Given the proliferation of microtransit services trying to match drivers and passengers, you might think they had ride-sharing and carpooling all figured out. But the recent demise of Leap Transit in San Francisco—to say nothing of the other transportation start-ups that have failed without a media whimper—reminds us that even in a big city it’s not easy to fill empty vehicle seats. And in the suburbs, it’s downright mathematically impossible.
Or just about, anyway, according to a provocative new thought-experiment by Steve Raney, principal at a smart mobility consultancy called Cities21. In a working paper, the former Silicon Valley tech product manager crunched the numbers on ride-sharing in the Palo Alto area and found the odds of matching drivers with passengers long, to say the least. Raney calls it the “Suburban Ridematch Needle in the Haystack Problem.”
“I wanted to gently inject some reality into this,” he tells CityLab.
Things get real pretty quick by Raney’s math, especially when it comes a ride-share service’s ability to match commuters in an area with high single-occupancy vehicle rates. Let’s walk through his assumptions and conclusions:
- You have 10,000 people working in downtown Palo Alto.
- A zipcode in nearby Redwood City with a population of 31,500 residents is home to the largest share of commuters who work in downtown Palo Alto: 500.
- If 10 percent of these commuters were willing to carpool to work, as per national averages, then the demand for a ride-share service is at most 50 people a day. (And that’s a generous assumption, since the vast majority of carpoolers are family members or coworkers, as opposed to complete strangers.)
- If all 50 of these workers keep normal hours with standard morning commutes—again, a generous assumption—then they would all head to the office in a two-hour window. But since not everyone leaves for work at the same time, that window should be broken up into segments. Raney uses six 20-minute segments for the two-hour peak commute period.
- The six segments turn Redwood City’s 50 potential ride-share users into groups of about eight. In other words, eight out of 31,500 people in Redwood City might be matched for a carpool into downtown Palo Alto on any given morning.
- If there are even just two competing ride-share services, that number halves to four out of 31,500.
Four into 31,500 rounds down to roughly 0 percent of the population being eligible for a morning ride-share carpool service. “That’s how you quickly get to something small,” he says. These calculations don’t even account for what Raney calls the “Day 1 Challenge”—the need to have critical mass on the very first day of a service because potential users will only put up with one or two failed matches before they give up.
Raney says his exercise isn’t meant to suggest that suburban ride-share can never work. “I really want these guys to succeed,” he says. Instead, it’s a “word of caution” to all the young would-be disruptors wading into the field of urban mobility and flexible transit without a strong understanding of or background in transportation and network planning.
Shared autonomous vehicles would change the math, but there’s an even simpler solution in Raney’s mind: cities and employers need to charge commuters market price for parking. That basic policy shift would raise the cost of driving alone to work enough to make travel alternatives like carpool or ride-share services more competitive. In low-density areas where transit is expensive to provide, he says, microtransit then becomes a pretty viable option.
“It’s an incredibly cheap solution if there’s demand—especially if you’re going to start shifting mode in suburbia, it’s not going to be super cost-effective to inject a bunch of new public transit in there,” he says. “It’s really hard, but it’s pretty promising, too. It’s going to be a huge breakthrough when these things can start thriving.”