Uber, Lyft 'Biggest Contributors' To Traffic Congestion In San Francisco, Data Science Study Shows

Highlighting another challenge for Uber Technologies Inc UBER and Lyft Inc LYFT amid the two companies' lackluster initial public offerings of stock, a study that scraped data from the application programming interface (API) for both ride-hailing outfits found they are the "biggest contributor[s]" to growing traffic congestion in San Francisco.

The research, published in the May 2019 issue of Science magazine, shows that between 2010 and 2016, weekday vehicle "hours of delay" increased by 62 percent compared to 22 percent in a scenario without Uber and Lyft.

Population and employment growth, along with an increase in deliveries have contributed to traffic congestion in the city, said the study's co-author, Greg Erhardt, an associate professor for civil engineering at the University of Kentucky and an expert in transportation models and travel forecasting.

But even after controlling for those variables, said Erhardt, "the bottom line result – of the 60 percent increase, at least half and more like two-thirds is attributed to Uber and Lyft. If Uber and Lyft didn't exist, traffic would still have increased by 22 percent. With Uber and Lyft, traffic increased 60 percent."

The study contradicts ride-hailing companies' oft-cited claims that their services reduce or even eliminate traffic congestion.

In 2015, Uber co-founder and then-CEO Travis Kalanick said he envisioned "a world where there's no more traffic in Boston in five years." John Zimmer, the co-founder of Lyft, predicted in 2016 that private car ownership "will all-but end in major U.S. cities" by 2025.

But a growing body of research undermines these assertions. Studies have consistently found that ride-sharing is associated with more driving, less transit use and increased traffic.

The Science study marks an advance in the research methodology because it leverages the treasure trove of data Uber and Lyft possess showing where and when people are traveling.

Generally speaking, the ride-hailing giants hold their data close to the vest, arguing that releasing the information compromises privacy and undermines their competitive advantage.

In the case of the Science analysis, transportation network companies (TNCs), an industry term for ride-hailing outfits, declined Erhardt's request to share data for the study. Notably, they did offer to release data showing how many riders were taking to their services to get to and from train stations.

"I said: ‘No thank you,'" Erhardt said. "In this era," he noted, "everyone wants to talk about big data. But the reality is most big data is controlled by companies. If we let those companies only release data that supports their narrative – that's an important ethical issue."

So Erhardt and Joe Castiglione, the deputy director for technology, data and analysis at the San Francisco County Transportation Authority, and a co-author on the study, came up with another solution.

They reached out to two data scientists at Northeastern University who had put together a computer program that mimics what users do when they request an Uber or Lyft via smartphone: connect to their application programing interfaces, and see the locations via app of the 10 closest vehicles.

In late 2016, Erhardt said, the researchers set up "essentially dummy smart phone calls" in a grid across San Francisco and put in a request every second for a period of six weeks.

The work yielded a huge amount of data, around 17 kerabytes. "What you get out of it is traces where vehicles are driving around," Erhardt said, "but only where vehicles are available for a ride. That's how we were able to infer where and when the trips occur."

The team then related that data to a second data set provided by Inrix, a global company that provides a variety of Internet services and mobile applications pertaining to road traffic. That information goes back 10 years, allowing Erhardt and his team to see how traffic has changed over time. Crunching the numbers to account for the increase in population and employment, the team fingered Lyft and Uber as the primary culprits in the city's worsening congestion.

Uber and Lyft did not immediately return requests for comment.

The study raises a number of policy questions, such as whether Uber and Lyft should be required to share their data with cities so policy makers can better plan for traffic improvements.

"Understanding the factors of congestion is key to our ability to address the problem effectively and maintain the accessibility of our downtown core," said Tilly Chang, executive director of San Francisco Transportation Authority, in an email to FreightWaves.

"We are committed to data-driven analyses like this report to serve as the foundation for deeper understanding and informed action."

The city of San Francisco is considering a usage tax on TNCs, to be invested in traffic improvements. A congestion pricing proposal would also impact ride-hailing operations.

Uber and Lyft provide a clear benefit for many users, Erhardt emphasized. "For the person who is in the vehicle, it's convenient, relatively cheap. The problem is there are impacts on other people, including those people who are left on the buses. And those people who are left behind are perhaps different than the people who use the TNCs."

Uber and Lyft are grappling with a trifecta of problems. Since going public last month, Lyft's shares have dropped to $52.48. Uber's stock is also sinking after its lackluster IPO. Driver strikes called attention to a host of labor issues connected to gig economy companies.

The Science study yielded other findings – Uber and Lyft rides are concentrated in the densest and most congested parts of San Francisco and during the most congested times of day (morning and evening rush hour), when they compete with buses, bikes and pedestrians.

Between twenty and thirty percent of Uber/Lyft traffic is generated by drivers driving around waiting for fares, and most of the drivers are coming from outside the city, some from as far away as the Central Valley.

Image sourced from Pixabay

The post Uber, Lyft, "biggest contributors" to traffic congestion in San Francisco, data science study shows appeared first on FreightWaves.