For years, Uber and Lyft have claimed their services do things which they do not, such as allowing drivers to earn $90,000 a year and reduce traffic in urban areas. Now, a new study suggests that, despite having the stated goal of reducing car ownership, the ridehail companies may in fact be increasing the number of new car registrations in U.S. metro areas.

Primarily through studies based on rider surveys, Uber and Lyft have claimed they reduce car ownership because they act as the final piece of the puzzle, along with public transportation and bicycling, to allow primarily urban residents to finally ditch their cars. There is little doubt that this happens on some scale; a survey conducted by Lyft claims that 49 percent of riders without a car would be “more likely to purchase” a vehicle if Uber and Lyft weren’t available and that the service played some part in the consideration of riders getting rid of 500,000 personal vehicles.

But the new study from researchers at Carnegie Mellon University found people ditching cars when Uber and Lyft came to town may be counterbalanced by people who buy cars so they can drive for those services. This, too, undoubtedly occurs on some scale. Uber was even offering a loan program (an extremely sketchy and ethically dubious one) so people who couldn’t afford a car could start driving for it. Other drivers were purchasing vehicles that qualified as Uber Black or XL (and their Lyft equivalents) so they could earn higher fares.

The study took advantage of the fact that Uber and Lyft entered different cities at different times from 2010 to 2017. Using that staggered entry to look at how car registration rates changed in the respective urban areas, the researchers found that, on average, vehicle registration increases by 0.7 percent after the ridehail companies enter a new market.

While less than one percent may not sound like a significant impact, Professor Jeremy Michalek who led the study said it’s noteworthy not just that it’s an increase at all, but an increase across the board. “It changes my view of what is happening with these additional modes.” He added that these results update and improve upon a previous study they conducted looking at state-level data that found ridehail reduces car ownership rates.

But, this isn’t quite the definitive undercutting of Uber and Lyft’s narrative as it might appear, both because of some limitations in the study itself, but also broader problems with trying to link any relatively small change with car registration rates in general.

For one thing, the study takes a look at entire metro areas, rather than the downtown urban cores where ridehail usage is highest and in theory will have the biggest impact on car ownership rates. On the one hand, looking at the entire metro area makes sense because most Uber and Lyft drivers do not live in the expensive downtown urban cores where the services are most popular (some also commute from outside the urban area entirely and therefore won’t be captured by the study). On the other hand, entire metro areas are already “very auto-oriented and little TNC [transportation network company] usage,” said Bruce Schaller, a researcher who has extensively studied Uber and Lyft’s impact on cities. “So the model inherently will be mostly noise, not signal.” Likewise, there’s a broader question about to what degree reducing car ownership in auto-centric outer areas of cities or suburbs is even a realistic goal for Uber and Lyft.

On top of that, the years the researchers studied—2010 to 2017—was one of economic growth. Schaller argues that when analyzing something as cyclical as car purchases, it is important to look at an entire economic cycle. When he did so, he found “car ownership is flat in the big cities with the most TNC trips.” Michalek says they accounted for this by using a “well-established statistical technique called difference-in-differences,” but Schaller says this approach “opens the door a bit to seeing growth as caused by something that is merely correlation.”

If nothing else, the takeaway from all this is that the evidence is sufficiently muddled that nobody can claim definitively what impact ridehail companies have had on car ownership rates. In some ways, even this indeterminate result reflects the very same narrative shift that surrounded the impact Uber and Lyft have had on congestion, a much more important impact than car ownership overall (since a car sitting in a driveway is not creating harmful emissions or adding to traffic). Early in Uber and Lyft’s existence, the companies suggested their services would improve traffic in cities because there would be fewer cars on the road operating in a more efficient manner. But, using similar methodology as Michalek and taking advantage of the staggered entry into different markets, researchers chipped away at that narrative over time. The biggest reason the ridehail companies increase traffic is because they’re not nearly as efficient as they made themselves out to be. The vehicles don’t have passengers in them about half the time. In 2019, the companies themselves were forced to finally concede they had in fact made traffic worse.

Whether we will ever arrive at a similar point regarding car ownership is yet to be seen. But in some ways, it doesn’t matter. Reducing car ownership, Michalek said, is not a very worthy goal in and of itself. “I think the reason we might want to reduce vehicle ownership is we want to reduce the negative impact of ownership, like congestion and emissions and crashes and those things.”

Both ridehail companies seem to acknowledge this. Uber and Lyft have made commitments to reach zero emissions by 2030, a much more significant but difficult goal for two companies that currently rely on polluting vehicles to make the vast majority of their money and continue to incentivize drivers to use less efficient vehicles by paying higher rates for luxury SUVs with poor gas mileage and a higher likelihood of killing pedestrians. But car ownership in and of itself is not a cause of nor solution to urban issues if people will simply replace many of their private vehicle trips with vehicle trips in other people’s private cars.

*By Aaron Gordon, VICE*