In my city, rents have skyrocketed near transit stops since it's so essential. When you're a 20-25min walk from a transit stop, prices are a lot cheaper, but y

In my city, rents have skyrocketed near transit stops since it's so essential. When you're a 20-25min walk from a transit stop, prices are a lot cheaper, but you become hard to get to and hard to get to work. New transit isn't really an option because of density and cost (the cost to build per rider is astronomical). But the city is also dead set against increasing the density near transit stops to decrease rent/buy prices (and increase transit ridership due to it being convenient). Public transit in many cities creates hot-spots in the real estate market that isn't good public policy either. Of course, the article is right that people shifting into less dense transit will have bad environmental and congestion problems. But I don't think that traditional public transit will be the way of the future. Rather, I think that self-driving, reasonably high-density vehicles will be the future. Imagine a nice bus that seats 15-20 picking people up along an ad-hoc route in the morning determined as riders hail the bus and are instructed to an ad-hoc stop within a block and dropping them off within a block of their destination. That's a lot more convenient than most public transit systems where you have to travel to stops, maybe change lines, not getting exactly where you want to go, etc. It could also cut down on vehicle miles travelled by creating optimized routes. If Uber Pool can do what bus service can do for barely more money, a self-driving bus will be way better than a standard public transit experience and as efficient or more efficient environmentally. In fact, I think the self-driving future in cities will be determined by good incentives. During peak periods, charge for congestion. Not broad-based attacks on vehicles, but an incentive for people to commute in higher-density vehicles where the charge can be spread among more people. It would be easy for a city to incentive Uber, Lyft, and others to offer higher-density options for commuters via congestion charging. Likewise, environmental incentives could be offered to push customers and companies toward more economical vehicles and routes. I think it's reasonable to assume that in a self-driving future, companies like Lyft and Uber would want a lot of economical vehicles like Priuses getting 50MPG in the city. For higher-density vehicles, 10% fuel savings could push margins up a couple points - especially if environmental fuel taxes are put on top of the price of fuel. Similarly, better routing can lead to fewer miles travelled leading to savings. For those that want the privacy of single-person travel, they can be charged an appropriate amount to compensate society. Uber can't do a lot of high-density vehicles currently because it relies on vehicles owned by random people. But when self-driving vehicles truly become mainstream, there's no reason Uber wouldn't want to expand into company-owned, higher-density vehicles. They could run these at a fraction of the cost that most public transit systems are running at. In lower density areas, maybe medium-density vehicles and in even lower density areas, single-person rides in small vehicles may remain common. When Uber can control its vehicle stock with self-driving vehicles, there's a lot of options for them to optimize in ways that will boost their profits while also helping the environment and congestion. Maybe you think Uber isn't interested in a low-rent, non-premium service. That may be, but so many are interested in transit and it would be reasonably easy for a competitor to put together such a service and undercut Uber on price for so many riders. Uber would want to respond. Ultimately, the article talks about bus routes doing 10 boardings per hour and how that's more than an Uber will do. That's probably true, but an Uber-bus would likely do more boardings due to better ad-hoc routes and more convenience. In my city, fares only cover a quarter of bus operating costs (never mind capital costs) and two-thirds of subway costs. Part of the problem is that a lot of transit systems work off the principle that they need to serve off-peak and lower utility uses in order to hit that critical mass that would make them a good choice for users. Ad-hoc, self-driving routes could relieve transit systems of their bigger loss-leaders using vehicles optimized for those areas. Similarly, off-peak service that often sees low ridership and loses money could be off-loaded. This is also an environmental win - subways are environmentally friendly when there's a lot of riders, not when they're mostly empty. A bus route that's losing over $10 per rider is bad for a public transit system and also bad for the environment since the bus probably doesn't have enough people on it to make it fuel efficient on a per-passenger basis. I think there's a genuine opportunity to do a lot better than current public transit with self-driving vehicles. Something that's a lot more environmentally friendly and a lot more convenient.