uber and Lyft. On average they would need a 20-percent service increase, Erhardt said.

“A more effective approach may be to think beyond the bounds of transit agencies to consider policies such as congestion pricing, or a broader re-thinking of how right-of-way is allocated in the urban streetscape between cars, transit, bicycles, pedestrians and other uses,” he said.

123 thoughts on Study: Uber and Lyft Caused U.S. Transit Decline

  1. They were coming to SF “anyway” to work as Ubers because as I said taxi service before Uber came along was inadequate and Uber drivers are needed in the city.

    Deadheading has impacts, but so does people driving their own cars to work, adding to congestion on the bridge, and then their cars need space to park in the city for 8+ hours. If they drive their cars and park them, the Uber cars are still coming into the city to work as well, meaning more cars total on the freeways, bridges, and in the city either parked or driving.

  2. From Market Street in SF you can go pretty much anywhere in the city on transit, so why were you waiting for a taxi? In that 30 minute wait, you could have arrived at your destination.

  3. “Cars did not become the dominant form of transportation by some mysterious conspiracy.”

    Actually…yes, they did.

  4. All public transit should be free to board & ride. Just as we allow private vehicles to freely access public streets. Would probly benefit from some rejiggering of the typical state/local budgets. All it takes is political will.

  5. Dear Angie Schmitt and Streetsblog: in every piece you write on Uber and Lyft, you need to always remind readers of the fact that these companies are not profitable, and that is why they are able to offer such cheap fares that easily compete against public transit and taxis. Operating for several years below cost, which somehow seems to be legal in this country, is a bet by investors that they will be able to put most bus service and taxi companies completely out of business, at which point they will then increase their fares fivefold.

  6. Exactly. The article only very briefly mentions that rail transit ridership has increased as a result of ride-sharing. Ride sharing is particularly good about getting cars of the road because it is highly useful for the ‘last mile’ transit connections.

  7. Again, you are inventing a problem that does not exist. Ride sharing gets individuals out of driving alone in their separate cars so they can carpool where they need to go with multiple folks in a single car.

  8. False. Next time you need to get somewhere, try either picking up passenger with the Uber driver app or riding with another person using the Uber rider app. You will understand better how ride sharing/ carpooling works.

  9. That is inaccurate. 94% of Uber drivers use the app on fewer than 10 rides a month, generally as a way to reduce their carbon footprint and to access carpool lanes.

  10. False. By allowing many folks to sell or never buy a car, by dramatically increasing carpooling, and by reducing driving alone in cars, Uber and other types of carpooling have dramatically reducing congestion, parking shortages, transit overcrowding, stress, obesity, climate change, inefficiency, income inequality, and air pollution.

  11. Obviously false. Having 2 people carpooling in a ride share vehicle, takes at least one car off of the road, reducing congestion that hurts buses and other vehicles.

  12. It’s not false because the majority aren’t pooling and these companies have made congestion, climate change, inefficiency, income inequality, and air pollution worse .EberyEvery I ommitted I largely agree with .

  13. This article should be retitled “Lack of proper road pricing caused US transit decline”, since rideshares are smartly and fairly using the underpriced roadways.

  14. Exactly. Ride-sharing is by definition reducing folks driving alone and optimizing an under-used resource. Why drive alone downtown when you can ride with someone else going the same way or pick up someone else going the same direction? Ride-sharing is particularly good for reducing congestion overall and for the ‘last mile’ part of a commute or trip. I.e., if a bus or train does not go near a destination, a commuter can catch a ride via Uber. That way, the commuter (and at least one other commuter – the driver) does not have to drive his personal car to the destination.

  15. Yes, and Uber and ride-sharing/ carpooling are an excellent way to assist with demand management. Uber also has introduced features to notify users of transit options that also go to their destination.

  16. Factually inaccurate. Uber/ ride-sharing/ carpooling is dramatically better than having multiple people driving alone in multiple cars.

  17. TNCs assist with demand management by congesting city streets, delaying transit and making it less attractive to riders.

  18. Wrong. Before Uber, If Person A and Person B both had to get to Destination, then both drove their personal cars. Now, they ride together and reduces car usage, congestion, pollution, and parking by at least 50%. Bother Person A and B can be either a driver or passenger as ride-sharing makes every single person able to be both, generally choosing to be a driver when one’s own car is absolutely needed for some further leg/ errand.

  19. Buses are preferable to Uber/ carpooling, but ride-sharing/ carpooling is much better than driving alone in one’s own gasoline-powered vehicle. Ride-sharing also makes bus usage possible for more people by A) reducing overcongestion on buses and B) allowing travel to destinations not served by buses. Ride-sharing is great for the ‘last mile’ part of a trip (i.e., ride-sharing from the nearest train or bus stop. One reason most people do not use transit is the lack of last mile service (completely and exclusively solved by ride-sharing).

  20. Wrong. All of the data strongly support the facts and description I provided. Just to give you one statistic, over 94% of Uber drivers carry less than 10 passengers a month. Still, that is at least 10 cars taken off of the road. That actually underestimates the benefits though. When someone catches a ride with a neighbor through ride-sharing to get to work or run an errand, he is not bringing his car downtown and thus will not need to take up a parking spot and also will not have his car with him to drive for other errands during the day. Hence, Uber does not just reduce car usage for an individual trip, but throughout the day/ night. As a bicyclist, an Uber passenger, and an Uber driver, I also have first hand experience and no well that of which I speak.

  21. How about cutting back funding for (over prices inefficient public transportation) and giving N.Y C riders a voucher to use the better transportation
    Of uber and lyft, way better bang for the tax buck!

Leave a Reply

Your email address will not be published. Required fields are marked *

ALSO ON STREETSBLOG