[potatolicious]

> "I highly suspect it is plain cheaper (and more convenient) to just subsidize private car rides for disabled people"

This is true, and also why NYC has ParaTransit for disabled people, which amounts to an on-demand car/minibus service that is paid for by the transit agency.

The service levels are pretty terrible, though, because it turns out the private operators that actually provision these rides don't have a lot of incentive to go through the costs of operating accessible vehicles to address such a small market. Wait times for a vehicle are extremely long.

More importantly, disability and motion-handicaps are not binary, nor are they always permanent. There exist a large number of people who have difficulty climbing stairs who are not wheelchair-bound (see: the elderly, the pregnant, people coming from the airport...), and there exist a significant number of people who temporarily suffer from motion handicaps, whose handicaps are too temporary to be enrolled in ParaTransit (see: a guy with a leg cast).

Accessible stations and vehicles help these people also, not just the stereotype of the permanent wheelchair-bound person. The reality is that accessiblity features in our infrastructure benefit a lot more than those who are permanently or completely handicapped.

[Mythanar]

ParaTransit specifically is a non-profit organization. "Non-profit" and "incentives" generally don't mix well. Better service would require paying market rates - by the city, or partly by the city.

And it'd still be cheaper than building accessibility over existing subway infrastructure.

You make a very good point that disability is not binary; there is an entire spectrum from Michael Phelps to Steven Hawking. In some ways, having three kids on your back can be considered a [temporary] disability ;) My main point is that going down the percentage of affected population drastically changes the ROI of changing the infrastructure as a whole versus addressing the people involved individually. And wheelchairs specifically is fairly far down that scale.

[potatolicious]

ParaTransit does not operate vehicles themselves - ParaTransit is simply a program that purchases these services from private, for-profit operators (like, say, Uber), and manages eligibility checks as well as payment to providers for rides delivered.

> "Better service would require paying market rates - by the city, or partly by the city."

This is what happens. The riders themselves pay the "regular" transit fare of $2.75, and the agency (which is actually a state agency, but partially funded by the city) pays the rest. The service is operated at private industry rates.

The problem here is there is simply very little incentive for cab company or driver to service this demographic. Accessible vans are expensive and they reduce "normal" passenger capacity. Accessible customers require more time - both traveling to pickups, as well as getting in/out of the vehicles, which is time the driver is not being paid.

The small size of the customer base also means drivers spend much more time than usual traveling to a pickup.

The natural equilibrium state of this market, given the cost of providing the service, and the size of the customer base, results in very sub-par service levels. We can improve this somewhat by paying dramatically more for these rides, but that undermines the ROI argument pretty badly.

> "My main point is that going down the percentage of affected population drastically changes the ROI of changing the infrastructure as a whole versus addressing the people involved individually."

I agree - but my contention is that modeling the accessibility issues with the NYC subway as "permanently wheelchair-bound population" vs. "everyone else" is very incomplete. The ROI on accessibility improvements to our infrastructure must take into account all users of such features, not just the most extreme/permanent users.

The beneficiaries of, say, elevators in stations is the union set of [pregnant people, elderly people, wheelchair-bound people, crutch-bound people, people carrying large amounts of cargo, ...] - plus probably more categories I've missed. The union of these sets is pretty darned large, and a heck of a lot larger than simply the wheelchair-bound population.

The ROI argument only works if we assume a much narrower pool of users than it would be in reality.

[AdamJacobMuller]

> The small size of the customer base also means drivers spend much more time than usual traveling to a pickup.

Wouldn't a service like Uber avoid this issue? The cost for a disabled persons Uber ride won't be significantly above the cost for my Uber ride. While the cost of a dedicated service, per ride, will be significantly higher.

[taejo]

Did you happen to read the paragraph above the one you quoted:

> The problem here is there is simply very little incentive for cab company or driver to service this demographic. Accessible vans are expensive and they reduce "normal" passenger capacity. Accessible customers require more time - both traveling to pickups, as well as getting in/out of the vehicles, which is time the driver is not being paid.

The cost for a disabled person's Uber ride is significantly above the cost of your Uber ride, and so there's no incentive for Uber to service this market at the same price.