These are good questions. They have been mostly answered by the law, rules and public paperwork. Past a certain size, having a dedicated team working on disability is not only the right thing to do, it's almost essential to ensuring compliance. (Below that size, a consultant.) Dealing with disability is complicated; "disability" deals, by definition, with edge cases.

> Why, ethically, should this cost fall on Uber instead of the driver? Either way an inefficient market is being created here because the government is stepping in and saying that a transaction made between two rational actors is unfair even if those parties agree to the terms of the transaction (unless Uber is a monopoly).

You are relying on "ethical" principles that I would say are alien to the vast majority of people.

There is no reason why Pareto efficiency should be ethically prioritized if the alternative is welfare enhancing in the aggregate.

Can you describe to me a situation in which reaching pareto efficiency would, in your opinion, not be the optimal state for the welfare of the aggregate? I'm trying to imagine a situation where resources are optimally distributed where that would be suboptimal for the welfare of everyone.

edit: I'm new to the concept of pareto efficiency, so hopefully this reads more as a sincere question rather than a challenge to your opinion

Sure.

Imagine a situation with two people and a state-like actor.

One person has all of the food in the entire world minus one loaf of bread. The other person has one loaf of bread. After they eat that loaf, they will probably die of starvation in the next 30 days.

This is a Pareto efficient distribution.

If the state-like actor intervened and re-distributed some of the food to the other person (with coercion), I would argue that aggregate welfare would be improved.

That's not a pareto efficient distribution. The efficient option would be for the person with bread to sell bread to the person without bread in exchange for future labor.

Pareto efficiency is defined to be an economic state where resources cannot be reallocated to make one individual better off without making at least one individual worse off. If a redistribution can make things better, it is not an efficient state.

> The efficient option would be for the person with bread to sell bread to the person without bread in exchange for future labor.

Okay, with work, let's say they are given the bare minimum to survive and they will get their work done (no rational reason to offer a higher wage). It would still be welfare enhancing to re-distribute.

> If a redistribution can make things better, it is not an efficient state.

This is just straight-up false, no other comment. Not what Pareto efficiency. The redistribution makes the person taken from worse off.

If they both have the bare minimum to survive, redistributing from one to the other puts the first below the minimum. If someone has a surplus, the efficient thing is to trade it. If they don't, redistribution hurts them as much as it helps another. There is no free lunch.

Again, that is the definition of pareto efficiency.

> If someone has a surplus, the efficient thing is to trade it.

With only two people, the most efficient thing to do is to trade it for the bare minimum the other person needs to survive. Beyond that is not Pareto efficient.

Yes, you give someone the bare minimum they need to survive. Then they want more. Now you're back to the situation where you have a surplus and they're willing to trade, the efficient thing is to trade your surplus. Continue repeating until one of you dies of old age, and the remaining survivor has all the bread.

So long as you have surplus, and that surplus has value, it makes sense to trade it, rather than have it sit idle.

Yes, someone living their entire life on subsistence amounts of bread while the other person has a great 3 meals a day is not a welfare optima.

I don't think you're understanding this. As long as person A has more bread than they're going to eat, and person B still wants more bread, it is still optimal to trade. A well fed individual can do more labor than a minimally fed person who can in turn do more labor than a dead person. As person B's quality of life improves, they are in an ever better position to trade. The bread is worthless to person A, it doesn't matter how little person B gives for it, it is still optimal.

It does not become Pareto efficient until person A is worse off by trading bread, by which point they no longer have enough to completely satisfy their needs.

This is a straw-man description of the ethics of Pareto efficiency. Even as such, if the second person has nothing the first person desires, what obligation does the first have to share their bread? Why can't the second person make bread or an alternative? Are they unable to, and if so, are they strictly a burden on the first person?

> what obligation does the first have to share their bread

None. However, part of the role of government (I would likely say it's entire role) is to make society as a whole better. It does this by imposing rules (obligations) on members of that society (taxes, a force to enforce the concept of private property, etc).

In the toy case described above, it is likely to the benefit of society to have both members survive and one member be unhappy; rather than have one member survive, but be happier. Hence, the government enforces an obligation on the first person to share their food.

Yes, you are coming from a very objectivist/libertarian perspective that I reject.

It's not a straw-man description of Pareto efficiency, it's a toy example designed to answer a very explicit question by the GP. What I described was a Pareto efficient distribution, and not the most welfare enhancing one.

What if neither person can make bread and the other just happened to inherit it all from their parents? Who is burdening who?

And that's before we get into the real-world ethics of whether we can honestly claim that contemporary wealth distributions are really derived from consensual interaction. Why should the descendents of nobility be entitled to wealth deriving from what are effectively violent takings?

I don't know - the whole deontological libertarian project in this sense seems like it has gaping holes.

I share the objectivist standpoint that, as long as wealth is accrued through fair transactions, having occurred without coercion (including monopoly influence) and because parties provided equitable value to each other, it (and its distribution to heirs over time) is ethically acceptable.

So then you'll agree that in a society where people without jobs are precluded from survival, labor is (in effect) coercive. Thus, any wealth accrued by arbitraging the labor of others was unethically gained and fair game for redistribution up to the point where joblessness is no longer an existential threat?

The fact that you need something to survive doesn't mean that every transaction providing it is coerced. Do you consider it coercion when you shop at the grocery store?

That makes sense, I guess there's subjectivity involved as to what that would look like as I imagine the death of one of those people would be an unwanted consequence to the person with all the food and as such the first person would be better off giving food to the starving person.

Your point makes sense to me now though, so thanks

> unwanted consequence to the person with all the food and as such the first person would be better off giving food to the starving person.

Doesn't really play out that way in actual real life (people still starve every day and both you and i probably have the resources to stop at least one of those starvation deaths), so didn't think of including this altruism in my toy model

Respectfully, I don't think we can say axiomatically that welfare is enhanced in the aggregate by Uber stakeholders ({drivers|employees|shareholders}) sacrificing in favor of the disabled because the Nth degree ethical consequences are unknowable.

My claim is that there is nothing about Pareto efficiency that is ethically preferable to a situation in which more people are better off.

Arguing that this policy would render some consensual exchanges illicit is irrelevant, because I don't think the most ethical outcome is axiomatically realized by solely consensual exchange without any coercion.

Being disabled you get to experience the wonders of the free and efficient market every day.

For example when you need to go somewhere, as a rational actor you can choose with your wallet:

* You can choose to go by foot using the sidewalks that aren't accessible.

* You can choose to use public transport which is not accessible.

* You can call a cab and pay an exorbitant amount as long as you don't tell them that you're disabled or they won't come and they might still leave when they see you.

* You can call an Uber which is quite pricy as well as long as you don't tell them you're disabled or they won't come and they might still leave when they see you.

Thankfully after three years of bureaucratic mazes and appeals, the magnanimous state has granted you benefits, although they don't constitute a living wage and don't allow you to pay rent nor food. (The glorious free market has given you a choice of one of twelve accessible dwellings in a radius of a 100km so that you may rationally choose the lowest rent.)

Since you can't pay for rent and food and transportation you'll need to find a job to pay for your lavish lifestyle. Now if it's before 2020 don't expect remote work, disability associations have been asking for it for 10 years for jobs that don't require physical presence but as everyone knows that's entitlement of the highest order. Although if it's after 2020 then it stops being entitlement because the average person needs it and wants it.

Now if you've managed to get yourself an education before hand, despite your health issues, your inability to go outside, and your general standing as a second-class citizen, then you should be able to get a remote job, and even if it stops being remote you should be able to pay to go there using the methods listed above. Although if you make below or up to the median salary, it'll cost your more to go to and from work than what you earn there, also you no longer have your benefits. But at least you get to be part of society as long as it's the only place you go because leisure is entitlement.

This is not fiction, this is how I got my first job, although Uber wasn't a thing back then, and you couldn't be remote. Also I didn't have benefits, I did freelance while I was doing a Masters at a college I couldn't attend because of transportation costs, save for the exams. Then when I got my first job I did pay more than my salary in cab fares using my savings and the rest of my college loans until I could move up and get a higher salary.

Keep in mind the following scenario:

A user with a disability books an Uber. They are unable to get to the Uber within the allotted time. The driver leaves. The customer then books another Uber, and this time successfully makes the trip. The customer calls Uber to refund the cancellation fee, which Uber does. Uber is then out the time the first driver took to reach the destination, the driving time, and the cost of the customer service contact (which can be quite expensive). They also have to deal with reputational and regulatory risk. This isn't an invented scenario - I've seen this happen.

Wouldn't it be less costly for Uber for the driver to just wait an extra 5 minutes?

> How should a person with disabilities indicate that they have a disability requiring extra time in a fraud proof manner?

There are already systems for allowing people with disabilities to qualify for reduced fares or paratransit - typically involving having your doctor send a letter to the local transit agency. Uber could either replicate such a system, or they could accept documentation from a local transit agency or they could allow people to self-designate and eat the cost of fraud (my guess is that stigma and lack of awareness of the program will prevent a lot of people from fraudulently using it).

> Should all disabilities be treated equally? > Who decides how much time should be granter per?

Ideally you'd make a decision based on an individual's needs, but if I was a product manager for Uber I'd just just set a high limit for everyone who qualified for the program initially and then figure out later if something more complex is needed based on how much time drivers are actually waiting. If this program helps Uber reduce the number of cancellation/re-bookings/1-star reviews

> Should this same policy apply to people with ephemeral injuries?

Yes, temporary disabilities are disabilities.

> Should Uber compensate the driver for this extra time to ensure the driver is still incentivized to pick up people with disabilities?

Yes, for exactly the reason you mentioned

> Why, ethically, should this cost fall on Uber instead of the driver? Either way an inefficient market is being created here because the government is stepping in and saying that a transaction made between two rational actors is unfair even if those parties agree to the terms of the transaction (unless Uber is a monopoly). If the driver is not compensated fairly for their time, why is this sacrifice ethically permissible?

First, I'm not convinced that Uber having to pay for extra waiting time is against UBER's best interest - see the scenario above. But even if it were on balance a net cost to Uber, society has an interest in ensuring that everyone is able to travel and do all the things that travel enables (any kind of in-person commerce, healthcare, etc).

> - Should this same policy apply to people with ephemeral injuries?

Are these sorts of things considered "disabilities" though, from the perspective of ADA and other disability-related laws? A bunch of years ago I broke my foot and was on crutches for 8 weeks. I don't recall it being that difficult to get to my Ubers in time (they may have been more lax with wait time back in 2014). It didn't occur to me to even think I should be entitled to extra time.

> - Should all disabilities be treated equally?

I would assume not? IIRC the key phrase in disability cases is "reasonable accommodation". I think it's reasonable to grant someone extra time if they have a mobility issue, but perhaps not if they have something like diabetes?

- How should a person with disabilities indicate that they have a disability requiring extra time in a fraud proof manner?

That might not matter? From what I understand, the ADA (and similar laws) are worded and interpreted to give the person claiming a disability the benefit of the doubt. I would guess that Uber is expected to eat the cost of fraud here, unless they can prove in specific instances fraud occurred. In general, though, the opposite case is much worse: denying an accommodation because you (erroneously) believe no accommodation is needed can get you in big trouble.

> - Should Uber compensate the driver for this extra time to ensure the driver is still incentivized to pick up people with disabilities?

Personally I would say yes, definitely. I know this isn't how Uber works from the driver perspective, but the "meter" should start running as soon as the driver gets to the pickup spot. Uber eats (or builds in) the cost of a reasonable "grace period" for riders, which is Uber's decision. Ignoring the disability question, if Uber drivers were truly independent contractors and had full autonomy, they could certainly decide to give riders zero time: either be at the pickup location already when the driver arrives, or you miss your ride. (Not saying this is a good policy, but in theory this could be something that's allowed; pre-Uber taxis in San Francisco would sometimes do this, and it sucked.)

Not sure if Uber should specifically note "this person is disabled so you'll be getting a little more money as compensation for waiting longer"; probably not? Though I guess after they accept the ride and arrive at the pickup point, they'll see their required pickup wait time be 8 minutes instead of 4 (or whatever), so they'll at least know at that point.

> - Why, ethically, should this cost fall on Uber instead of the driver?

Because, despite Uber's insistence that drivers are independent contractors and are not employees, Uber still sets ride pricing and does not give drivers all that much flexibility in what fares they are willing to take (that is, they don't give drivers much information up-front about fare details, to avoid discrimination).

Ultimately, though, the cost will more likely fall on riders (all riders, that is), not drivers; the rational response here from Uber will be to raise the cost of every ride slightly, in order to cover the added cost to transport people with disabilities.

> the rational response here from Uber will be to raise the cost of every ride slightly, in order to cover the added cost to transport people with disabilities.

And ironically, one of the reasons a lot of people like Uber is because it's cheaper than cabs... because it ignores regulations. Complying with regulations has a cost.