Uber Wants Us to Think It's Environmentally Friendly, But Is It?

For a company whose business depends on sending fleets of new cars out onto the road, Uber's environmental proposition is a strong one.

The theory is quite simple: Uber, along with other ride-hailing startups, will decrease carbon emissions in cities where it operates by lessening people's reliance on personal automobiles, minimizing traffic congestion on roads, and eventually disincentivizing the perks of owning a vehicle in the first place.

"A city that welcomes Uber onto its roads will be a city where people spend less time stuck in traffic or looking for a parking space," Uber CEO Travis Kalanick once said. "It will be a cleaner city, where fewer cars on the road will mean less carbon pollution—especially since more and more Uber vehicles are low-emission hybrid vehicles."

But so far, we've seen a staggeringly small amount of meaningful data to verify any of Uber's grand claims about its ability to reduce the transportation industry's carbon footprint. When nearly every other aspect of the gig economy has been quantified, why are we left clueless when it comes to the issue of Uber's environmental impact?

Uber has a long history of holding its user and driver information close, especially when it entails valuable and potentially proprietary activity data. Because of the startup industry's brutally competitive nature, young and successful companies such as Uber often err toward secrecy, which can hurt their transparency in the short term. Earlier this year, the California Public Utilities Commission issued the company a $7.6 million fine for failing to report accessibility and driver safety data back in 2014.

"Believe it or not, if you had a big data set and started to crunch the numbers, you'd be able to see an awful lot about their business—when and where they're most successful. Information abstracted from data like this could inform new or existing competitors," Dr. Susan Shaheen, co-director of the University of California, Berkeley's Transportation Sustainability Research Center, told me.

Some of Uber's unwillingness to open parts of its API to researchers and stakeholders can also be interpreted as due diligence, or perhaps just paranoia. Over the last couple of years, the company's security credentials have been patchy at best. Multiple data breaches, accidental leaks of customers' personally identifiable information, and deeply troubling evidence that Uber employees were able to track riders through its "God Mode" tool might have influenced the company's hush-hush attitude.

Instead, Uber has decided to periodically release its own in-house reports on ridership, sustainability, and emissions—all of which are provided with unclear methodology, and inherent bias.

In a 2015 blog post touting its new ride-sharing service UberPool, the company asserted that "San Francisco's uberPOOL efforts prevented about 120 metric tons of CO2 emissions from Feb-Mar 20th, equivalent to the output of over 128,000 pounds of coal." According to the post, anonymized UberPool data collected over a one month period was fed through an Open Source Routing Machine to calculate the mileage and emissions customers saved by carpooling, rather than opting for solo rides.

What Uber failed to disclose was that three separate promotional offers for discounted UberPool fares were issued during that time period, which could have artificially inflated the number of people who used the service. Additionally, the study assumed that every vehicle users hailed was an efficient Toyota Prius getting an ideal gas mileage of 50 miles per gallon. This conjecture is confusing, seeing as UberX car requirements for San Francisco don't limit drivers to hybrid cars only. And even among Prius owners, reports of actual versus advertised fuel economy vary, suggesting it's much lower than Uber guessed, at 30 to 45 mpg. Perhaps most egregiously of all, Uber never accounted for the possibility that UberPool riders might have otherwise taken more sustainable modes of public transportation, biked, or walked to their destinations.

Uber has also made bold declarations about drastically reducing the amount of cars on roads in congested cities like New York. The company stated one of its "simple" goals was to remove 1 million cars from New York City's streets. But how does that stack up against its actual deployment numbers?

Of the 30 largest US cities, New York had the highest share of carless households in 2014, coming in at 56.5 percent. That's out of roughly 3 million total households, which leaves a whopping 1.3 million cars spread out over a mere 322 square miles of land. In 2015, Mayor Bill de Blasio accused Uber of adding approximately 20,000 UberX cars to New York's busy streets. In 2014, that number was hovering somewhere around 10,000. The Taxi and Limousine Commission estimated that 14,088 black and luxury cars were operating on Uber's platform within five boroughs, as opposed to the city's 13,587 yellow cabs.

The average passenger vehicle emits 4.7 metric tons of CO2 per year, according to the Environmental Protection Agency. Therefore, in three years, Uber should have contributed 94,000 metric tons of CO2 into the atmosphere in New York City alone, which equates to nearly 101 million pounds of coal burned.

But the data isn't that simple. According to open data compiled by Todd Schneider, New York City's yellow cabs averaged just under 352,000 trips per day in January 2016, while Uber vehicles made 150,000 trips per day. Uber's New York contractors also work fewer hours during the week than their taxi-driving counterparts. And it's unclear whether Uber's "new" cars were recent additions to its fleet, or simply existing taxi vehicles that had been absorbed into its platform.

Uber has also kept mum about the efficiency of the proprietary algorithm it uses to direct drivers to new passengers. We know, theoretically, how its routing engine works, but there are a lot of unknowns in regards to how successful it is at charting actual trips, and whether or not drivers are relying on it. However, both drivers and passengers have complained about Uber's in-app navigation feature before, accusing it of suggesting longer routes. It's also not uncommon for drivers to use alternate GPS services, such as Waze, to navigate cities during peak traffic hours.

How this affects something like UberPool, which must use the shortest linear route possible to be as sustainable as Uber claims it is, is vague. The company presented its carpool service as a way to raise overall earnings for drivers by combining two separate rides into one trip along a similar route, but according to numerous negative posts on the Uber Drivers Forum, not all of them agree with that sentiment. Because many drivers feel they're doing more work for a negligible fare increase, there's little incentive to accept UberPool trips, especially when they might make more off single-rider trips during surge pricing. Some drivers admit they turn down 100 percent of UberPool requests.

That doesn't mean ridesharing can't be environmentally friendly, though. In 2014, MIT researchers conducted a computational study of 172 million New York City taxi trips completed in 2011. By modeling different shareability networks throughout the city, the authors were able to see that certain carpooling algorithms were able to cut trip lengths by 40 percent or more, reducing driver costs, emissions, and rider fares. On average, they found, every trip within their sample set was shareable with 100 other trips.

Uber reiterated this vision in a statement to Motherboard, saying that "When people forego car ownership, a city's overall vehicle miles traveled (VMT) can also decrease—which can mean less gasoline is consumed and fewer emissions are generated. Furthermore, Uber helps enable smarter transportation policy choices. Fewer personal cars require fewer parking spaces and parking lots, freeing up space for more productive and sustainable uses."

Yet the company has heretofore made it difficult to independently study how far along this track it might be. It was this sort of opacity that instigated a massive and unprecedented review by UC Berkeley transit researchers and the National Resources Defense Council of Uber's environmental impact claims. According to Dr. Shaheen, who is heading up the study, Uber has agreed to submit company data, platform data such as anonymized driver and rider information, and will also be issuing a rider survey which aims to collect sociodemographic and attitudinal data from its customers.

All research will be conducted by a neutral third-party, without funding or control from Uber. Unlike previous reports and analysis released by Uber's own data team, the study will be independently verified. Lyft will also be analyzed, and the team's results will be published as early as fall of this year.

"The way that the study came about is that I was being asked by regulators, legislators, and stakeholders to weigh in on a variety of legislation related to Lyft and Uber. But nobody had ever studied the environmental impact of ride-hailing companies before," Amanda Eaken, deputy director of urban solutions and sustainable communities at NRDC, told me.

A spokesperson for Uber told Motherboard she didn't think they would have anything to say on the study.

Some of the key metrics the investigation will look at is whether ride-hailing companies like Uber have influenced changes in auto ownership, if they've had notable impacts on miles traveled, and what that all equates to in terms of greenhouse gas emissions.

Furthermore, the study is coming at a time when Uber is heavily shifting its resources toward UberPool, which will provide experts and regulators with helpful comparisons between carpooling services and classic services such as regular, single-ride taxicabs.

It's possible their findings will suggest competing trends, such as some Uber riders taking cars to and from mass transit stations, and others choosing to hail cabs in lieu of more environmentally conscious options like biking. In cities with multiple public transportation options, Uber might also be encouraging people to hop in cars to avoid things like bus transfers.

Dr. Shaheen led a smaller, less exhaustive study in 2014 that observed a randomized group of 380 ride-hailing customers in downtown San Francisco. The survey was carried out before the emergence of carpooling services like UberPool, and should not be interpreted as representative of today's market, but nevertheless revealed a curious theme. Half of the trips the group surveyed replaced non-taxi modes of transportation, including public transit, walking, and biking.

Will that trend be duplicated in the ongoing study? It's still too early to tell. It would be to Uber's benefit if results indicated that ride-sharing instigated behavioral shifts such as more people feeling empowered to take public transportation, or deciding against purchasing a car. After all, Uber needs to cohabitate with existing transit options, and should not only be complementary to other environmentally friendly modes of transport, but also adaptive to ongoing initiatives cities are pursuing to make their carbon footprints as small as possible.

"There's a sweet spot to these services. Are they convenient, cheap, and quick? But are they also green and equitable?" asked Tim Papandreou, chief innovation officer at the San Francisco Municipal Transportation Agency. "We don't want everyone chauffeured around the city. There's no room for that in San Francisco."

If UberPool is found to be as sustainable as Uber assures it is, a transportation authority like the SFMTA might consider developing carpooling networks inside of city limits to incentivize the company to focus on its ridesharing services, Papandreou told me. But this first snapshot of Uber's footprint is just that—a snapshot. Several years of additional studies would need to be completed before regulators could propose any changes to the industry.

Regardless, riders, drivers, and legislators alike will soon have the best comprehensive look yet at Uber's environmental impact. Where that takes us is yet to be seen.

In the wake of all this, the CPUC has ordered Uber to validate its carpooling claims through an independently conducted study of its own services. Currently, there's no relationship between Dr. Shaheen's research and the CPUC ruling. But either way, a verdict has been hailed, and not a minute too late.

Uber Earth is Motherboard's exploration of the ways Uber has already changed the world and how it stands to do so in the future. Follow along here.