For most Americans, public transit basically has three virtues. The first two cater to liberal sensibilities: it’s environmentally friendly, and because it’s cheap, it’s effectively a sort of transportation safety net for the poor. On top of those feel-good benefits, there’s a “business” case, which is that public transit is good for economic development.

These three virtues are broadly understood—or at least, understood as standard arguments in favor of transit, even if not everyone finds them convincing—by most people, even those who couldn’t get into the weeds about dedicated lanes for streetcars or all-door fare payment. They are accessible to average voters and local elected officials alike.

By Ibagli (Own work) [Public domain], via Wikimedia Commons
By Ibagli (Own work) [Public domain], via Wikimedia Commons

But as city centers are changed by both demographic shifts and technological innovation, it seems increasingly necessary to add a fourth virtue to the broad public debate over urban transportation: Public transit is space-efficient.

But the knock on transit is that it’s slow.  Or slower than a private car.  And the results of a recent race in Chicago shows that transit is considerably slower than UberPool, Uber’s carpool like ride sharing service. The race was results were confirmed by a study from DePaul University’s Chaddick Institute for Metropolitan Development, which found that Uber Pool was significantly faster than Chicago’s public rail and bus system for most trips. While Uber Pool also cost more, its price was significantly below that of regular Uber service—and even further below the cost of traditional taxis. The implication of the head-to-head comparison is that, for a significant number of people, switching some proportion of the CTA’s 1.7 million daily trips to Uber Pool might be worth it.

And even Uber Pool’s cost disadvantage over public transit might disappear if a new pilot program in Boston catches on. This month, Uber announced that it would introduce monthly Uber Pool passes—like transit passes—for just $2 per ride, or less than the MBTA’s $2.25 fare. While these temporary low fares are no doubt a money-losing loss leader, if UberPool fares are anywhere close to the price of public transit it would seem like we’re probably looking at a massive shift from public transit to these sorts of ride-hailing services. And wouldn’t that be a good thing, if it’s both faster and nearly as cheap?

But here’s where the importance space-efficiency comes in. When one person switches from the bus to Uber, two things happen. One is that they get a faster trip almost by definition: A vehicle that makes many stops (the bus) is going to be slower than a vehicle that makes few or no stops (the Uber) unless the bus has some other advantage, like transit lanes that allow it to avoid traffic congestion. And the vast majority of American bus lines are given no such benefit.

The second thing is that they switch from a very space-efficient vehicle, where they probably take up only a few square feet on the road, to a very space-inefficient vehicle, where they take up many, many times more.

Screen Shot 2016-07-05 at 9.38.11 AM

Screen Shot 2016-07-05 at 9.38.00 AM

(Yes, the above image isn’t totally fair—Uber and autonomous vehicles could theoretically reduce the space needed for parking by quite a bit, and potentially services like Uber Pool could shrink the space it takes up on the road as well. But regardless, there’s no doubt that while traveling, Uber and autonomous vehicles consume space much more like private vehicles than like public transit, pedestrians, or bikes.)

When one person makes that switch, it doesn’t affect traffic very much. But if many people were to make that switch, particularly on streets that carry a decent number of public transit passengers, the implications are quite serious. Even in an Uber Pool, riders will be increasing the road space they take on manyfold—and thereby radically decreasing the number of people who can use the road at a time. In other words, they will be creating serious traffic congestion.

In many cases, there’s just no plausible way to physically move everyone who currently uses a road if bus riders were to switch to Uber-like services. On Lake Shore Drive in Chicago, roughly a quarter of all the people on the road at rush hour are on public buses, and travel speeds already regularly fall well below the speed limit. Were a large proportion of those bus riders to switch to Uber, they may very well see their own travel times become longer, as they suffer from the congestion that they have helped to create. Uber makes it possible for one person, or even many people, to take a faster trip; but if we were all to take Uber, we’d all have even slower trips.

Crucially, though, no matter how slow traffic gets, Uber will almost always be faster than street-running, mixed-traffic transit services, meaning no one will have an incentive to fix the problem by switching back to the bus. This is an issue that can only be resolved by public policy: Re-prioritizing space-efficient transportation by, for example, creating bus lanes.

The broader point here is that the kind of dense, pedestrian-friendly neighborhoods that are increasingly in demand can’t function without very space-efficient forms of transportation. For very dense areas, like central business districts, that’s often because you literally can’t fit all the people traveling to and from them on roads in cars. In lower-density residential neighborhoods and commercial corridors, travel congestion may be less of a barrier, but parking becomes the pinch point: if everyone arrives by car, the amount of parking required would end up leveling half the neighborhood. (Recall the parking requirements in downtown Kalamazoo, which call for parking lots so big that they wouldn’t fit on their parcels, even without any actual building.)

If this virtue of public transit (and, of course, walking and biking) is understood, then calls for there to be “enough” parking, or to “solve” the congestion problem in popular, dense neighborhoods, stop making any sense: allowing everyone to arrive by car is just physically incompatible with those types of communities.