Study: Uber and Lyft Add Traffic, Reduce Efficiency on Denver and Boulder Roads

About a third of people would have walked, biked or taken transit, but they took Lyft and Uber instead. Image: Alejandro Henao
About a third of people would have walked, biked or taken transit, but they took Lyft and Uber instead. Image: Alejandro Henao

Ride-sourcing companies like Uber and Lyft add tons of traffic to Denver and Boulder streets, and make the transportation system less efficient by cannibalizing transit, biking, and walking trips. That’s according to a study by Alejandro Henao, a PhD candidate at the University of Colorado.

Uber and Lyft famously hide their trip data from cities, so Henao went and got the data himself. He became a driver for both companies and surveyed 311 passengers over about four months in Denver, Boulder, and various suburbs. Each passenger answered 28 questions — things like where they were going, how they would’ve traveled otherwise, and demographic information.

“The main goal was to understand how people are using Uber and Lyft — what modes they were replacing,” Henao told Streetsblog. “And also, looking at it as a transportation engineer, I wanted to see how efficient this service is, compared to other services.”

About 34 percent of people surveyed said they would have either taken transit, biked, or walked instead of using the car service. That’s about 106 people who replaced transit, biking, and walking with sitting in a car, which means 106 more cars on the road that weren’t there before.

And here’s the other thing — those people aren’t adding a car to the road solely during their trip, according to Henao’s research. The mileage ticker starts when Henao is hailed by the app, continues through the pick-up, and ends with the drop-off. So if he drives two miles to pick up a passenger, then takes the passenger three miles away, the person in his back seat actually accounts for five miles worth of car traffic.

A bus going that same three miles on a fixed route carries a lot more people in a smaller amount of space, and doesn’t add excess mileage to the streets. But bad transit is also a culprit, according to Henao’s research. If the bus doesn’t come frequently or takes too long to get its destination, people will pay more for the immediacy of a personal car.

“One of the main reasons that these people stated that they’re using Uber and Lyft is because public transportation is unavailable or is poor,” Henao said.

Henao experienced the frustration of extra driving, and used his trip data to measure its effect on the overall transportation system. He examined the group of people who replaced other modes with Uber and Lyft rides, and measured the impact of those trips with Uber and Lyft and without.

With ride-sourcing, it takes 1.6 vehicle miles traveled to move a person one mile. Image: Alejandro Henao
With ride-sourcing, it takes 1.6 vehicle miles traveled (VMT) to move a passenger one person mile (PMT). Image: Alejandro Henao

Using other modes, Henao’s passengers could have moved 112 miles for every 100 miles worth of car traffic burdening the system, Henao found. With Uber and Lyft in the picture, it takes 100 miles of car traffic to move 60.8 “person miles.” In other words, more car trips moved fewer people.

Uber and Lyft “could be part of the solution for better transportation systems,” Henao says. After all, 12.3 percent of Henao’s passengers said they wouldn’t have traveled at all without the car service. They could also help cities treat their expensive and space-hogging parking addiction, Henao said.

But mostly the study is a cautionary tale to cities that think private ride-sourcing gets them off the hook for providing convenient transit. It’s not a silver bullet for mobility

“Transit agencies cancelling bus routes and subsidizing Lyft and Uber rides — you see it happening more and more all over the country,” Henao said. “But we don’t fully know the effects.”

  • TakeFive

    Props to Alejandro Henao!

    Not sure that 106 Uber/Lyft trips over four months reflects “tons” of traffic. The obvious benefit is their ability to create individualized routes easily and quickly which is not something that RTD is even capable of doing. Uber/Lyft could be most productive by adding first mile/last mile ability; but that depends on having transit that appeals to commuters.

    • mattlogan

      Multiply 106 by the number of Uber/lyft drivers in Denver.

      • TakeFive

        … and how many would that be? And what would be the amount of total traffic over four months in Boulder, Denver and the surrounding suburbs as opposed to the Uber/Lyft traffic? Is that spread throughout a 24 hour day or is it mostly at Rush Hour? Feel free to offer whatever intelligent assessment you can.

    • David Sachs

      Full study, which was linked in the article.

  • Roads_Wide_Open

    Doesn’t look like a lot of data points to make sound conclusions, though I see the trends. Would be interesting to see the pick-up/drop-off locations mapped though. If he primarily stayed in the CBD areas, I’d think the 1:1.6 ratio would be closer to 1:1.

  • JZ71

    One major false assumption: “A bus going that same three miles on a fixed route carries a lot more people in a smaller amount of space, and doesn’t add excess mileage to the streets.” That assumes that the bus is operating near capacity / carrying multiple passengers. In many suburban areas, the bus is just a bigger vehicle carrying a handful of riders, so it becomes more of a “wash” . . .

  • mike

    How are households that are no car/car-lite because of the occasional availability of rideshare accounted for? Just because one might have otherwise taken a bus for that specific ride, doesn’t capture all the other trips that person takes by foot/bike/bus because of the availability of rideshare. Anecdotally, here in DC it feels like everyone in the city is dropping to no car or one car households because there are way more options these days than even a few years ago.

    • Asher Of LA

      Great point, and one that’s hard to address with the dataset Henao has.

      Once you have the car, you tend to use it much more – but TNCs can substitute for say, 20% of your car use, with the balance from transit/biking/walking/shared rides.

      I’d contend Uber is simply too expensive for most to replace 100% of your car rides. It would be interesting to study those DC households and see what percent of their travel is by Uber, vs similar but car-owning households and their car VMT.

    • Alejandro

      Good point. This article is in regards to the “specific trip” data. Data on general travel behavior changes (including car-ownership changes) was also collected, which can be found in the full dissertation….stay tune for more papers about this. You can find more info at

  • SingleOccupantDriver

    Congestion and air pollution is mostly caused by the width design choice of vehicle makers.

    I propose large cities work with Uber, Lyft, FedEx, Amazon, UPS, taxi companies, and their public transit groups to build and lease single-width, 100% electric narrow cars and buses so that the trips to pick up passengers don’t require twice the road-width.

    Bicycles wouldn’t be as great for cities if they had side seats, which some novelty bikes do. I can’t imagine a single person thinking cities allowing car or bus width twice as wide as they are currently to be a good idea. The same logic applies absolutely to side by side seated vehicle design.

    It’s time to design and build single-width road vehicles to fix congestion and improve air quality in cities.

  • joerubino

    Regardless of what you think about this study, there is no doubt that Uber and Lyft have created incredible unintended consequences re: traffic congestion and pollution. NYC streets used to seem like a sea of yellow with their 13,000 taxis. Now, added to that group, there are 65,000 Uber cars. That is not an exaggeration. San Fran has 1800 taxis, which always seemed like a lot in that small (45 square mile) city, yet there are now 45,000 Ubers on San Francisco streets. Think of that number for a second. Uber was supposed to REDUCE cars on our streets, not add to that number exponentially. These additional numbers of vehicles on city streets are having a deleterious affect on the act of simply walking around either city. To say nothing of the increased pollution. Visit either of those cities to see for yourself if you doubt my word on this. Regular drivers in San Fran can barely get through an intersection without dodging a ride-hail car heading to or from a pickup.

  • hibbidydibbidy

    If they were truly “ride sharing” then it would reduce traffic because we’d be sharing a ride to a nearby destination.

    However the driver of the car has to drive from where they are to where you are, then go from there to your destination, and then finally go to where they want to be. This adds 2 additional legs added to the trip compared to you just driving yourself.

    Sure, a taxi would do a similar amount of driving but nobody claims that taxis cut down on the number of miles driven.

  • JulieMcCabe7

    Hi Alejandro, I attended 10/18/17 transport meet in Boulder. You did a great job! Too bad everyone in Boulder can’t take a pill and download all the points made by the panel. You equate cost transportation in vmt vs pmt. Another way of saying this is $ efficiency, cost/benefit analysis based upon a graph showing: parking, cost, GHG emissions, congestion, etc. This will show the variables of choice in making a decision: it’s windy and cold, call Lyft b/c the bus stop sucks, or there’ll be no parking, etc. My bet, few are so politicized that they take mass transit solely because of GHG issues. GHG is a big picture issue people think about, on which gov.s act. Lauren (think ?) said future mobility will incorporate mass transit. My BIG complaint about city mass transit – it runs nearly empty except for routes for students and CU. (CU’s on campus parking is limited and expensive.) Our “non-profit” politicians default to more mass transit to combat Boulder’s traffic woes. I don’t agree highly subsidized mass transit (i.e. too few riders) is worth the cost. High subsidy = NOT contributing to any solution. The RTD saga in Denver with its light rail costs is going to be a headache until RTD figures out how to incorporate its systems with the ones discussed last night. Have a good day. JMc


The UN secretary general, António Guterres, addresses the closing meeting of the 73rd session of the United Nations general assembly at the UN headquarters on 16 September. Photograph: Xinhua/Rex/Shutterstock

‘We’re Losing the Race’: UN Secretary General Calls Climate Change an ‘Emergency’

This story originally appeared in The Guardian. It is republished here as part of Streetsblog’s partnership with Covering Climate Now, a global collaboration of more than 300 news outlets to strengthen coverage of the climate story. The UN secretary general says that he is counting on public pressure to compel governments to take much stronger action against what […]

Car-Free Zones Eyed in SF, Elsewhere

A string of pedestrian injuries and deaths in San Francisco’s Tenderloin district has spurred city leaders to demand a ban on cars in some densely populated neighborhoods — the latest in a nascent and long-overdue move by activists nationwide to get reckless drivers off at least a tiny handful of city streets. San Francisco Supervisor […]
A Comcast Xfinity van tips over within eyeshot of where cyclist Alexis Bounds was killed July 24. Photo: Anonymous

Wednesday’s Headlines

Dangerous place: A photo shows an Xfinity van rolling over near the ghost bike placed in remembrance of Alexis Bounds. Input sought for Front Range Passenger Rail. The five options CDOT is considering to upgrade I-25 in Denver. More headlines ...