The average Uber or Lyft customer is responsible for adding more than twice as many car miles to his city's roads as he was before he started using app-taxis to get around, a new study finds — adding another stone to the growing mountain of evidence that "rideshare" companies are clogging city streets, rather than purging them of privately owned vehicles by getting people to carpool or connecting them to transit.
In a study of four of the largest metros in the U.S. (and a handful of California suburbs in which app-taxis are particularly prevalent), analysts at Schaller Consulting sought to quantify exactly how many vehicle miles travelled that e-taxi customers collectively add to our roads, when compared to the vehicle miles that those same residents would have been responsible for if Uber or Lyft weren't around.
The study isn't just an accounting of all the miles city residents actually spent in the back of a stranger's DIY-cab; researchers also factored in all the miles that drivers spent circling or heading to their pickups — aka dead-heading — which accounts for as much as 48 percent of the average e-taxi driver's total mileage in some cities, according to the report.
On the flip side, if all the e-taxi customers in a given city collectively chose shared "pool" options a lot — as the most optimistic proponents of rideshare once dreamed everyone would — that would reduce their collective VMT impact. (In cities where residents tended to choose e-taxi trips over taking the subway, that would be reflected in the city's mileage comparison, too.)
The results? In every region in the study, the customers of transportation network companies at least doubled their total vehicle miles travelled compared to the modes they told researchers they would have taken had Uber and Lyft not been around. Schaller found collective VMT increases of a whopping 97 percent in Chicago, 114 percent in New York City, 118 percent in San Francisco, 157 percent in Boston, and 118 percent in the California suburbs.
"Over the last roughly half dozen years, Uber and Lyft have made a lot of claims about their benefits, not just to individual users, but to cities as a whole — and specifically, they claimed they'd reduce VMT as shared trips became more common," said study author Bruce Schaller. "And when that didn't happen, there were still a lot of 'what about' questions: What about when rideshare is being used for the first mile and last mile [of a trip otherwise taken on transit]? What about people giving up their car and using rideshare instead?
"We can keep asking those questions, but this study shows it's never going to pencil out," he concluded.
The news that "transportation network companies" are increasing congestion, emissions, and even private vehicle ownership itself is, well, not exactly news. But the Schaller study is unique in its efforts to quantify exactly how preposterous it is to treat the app-based cabs as a path-breaking new mobility mode that could work in concert with transit agencies to get privately owned cars off the road. In fact, they have proven to be, at least in the largest cities where they've most taken root, merely digital taxi companies that leverage artificially low prices to take passengers off of transit and put them on the least sustainable mode instead.
And before you wonder if some magical mix of shared trips and reductions in dead-heading could somehow transform app-taxis into a green option: don't bet on it.
Schaller found that, on average, in order for rideshare companies to simply match — not even reduce — the VMT for which their riders were responsible before they started app-hailing, the company would need to accomplish five feats, all of which are virtually impossible:
- Incentivize riders to share 85 percent of their trips with other passengers (pre-pandemic, in the city with the highest share of pooled trips, it was only a third — and in the lowest, it was just one-eighth),
- Incentivize riders to pool with at least two other passengers for most of their trips (even pre-pandemic, most of them only shared with one other person, if they shared at all),
- Attract 70 percent of their customers from other taxi companies, or from drivers who left a private vehicle at home (pre-pandemic, only "one half or less" were trading a dirty mode for an e-taxi; the rest were trading a bike ride, a walk, or a transit trip),
- Decrease total dead-heading to 20 percent of total mileage or less (pre-pandemic, it was as much as 49 percent in some cities, and never lower than 35 percent), and...
- Somehow work it out so that non-shared mileage of pooled trips (e.g., the time between when the first and second passenger in the car pool are picked up) is reduced to 20 percent of pooled passenger miles (versus 48 percent pre-pandemic).
Repeat: app-taxis would need to do all five — and even if they did, they would only equal their users pre-app-taxi VMT, not reduce it.
According to Schaller, cities simply can't afford to keep betting on app-taxis as a solution to their transportation problems — and that will be an even worse bet if Uber and Lyft succeed in their multi-billion dollar quest to make their cars fully autonomous, which companies have long claimed could help make carpooling more efficient and attractive. (But probably not enough to shift 85 percent of e-taxi trips into crowded Waymo vehicles with bogglingly well-optimized routes.)
"The takeaway for city leaders right now should be that [app-taxis] need to be limited in big cities, just as taxis have almost always been historically," added Schaller. "And looking ahead to autonomous vehicles, there should be a shift in thinking about the potential of shared vehicles to reduce VMT. We really shouldn't be thinking about shared sedans and SUVs at all, even if they drive themselves; we should be thinking about shared vans, and buses. That's where AVs could have tremendous benefit."
But until that happens, the best ridesharing options for the health and safety of our cities are still the same as they've been for generations: the bus, the train, and the paratransit van.
Read the full study here.
