Hacker News Reader

Comment by paublyrne | Hacker News Reader

This makes me wonder aloud ... why shouldn't we charge people who are more affluent more than those who are poor? We apply this idea already in our income tax systems.

I'm not sure I subscribe to this idea, but I notice I don't feel morally aghast at this story.

View on HN

Charging able-to-pay people more is called "price segmentation" and is quite common... but it usually happens by adding some sort of features at a (much) higher margin rather than by price discrimination on exact same product.

The difference between letting folks opt-in to a more overpriced service versus selecting it for them is huge, because it centralizes all the economic choices in Uber HQ. Uber knows you will pay more, but does it know that you have 5 kids to support with what's left over? Or that your costs come out of a non-profit budget? Or that your need for rides is associated with disastrous new medical expenses? These concerns are built into our tax code, but Uber doesn't have the interest or ability to align with any kind of moral consensus about what "able to pay" should really mean... they just have a lot of insight into what you're willing to pay for rides.

Early on, Uber was championed in some circles as a good example of technology enabling a two sided market to compete with highly centralized industries. Over time, they've stopped being a market, and insisted on becoming a broker that keeps both sides at an information disadvantage. This is a good way to collect rents, not to drive competitive service.

The funny thing is that Uber already has segmentation with their Black cars. Presumably if someone's on an expense account and doesn't give a fuck about price, they'd be using that service.

I think this isn't so much segmentation as it is simply extracting more money from people who probably aren't going to bother looking for alternatives like Lyft or a vanilla cab.

The black cars cost much more, but when there's chaos at an airport, the black cars are much easier to get from what I've seen. You're definitely right, the people who don't care or value their time more will choose black.

Price segmentation often happens not by adding features to the high end pricing, but removing them from low end prices, even at extra cost. Think CPUs shipping with disabled cores, etc.

You've got the cause and effect backwards. CPUs ship with disabled cores because that's the only way to recover from manufacturing errors that would otherwise make that segment of the wafer completely useless. Instead of scratches and misplaced dopants costing Intel the full price of a $500 i7, they disable the damaged silicon and sell it as a cheaper i3 or i5. It's literally the only way to make modern CPUs economical and the price discrimination is more of a consequence of physical phenomena. Likewise with RAM clockspeeds: each chip is "burned in" during fab and the most stable ones (with the fewest imperfections) are sold as overclocked RAM while the worst are sold as underclocked low power mobile chips.

If you don't care at all about cost (aka yield), like with RAD hardened CPUs or high throughput serial processors like those from IBM, there is no real room for discrimination and you just pay for the NRE/fabrication plus a profit for the vendor.

I had to use Uber and Lyft extensively after a car accident left my car being repaired for months.

There was no change in my income but I became a frequent rider.

My trips started getting more expensive, Lyft stopped telling me “This is a good deal” on prices before trips, and all kinds of shadiness.

It feels like you’re being mistreated when you’re punished for using a service more than others.

I’d much prefer they give out a fixed amount free credits for new accounts, make surge pricing crystal clear, and then keep prices at a base rate.

Instead of trying to play 3 card monte with the real price of a ride.

Lyft used to make surge pricing crystal clear, but now it's impossible to see.

It's easy to see why people would be irritated, but at the same time, if a company is losing vast amounts of money providing a service, then it's inevitable they're going to start getting creative about charging more.

Because you'd typically have to collude to do it. If just one company did this, anyone being charged more would switch to the other company. You'd have to have some kind of agreement among all the competitors to do it, or build it into law.

Some mobile games with in-app purchases do various flavors of this, but it's more based on willingness / capability to pay than affluence. And of course they're structured such that switching to a competitor would make everything you've previously paid in moot.

> Because you'd typically have to collude to do it

There's all sorts of ways that companies charge the wealthy more for the same product without collusion.

1. Coupons

2. Generic brands that are repackaged name brands

> anyone being charged more would switch to the other company

And this is a common way to charge the wealthy more for the same product. Someone making $500K+ is less likely switch ISPs every year, or even spend the time to threaten their ISP they are going to switch.

That's doing it by proxy, though, and isn't what I took OP's question to mean. S/he specifically mentioned a comparison to income tax, which is much more direct.

Because price discrimination is the producer's attempt to poach part of the consumer benefit of trade for themselves, to take more than their fair share of the economic specialization. (To be explicit, "fair" is when the whole area below demand and above equilibrium price goes to the consumer, and the whole area above supply and below equilibrium price goes to the supplier.)

And it requires that suppliers be able to know or discern things about their customers. If it becomes too severe, then the consumers that were overcharged have an incentive to either deceive their suppliers (as in the article) or to discriminate in the same way for the good or service that they supply (such as by adding a 50% surcharge to whatever it is you sell, just for known Uber owners).

It is economically more efficient for suppliers to always sell at the same price for everyone, and to not know anything about the customer other than the color of their money, but that is not a Nash equilibrium. A supplier can always get ahead by discriminating, if everyone else is not. It is therefore beneficial to discourage the practice by statute or by social approbation.

Income tax is often progressive, but doesn't affect any one good or service in particular, beyond financial instruments. We basically charge richer people more to have a higher savings and investment rate.

And mostly, people don't want to have to haggle, use coupons, and read sale flyers every time they buy anything.

This isn't a moral objection, exactly, but if every company did that for every product or service, being "rich" could potentially be meaningless. Like, getting a raise at work would just meaning paying more, so the net impact is zero.

Result of that would be disincentive to do anything that raised one's wages, I suppose.

That's assuming the prices go up linearly with your salary.

Strictly speaking, if they did that then making more money would make you poorer, because of progressive income taxes.

Game theory suggests that at least one company would defect and offer a reasonable price (with high margins).

So it's not infinitely elastic.

> why shouldn't we charge people who are more affluent more than those who are poor?

Some places have this thing called "public transportation" that is typically financed in this way.

...by forcing possible riders to pay, whether or not they use the service, although your point is an astute one.

The affluent support public transportation to reduce congestion, thereby easing their own commutes.

For a lot of public transportation, the price of the ticket doesn't fully cover the operating costs of providing the ride.

Doesn't this already happen online? Websites can figure out you're on a Mac or in certain zip codes, etc, and some of them alter pricing depending on that metadata.

I don't know how you could bring that into the physical world successfully. But I always thought it would be cool if gas prices changed depending on the type of car you drove up in. Or if ATM fees changed based on how much money you have in your account.

It's basically the reverse of gouging people with low credit scores.

Charging affluent people more would be handing your affluent customers to your competitor. I don't think shareholders would be appreciative

Because the freedom to switch is often a luxury. At the minimum, it requires foresight, which many lack. At the maximum, it requires buying the same thing twice.

Change pricing to "n% of AGI"? How much for the loaf of bread? .0000012 Tesla? .001 etc?

Uber has no idea if you are poor or not. Even if you want more price segmentation (maybe okay) you certainly can't ask each passenger their income. And Uber isn't indicating if a price is cheap or not, creating a very large disadvantage to the poor as they are unable to accurately judge the expected value of the service before planning their method of transportation somewhere.

>Uber has no idea if you are poor or not.

Seriously? Of course they do. They have your online id and your credit card so they know more about you than you do.

No, they don't. They have indicators and clues about your income based on your rides and credit spending at uber, but they do not know if you are poor. All those indicators could and often are very wrong if that is what they draw assumptions from.

Their indicators can be wildly off because they do not know the person themselves and they can only make guesses.

Also,

> Seriously?

Yes, seriously. Of course seriously. Stop being insulting and rude.

I would have thought that they know enough to make a decently accurate guess.

Not that I know what they do of course, but without much thought I can think of a model that would allow me to predict income.

- They're using Uber, that cuts out a decent chunk of poor people.

- What areas do you visit and at what times?

- Property prices in those areas, are they workplaces or urban commuter areas? Can likely predict value of someone's house if they use it from home.

- Type of Uber taken

- Size of tip

- Frequency of travel

- etc. etc.

I reckon that given some time and some people's ride data I could estimate their income, house value, and even perhaps job. I would guess Uber have done this and then some.

Of course it's not going to be 100%, but Uber don't need that.

It's not important if a specific person is poor, it's just important to be right often enough to offset the expense of being wrong. Since they have your billing address, it's pretty trivial to look up the median income for the zip code or even neighborhood you live in, or the average sale price of the home in your neighborhoods, etc, to figure out which wealth/income bracket you are in.

Many companies provide household income data (e.g. Acxiom, Epsilon). It's available to anyone who'll pay for it.

There is nothing to suggest Uber "[does] not know if you are poor", they could very easily be buying this data on their users.

Hopefully the various comments on this thread have shed some light on why I expressed incredulity at your assertion. Incredulity isn't intended as rudeness but I unreservedly apologize since you took it as such.

Reading your response I suspect you failed to account for the fact that what "Uber knows" is not limited to the set of facts that you choose to disclose to Uber. They can and likely do use those facts to lookup your information in third party databases. They also know facts that you may not have chosen to disclose but unwittingly did so.

What about having a (potentially temporary) card number gives Uber knowledge of a person’s annual income?

Where you're coming from and and going to. Your name and home address.

It doesn't need to be 100% perfect in all cases to be useful.

that doesn't answer the question, which is about credit cards.

Then I don't understand the question, but you can probably make pretty decent guesses about a person's income just by looking at the type of credit card and the issuing bank. And definitely if you include credit card billing address. There's tons of databases that translate a zip+4 to a median household income.

I don't know about temporary card numbers, which afaik don't exist where I live but : the data-industrial complex indexes people by credit card numbers so given the number you can get a full profile. Income, etc.

Data brokers like Experian can tie that number to the rest of your profile. Even something like "this is an AMEX" can give you some info.

The know where you live too which is a bigger indicator.

And specifically where you're going, which likely includes your employer, your home, your friends' home, and probably some restaurants you like

They also know the model of device you have, its retail price, and possibly what cellular carrier you have.

They've also got your e-mail address, which has some predictive value for income. And access to your contacts, which likely includes the names / e-mail addresses / mailing addresses / phone numbers of people with which you interact. If you have an entry for yourself with your address, they can probably know your home's value and whether or not you're listed as an owner.

With enough cross-referencing between users, I imagine they could get a reasonably good guess.

how do they get access to your contacts? why do they get access to your contacts? why do people give them access to their contacts?

https://play.google.com/store/apps/details?id=com.ubercab

    This app has access to:
    [...]
    Contacts
      * find accounts on the device
      * read your contacts

But why?

I'm pretty sure it's used to assist with selecting other accounts for fare splits, so that could be one justification.

ah... didn't think of that one... Clearly not an Uber user here. At least there's a possibly helpful reason to do it.

I'm on iOS and signed up for Uber long before the privacy scandals went mainstream. iOS would have required that I be prompted for access to contacts, but I don't remember what the justification was.

I do know it knows where "Home" and "Work" are without me explicitly telling it. I assumed I'd be able to type in the name of a contact and it'd resolve that to the contact's address I have listed (an actual valid use case), but I just tried that and it didn't work. So I don't know OTOH what valid justification they have for using it.

I do see if you go to invite friends (for "free rides"), it allows you to connect contacts to send codes to. I don't think that's how I got prompted, though.

New customers are likely enticed with the low end of the rates, and as you use Uber they learn a lot about you. If you're using Uber to get from a low income area to the rail station or to Walmart, you are poor. If you use Uber to travel from and to various international airports, you can probably afford to pay a lot more (or somebody is paying a lot of travel costs for you).

> low income area to the rail station or to Walmart, you are poor.

That is a wickedly evil definition of poor and extremely distopian if you truly believe that. Your scenario does not make someone poor or define them as poor or anything like that. It is a data point that Uber can use to guess your income but it does not mean you are poor. To define poverty like that is so extremely mean and also possibly straight false.

Of course that's not a good way of defining poverty. For all I know, somebody living in a low income area and shopping at Walmart could be a billionaire that lives frugal. But to Uber that doesn't actually matter: somebody who seems to spend little money on living expenses or travel won't spend a lot on Uber either, and likely wouldn't stay a customer if Uber inceased prices for him. In the end, spending is much more important in this context than actual wealth.

They have an idea. You signed up with a credit card and a phone. Experian will answer the rest.

>Uber has no idea if you are poor or not.

I would think that there is no way they don't know that. They know where you live and which bars/restaurants/whatever else you frequent via uber. They might even know where you work if you've ever taken an uber to work. I'm sure they've either purchased from a third party or taken off your phone all sorts of other data they use to build a customer profile on you.

> Uber has no idea if you are poor or not.

Wow. This is just wrong. Uber knows where you live and its trivial to link that to general income level, either indirectly based on zipcode/neighborhood or directly via public property records.

See https://en.m.wikipedia.org/wiki/Acorn_(demographics)

>why shouldn't we charge people who are more affluent more than those who are poor

Because in the case of taxes, you are paying to a collective fund to advance collective interest.

Paying more for some product or service will just fill the pockets of the owners. It is just thievery.

Businesses try this, look up "capturing consumer surplus".

We apply this idea in a lot of places. It's basically the entire idea of market segmentation: get everyone to pay exactly as much as they are able/willing to pay. What Uber does is notable, but I don't see anything evil here.

The mechanics are different, but the goal is always the same: Store brands that are just repackaged brand products sold cheaper, slightly more comfortable travel experiences that cost disproportionatly more, support contracts and "enterprise features" for absurd prices, lootboxes in games, etc.

It's usually the opposite. More affluent people usually pay less for the same goods and services. If they didn't they would cease to be more affluent.

Sure, but they do this by avoiding usurious debt, not from getting a "rich guy discount".

Costco is the epitome of a rich guy discount. If you can afford a small membership fee and have some reasonable financial planning skills and live in the suburbs you get a discount on everything. Just one example of many.

Reply on news.ycombinator.com