Will “surveillance pricing” help or harm consumers?

Companies track your data online. So it's no surprise that they know a lot about your habits and preferences.

Now, they're trying to use that data to offer different prices to different customers, for the same items.

Today, On Point: Will “surveillance pricing” help or harm consumers?

Samuel Levine, director of the Federal Trade Commission’s Bureau of Consumer Protection.

Lindsay Owens, economic sociologist. Executive director of the Groundwork Collaborative, a DC-based progressive think tank focused on public policy.

Sanjog Misra, professor of marketing at the University of Chicago Booth School of Business.

Part I

MEGHNA CHAKRABARTI: Back in the year 2000, Amazon launched a quiet little experiment. For six days, between August 31st to September 5th, Amazon varied prices on 68 DVDs. DVDs. It tells you how far back this was. But anyway, the prices varied from customer to customer for the same DVD, but people felt something didn't quite smell right.

So they compared prices in chat rooms and were outraged. Amazon issued refunds to almost 7,000 customers because of this. The company also issued an apology. "This random testing was a mistake, and we regret it," founder Jeff Bezos said in a statement. Quote, "We've changed our policy to protect customers."

And then he added this clarification, quote, "We never tested, and we'll never test prices based on customer demographics." A quarter century later, that customer uprising kind of seems almost quaint, doesn't it? These days, Amazon changes prices across its platform millions of times a day, but the company insists, just as Bezos did back in 2000, that it never changes those prices based on customer demographics. Dynamic pricing is still based on that ancient economics fundamental of supply and demand. Even if the way that demand is measured, and exactly how a market is defined, has become much more precise and personalized than ever.

In fact, for years now, that's what marketing and pricing consultants have called it. Quote, personalized pricing. It's become AI driven technology that follows you around the internet, gathering data on where you go, how long you linger on a page, where you're physically located, what your social media posts are, who you know, when you get paid, and more.

In short, the technology allows companies to understand what you care about and set prices on your individual willingness to pay, rather than on a set value of the product or market demand. We all know that airlines vary ticket prices on a variety of factors. That's no surprise. But what if we're in a world where two travelers, both bought their tickets at the same time, both sitting in economy, same legroom, same flight, same ticket class. But they're paying different prices because of who they are. Grandma in 23C, desperate to see her grandkids in Chicago, is paying $475 for her seat. But the traveler in 24C, who decided to visit the Windy City on a whim and is unlikely to ever go there again, is paying only $375 bucks.

LINA KHAN: For years now, Americans have had their personal data and information closely tracked and surveilled by a whole set of companies in ways that can really threaten people's privacy. What we see now is that firms could also be using this enormously personal data about people to also set person-specific prices based on what they know about you, based on what you're feeling, who you're seeing, what you're browsing, and so we want to make sure that we are looking under the hood and understand what's going on here.

CHAKRABARTI: This is On Point, I'm Meghna Chakrabarti, and that was Federal Trade Commission Chair Lina Khan on Bloomberg last month talking about what she called the quote, shadowy ecosystem of pricing middlemen. It's also called surveillance pricing, rather than the friendlier personalized pricing. The FTC announced last month that it's launching a probe into eight companies to learn how A.I. is used to differentiate prices based on individual customer characteristics. It's important to note that the FTC is not taking any law enforcement action in association with this probe. And industry experts say that surveillance pricing can actually provide more value to customers, which we'll explore more in a moment.

But let's start with Samuel Levine. He's the director of the Federal Trade Commission's Bureau of Consumer Protection. Sam Levine, welcome to On Point.

SAM LEVINE: It's great to be here, Meghna.

CHAKRABARTI: What exactly is the FTC hoping to learn in this probe into surveillance pricing?

LEVINE: Yeah, so you know what you said at the top is absolutely right.

Most Americans are used to seeing one product advertised for one price. That product is available to me for $10, it's available for you for $10, it's available for your neighbor for $10. But surveillance pricing threatens to upend that really fundamentally, in all the ways you said. I can be charged a different price than you because of maybe the iPhone I have, my location, my demographics, credit history, my friends, and that threatens to really upend the pricing models in ways that the FTC wants to make sure we understand.

We know that companies have collected a vast amount of data about consumers over the last two decades. And we have real concerns that this data could be used to target people with prices based on their ability to pay in ways that end up ripping people off or undermining competition. So again, we're not accusing any of the companies who sent these orders of wrongdoing, but we want to make sure we understand what's happening.

We want to make sure that these practices are not threatening consumers or threatening fair competition.

CHAKRABARTI: Now should specify that the eight companies that the FTC has requested information from, and that's what this is, right? A request for information on practices. But those companies include MasterCard, JPMorgan Chase, Accenture, McKinsey, the other consulting giant.

Then there's some interesting ones that maybe people haven't heard of. The software firm TASK, which has McDonald's and Starbucks as clients. Revionics, a company that works with Home Depot, Tractor Supply, Hannaford, the grocery chain. And then there's BloomReach and PROSE, which at one time was named Microsoft's internet service vendor of the year.

So these are companies that provide these, let's say marketing and data tracking services, to what, a huge number of industries that Americans interact, or companies that Americans interact with every day.

LEVINE: That's what we're trying to find out. But what we already know is yes, that's exactly right.

We know that these firms serve industries, including grocery chains, dollar stores, retailers, quick service restaurants. I think just your reading off the list demonstrates the breadth of the services these companies provide. And our hope with the study is that by getting more information from these companies, information on what services they're offering, what data they're relying on, who their customers are and how it's impacting prices. We can get a broad visibility into how surveillance pricing is working its way across the economy.

CHAKRABARTI: Now, I think a lot of people might be, are already familiar with dynamic pricing, for example, in our editorial discussions, the one that comes up the most is say for Uber and Lyft and ride hailing services like that.

But it sounds like what you're talking about with surveillance pricing is different, because it's so personalized to an individual's desires, and wants and willingness, versus even local market demand. But I do wonder how that is different than practices that have already been in place for a long time.

I was thinking the other day when I go to my local pharmacy, for example, which happens to be a CVS, and I buy toothpaste, and I swipe my CVS card. I get that, I would still get it on paper, but I get this two-mile-long receipt with specific coupons that are clearly customized for me, which means that I have the opportunity already to pay something different than someone else who might buy their next tube of toothpaste.

How is that not a potential concern for the FTC?

LEVINE: Yeah. So it's absolutely right that we already see customized pricing. Even if you go back to the days when you clip new coupons from the newspaper, those coupons were geared toward more price sensitive customers. I'm a person who also likes my CVS receipt printed so I can try to use those coupons geared toward me, there's no question that these practices already exist in some form. I think our concern with this study, and it's based in part on how these services are advertising themselves, is the granularity and the degree to which they can be personalized and adjusted in real time.

It's true, CVS can look at purchases over time and retails tourists can do that generally. But what we know based on the FTC has been involved in data privacy work for decades, companies are collecting a vast amount of information about us from the websites we visit, our credit history, our location, our devices.

This information can then be aggregated by data brokers, and it can allow companies to get a really complete picture of our ability to pay, our needs, our desires. So I think what we might see, and what we have concerns that we're going to see, is pricing that's a lot more targeted. So for example, a company might know that my payday is this Friday, and charge me more for products that I'm trying to buy this Friday, maybe more than they would charge to you when your payday is a week earlier.

So that kind of really granular targeting, I think, moves away from traditional coupon clipping and toward a model where it's no longer the case that a price advertised to me is also available to my neighbor.

CHAKRABARTI: How would the FTC determine where practices cross the line in terms of violation of privacy or unfair pricing?

Because what you've described is the world that we're already in, essentially, and there's no putting the genie back in the bottle, in terms of how much data we're individually spilling out there almost every second that we're online.

LEVINE: Yeah, it really is very pervasive already. You're absolutely right.

The way I think about it is that we've had this unchecked commercial surveillance, this data collection about Americans for more than two decades now. We already know that it's really undermined our privacy. I could look you up online. You could look me up online. You could find a lot of information.

You could go to a data broker, get even more information. But I think what we're seeing now is practices that have threatened our privacy now threaten our pocketbooks. Now we face this possibility that because of all this data being collected about us, we're being charged more because of who we are or the devices we have, or that companies are able to collude, that they're able to see how much people are being charged in a certain region.

So for example, there was reporting about a decade ago about an office supply store that was charging more in locations where they didn't have competitors, that was exposed, the company changed its practice, but that's the kind of practice we want to make sure is not harming consumers. If it is harming consumers, if it's causing them injury that's not avoidable, that can be unfair under the FTC act. That could also be a potentially an unfair method of competition. So we have a number of tools to determine whether the law is being broken and to protect the public. Again, the goal of this study is really just to lift the hood, understand what's going on.

Part II

CHAKRABARTI: The FTC announced last month that it's opening a probe. It's done what's called a request for information from eight major companies in the United States that advise and help other companies find out all the data they can about you.

And based on that, are potentially putting out their prices personalized to you. The FTC is wondering if this is a violation of your privacy, and perhaps even unfair business practices. And here is how FTC chair Lina Khan described, let's say, a theoretical example of how surveillance pricing could look if implemented.

LINA KHAN: For example, you can imagine that a hotel that knows you've already bought plane tickets may show you a higher price than they show somebody who hasn't really decided on where they're gonna visit. You can imagine a restaurant that knows you have a one-hour lunch break could hike prices for you during that hour.

And so these are just some of the types of practices that we could see take hold if this type of surveillance pricing is allowed to develop.

CHAKRABARTI: That's FTC Chair, Lina Khan, last month. And I should also note that she's highlighting examples of when companies, based on their really granular knowledge of you, could potentially hike prices, but the opposite potentially is true as well.

They could drop prices for you. We'll talk about that in a moment. But joining us now is Lindsay Owens. She's an economic sociologist and executive director of the Groundwork Collaborative. That's a Washington, D.C. based progressive think tank that focuses on public policy. Lindsay Owens, welcome to On Point.

LINDSAY OWENS: Thanks for having me.

CHAKRABARTI: Okay. So how far back does any form of personalized pricing or surveillance pricing go in American history?

OWENS: Yeah, personalized pricing is certainly not new. So there are a couple of markets that Americans have become quite familiar with, that have used personalized pricing really since their advent.

So the first one that you might think of is the credit score. Americans are charged different prices for borrowing money, depending on a whole host of personal characteristics that the credit bureaus collect about them. The second example that Americans are quite familiar with is the insurance market.

Your insurance premiums, the cost you pay for things like homeowner's insurance, auto insurance, health insurance vary based on a whole host of information that these companies collect about you, and what they deem as your risk profile for these goods.

I think what is so different about the broad expansion of personalized and surveillance pricing that the federal trade commission is so rightly taking a close look at, is there really is not a business case for charging me more than you for a sweater at the Gap, right?

This is not, Gap is not assuming greater risk when I buy a sweater than when you buy a sweater. And I think the sort of personalized pricing and surveillance pricing expansion that we're seeing across the economy is broadly concerning, because ultimately, it's not about covering their costs.

The sweater costs them the same to make whether I buy it or you buy it. But rather it is about extracting more profit, extracting more from me, if I'm willing to pay a little more for the sweater than you are. So I think this is really different than what we've seen in the past, and quite concerning.

CHAKRABARTI: Yeah, the contrast to the insurance example is a good one because, like you said, it makes sense, because insurance companies have to measure risk. Because that's going to be a direct correlation to the cost that they will incur in covering those risks. Whereas perhaps the only risk that goes along with whether or not you're going to buy a sweater at the Gap is a fashion risk.

But I'm also seeing though that, look, for as long as business and capitalism has been thriving in the United States, that some form of personalized pricing or surveillance pricing has been a dream of companies. That, for example, the Interstate Commerce Act of 1887, in part, was brought about to prohibit special rates and rebates and preferential treatment by the railroads.

So it's pretty old, but there are some also more recent examples, which are quite eye opening, that maybe folks didn't know about. For example, in the 1990s, Coca Cola tried something about thermometers and vending machines. Do you know about that?

OWENS: Yeah, absolutely. So this is a really famous example that I think highlights one of the ways, one of the important ways in which this practice is curtailed.

Of course, we need policymakers to step in, but I do think consumer backlash can be a really effective deterrent at moving companies off some of the more unseemly personalized pricing strategies. So Coca Cola is a great example. In 1999, the CEO of Coca Cola announced that they were considering installing thermometers on their vending machines.

And basically, the thermometers would allow Coca Cola to increase the price of a can of Coke on a hot summer day when more people were interested in a cool beverage. And the consumer backlash was incredibly swift. Coca Cola immediately reversed course, and still today,  there are no thermometers on Coca Cola vending machines.

CHAKRABARTI: But even that example, though, I would say that example perhaps is would be more accepted today than it was in the '90s, right? Because we essentially again, going back to the ride hailing app situation, people do pay for surge pricing, right? Depending on whatever circumstances lead to higher demand at the moment.

There's another one that I did not know about, but that has to do with the Princeton Review charging more for SAT prep courses, based on what?

OWENS: Yeah. So this is a finding that a journalist discovered, that Princeton Review was charging more based on zip codes. And in particular, zip codes that contained a high percentage of Asians were being charged more for Princeton Review test prep services for college prep.

Tests like the SAT and the ACT, they have said that they abandoned the practice, but they now appear to get most of the information they need for differential pricing actually from users IP addresses. And so IP addresses, as well as information about what kind of computer you're logging on, those are the types of data that companies now collect and use to personalize prices.

CHAKRABARTI: Has it reached a point, though, where the personalization can become so minute that, for example, people sitting in the same restaurant buying the same meal, or even, this is actually a more realistic example, using McDonald's app to buy a Big Mac, that they could be paying two very different prices at exactly the same moment.

And if that's the case, how is anyone supposed to know what the actual price of a Big Mac is or should be?

OWENS: Yeah. So this is the fundamental question. When the price tag was rolled out and more or less invented with John Wanamaker and his department store in Philadelphia in 1861, prior to the price tag, you really haggled for almost anything.

With the public price tag came a uniform standard, transparent public price. And I think what we're seeing today is really the erosion of that public price, maybe even the death of the price tag, if you will. And a key piece of how this is being executed is via isolation. If you and I were standing in line at the Gap together with matching pink sweaters, I think that we would be a little surprised if one of us paid $15 more than the other.

But if we're sitting at home in our couch, logging on to the Gap website, I have no idea that I overpaid and that you got a better deal than I did. Similarly, if we're shopping for groceries online and then picking them up in the store, but the transaction and the pricing information takes place on our phone, or on our computer, or in the Safeway app, or the Kroger app or the Albertsons app, we don't realize we're being offered a personal price, a personalized price.

And so I think isolation has been really critical to the broader dissemination of personalized pricing, because it obscures the fact that something unfair is happening, that price discrimination is taking place.

CHAKRABARTI: So Lindsay Owen, stand by for just a moment. Because I want to introduce Sanjog Misra into the conversation.

Sanjog is a professor of marketing at the University of Chicago Booth School of Business. Professor Misra, welcome to On Point.

SANJOG MISRA: Thanks for having me, Meghna.

CHAKRABARTI: So I want to explore actually both with you and with Lindsay, what the business case for personalized or surveillance pricing is. But before we do that, you've done some really innovative research on the effects of personalized pricing with ZipRecruiter.

Is that right?

MISRA: Yes.

CHAKRABARTI: So can you describe what that project was?

MISRA: Sure. So just let me start by saying, actually, ZipRecruiter did not implement personalized pricing. We tested it out and we were intrigued by what the outcomes were. So we typically think about uniform pricing as being fair, and it makes sense, right?

Everybody gets the same price. What people often don't look at is what the outcomes of such a pricing policy are. And what we found was that small businesses, who are unable to avail of a service because prices are too high. So think about a uniform price being too high. If the prices are too high, there's a lot of small businesses who can't afford the service.

They get cut out of the market. And so when you personalize prices, yes, it's true that some large businesses, these are, the ZipRecruiter's clients are businesses, right? And so some large businesses ended up paying sometimes a little bit more, sometimes quite a bit more, but there were about 60% of small businesses that were cut out of the market that now were able to avail of the service, simply because we could tailor the prices to them.

I think that's the broader theme here in terms of, I think the question that's being asked about whether or not data is being used. And for what purposes? I think it's a valid question. I think on balance, it's also important to think about what the consequences and outcomes downstream of personalization are.

And can it actually help consumers?

CHAKRABARTI: So for clarification on what exactly the experiment with ZipRecruiter was, there was an algorithm that was ZipRecruiter deployed, right? In order to personalize prices for its customers based on a series of questions. Now correct me if I'm wrong, but I note that as a result of doing so, for that period of time, ZipRecruiter's profits rose by 84% over the old system, yes?

MISRA: Yes. So part of it obviously is the fact that even the average uniform price that was being charged was too low from a profit maximizing perspective, right? So the firm was, as a startup, sometimes you start with low prices just to gain some traction.

So they, deliberately or mistakenly had suboptimally low prices that they were charging. So even if you had not personalized, profits would go up by about 60 to 70%. We show that in the paper, the personalization piece gave them an extra maybe 10 to 15% over that. They choose not to implement that, just to be clear.

And so they gave up on those profits. And what they found was, what they learned from that, though, there was heterogeneity in kind of what, how people valued the product. And by offering a menu now, where smaller businesses got fewer, a kind of whatever you call it, a simpler plan, and they got a lower price and larger businesses who wanted a lot more got a higher price that gave you kind of profit.

So that's the pricing they use today, right? It's also tailored on characteristics, but now the choice goes to the consumers. But we wouldn't have learned that, had we not done the first study.

CHAKRABARTI: So there is some form of customization, but it's not the same as when the study was done. Understood. Now I also see that as a result or soon after the study, you became an advisor to ZipRecruiter.

That's, are you still advising them in any capacity or you're no longer doing that?

MISRA: I'm not.

CHAKRABARTI: Yeah. Okay. Okay. So let me turn back to Lindsay here for a moment. Because look, one of the arguments that I think is a positive one for the kind of personalized slash surveillance pricing we're talking about is what Professor Misra just said, that we're focusing on the potentially unfair practices, but there's a whole other world in which this kind of very customized pricing could actually open the door to consumers, whether they're individuals or companies, to products they may have not otherwise had the opportunity to purchase.

OWENS: Yeah, so it's absolutely the case that the same technologies that are being used to jack up the price on consumers can be flipped into reverse and also used to offer discounts to consumers. That's certainly the case. And I think there are a couple of things worth unpacking here.

The first is, what is the role of policy makers in taking a look at practices like this? And I think policy makers are really focused on preventing harms. Are these practices being used in uncompetitive environments, where we're seeing a lot more of the technology being used to dial up and gouge or surge, and a lot less on the discounts.

And also, there are questions here about whether or not any of this is transparent, right? Consumers might be okay with some of these practices, but not know that these practices are being used, or not know that their data is being collected and used in this way. And so I think even in the case where these technologies are being used to offer discounts. I think there are still important policy questions, public policy questions to answer. The other thing that I think is important here is, ultimately, who sets the rules of the road when it comes to pricing and markets. And I think when I take a look at these examples, while of course, surveillance pricing and personalized pricing could be used to steal from the rich and give to the poor, right?

Charge wealthier people more, offer poor people discounts. The history of American capitalism, I think, is filled with more folks who fall into the kind of Robber Barron category than the Robin Hood category. And I think, this idea that personalized pricing is going to bring about sort of progressive pricing regime, filled with Robin hoods is a little bit hard to swallow, particularly given the fact that what we've seen over the last four years is a story of price hikes and price increases.

This is a story of price over volume, to use Wall Street speak. This is not about selling more glasses of lemonade. It's about selling each glass of lemonade at a higher price point and a higher profit margin. And most recently, that's what we're seeing personalized pricing being used for.

Part III

CHAKRABARTI:  Professor Misra, as I promised, I'd love to hear your response to Lindsay, who just to summarize, basically said that there's no guarantee that this kind of technology or pricing strategy would be used to actually help consumers instead of just raise profits.

MISRA: No, so I completely agree with that. So let me just step back for a second. There's kind of two harms that kind of legally we look at. One is obviously the violation of privacy, like your data being used without your permission, without your consent. And the other is, whether or not this leads to unfair business practices.

And to the extent that's what's being investigated. I'm all for it, right? I ... completely understand that any new technology can be used for good and for bad. I think the balance is what I'm preaching here, if you will, which is to say, we need to look at both sides.

The fact that it can be used for good, it can be used for bad. And if we can design policies, or we should design policies that help us or steer us in the right direction, just to give you, go back to your example of the grandma and somebody else flying to Chicago, right? Imagine that the grandma is on social security and cannot afford to pay, right?

And we now say we are not allowed to use data to personalize prices. The fact of the matter is that individuals who are poorer are willing to trade their data, in order to get better deals, in order to avail of themselves of any price reduction that might help them. The SNAP program is a personalized pricing device.

And in fact, I've worked with the SNAP program. And so this is the food stamps programs, more colloquially, and people consent to give all of their data just to get food on the table. And had, suppose we did a personalized messaging algorithm there just to get people to stay on the program and recertify and fill out forms, but it's the same technology.

It's the same idea. And if there's a blanket policy of saying no, you're not allowed to use personal data, that would have not the kind of impact that we want again, just to, I completely agree with the point that if it's being used for unfair business practices, we should look into that. One other point I'll quickly make, think about buying a car, before any of the technologies that we had, prices were personalized, you went in and you haggled with the salesperson and the same car, bought on the same day with the same exact specs were sold at different prices.

They continue to be sold at different prices. My point is not that practice is good, but why is that allowed? It's allowed because other dealerships, other car manufacturers are also allowed to do it and it fosters competition. And now prices might actually be lowered. Is there, do people end up paying some, a few people end up paying higher prices?

Obviously, they do. There's a continuing research project that me and my colleagues here at Chicago are working on is to look at how competition interacts with personalization and what we found is invariably that when you have competition, multiple manufacturers, multiple providers will actually personalize in lower prices because that's competitively the right thing to do.

Again, I'm not saying that this is all, like there's a bunch of Robin Hood's out there. I completely agree. There are not. But at the same time, we have to look at the downstream impact and then think about, are our policies going to impact the disadvantage, the minorities, the people who are poor or specifically because if we have uniform prices, they will be the ones who will be impacted.

CHAKRABARTI: About that car buying analogy, it's one that's important, because a lot of people are familiar with it, but I would say the difference there, though, is that, A, the price, that the sticker price originally, everybody knows the sticker price on the car, which isn't necessarily the case with this surveillance pricing.

And B, as the consumer, you have the opportunity to haggle. I don't know how the haggling works when you're just given a price, let's say online for an item that you need, and you have no idea what any other potential price is, and how are you going to haggle anyway? It just is what it is, based on the surveillance done by the company, Professor.

MISRA: So I think we underestimate the power of the consumer here, right? So what's happened with the automobile sector is a great example. Prices have fallen, negotiated prices have fallen in the automobile sector, because consumers now have access to technology, too. They do search a lot more and then they go armed with that information to negotiate prices down and there's tons of evidence that's happened.

Now, the same thing when we buy online, if you think about just how we purchase, we don't just go to one place and say here's Gap and they've given us, given me a particular price for a particular sweater. I'll just pay that. I go digging for coupons to see whether or not I can find a Gap coupon.

I go looking to see whether or not there are competitors. And since all the competitors of Gap are also monitoring the prices. And this again, research on this too, that shows that competitors continuously monitor prices and then strategically drop prices when they see their competitors have higher prices.

So there's a lot more going on in terms of how algorithms and pricing interact than just this idea of firms are going to collect prices and then charge me a higher price, which may happen again.

CHAKRABARTI: Yeah, I want to go to Lindsay here for a quick second here, but just to clarify, has the technology, AI technology, become sophisticated enough that competitors can monitor the individual prices, that one company can monitor the individual prices that its competitors are giving to its millions of customers?

MISRA: No, that we are not, you're not allowed to do that, first of all, legally, that would become an issue, but they're conditioning on the same set of variables. So we know, think of what's happened now as being, we have markets which are micro markets, right? Now you can think of micro markets as being, if you're in a micro market of one customer, let's say, we can jack up prices.

But the catch is that everybody else is in the same market as you, so competition actually intensifies. And so what might happen, again, there's no, I can't say what the equilibrium will be, but it could be that competition intensifies and prices actually drop. Because now, because there's a lot more people competing for your attention, for your demand.

CHAKRABARTI: But it seems like all of this is really predicated upon the availability of the information which, as you said, there's actually some laws in place that may prevent that. But Lindsay Owens, I know you've been listening patiently here. I'd love to get your response to this, really the professor's argument that people can still essentially vote or exercise their desires positively or negatively based on where they shop, no one's forcing them to buy the products that are given to them at their individualized prices. And wouldn't that have a positive effect here?

OWENS: Yeah, absolutely. Before I answer that question, I did just want to go back to the example of haggling for a car. I actually think this is a great example of a place in which price discrimination is rampant and not welfare positive.

We know for a fact that women pay considerably more for cars than men, precisely because of this haggling, precisely because of this gray area there is to set the price differently based on sex characteristics in this case, this type of discrimination. And so I don't think that personalized pricing, the history of personalized pricing is really one of benefiting customers, I actually think in many cases, what we see is rampant discrimination on a whole host of characteristics, including things like gender and race.

Just one point to make on the car example. But on your point that consumers can vote with their feet, I think again, we need to go back to this issue of isolation. Many customers have no idea that they're being served a personalized price. When you're shopping online, frequently your user interface looks different than somebody else's user interface.

That's a practice called dark patterns, where companies can slot you into a very specific effectively display screen, where you get a different price than your neighbor. And so it's pretty hard to beat the machine if you will, if you have no idea what you're up against.

Similarly, when prices are changing by the nanosecond, it's going to be tough to do thorough, systematic comparison shopping. So I always say when I get this question of what consumers can do to avoid getting gouged by surveillance pricing or personalized pricing, it's really not each consumer's job to beat the machine. It's policy makers job to make sure the rules of the road are fair, and that the machines are playing by fair rules of the road. And I do think public pricing, it's uniformity, it's transparency. It's predictability. Remember, people are on budgets when prices are shifting, when they don't know what they're going to be charged for an item that they're used to paying for, daily or weekly.

That can really impact the family's ability to put food on the table, to make ends meet at the end of the month. And I do think there is a lot to be said for public pricing, the fairness it provides and it's a reason why it's been common for 150 plus years now and why I think every time consumers find out about new forms of surveillance and personalized pricing, there's considerable backlash.

CHAKRABARTI: Yeah. And we've also seen in the past reports of differentiated pricing working to the disadvantage of the already disadvantaged. There's been investigations on how, for example, one where broadband internet to a million different residential addresses, that the worst deals were given to the poorest people in those million addresses.

And you talked previously about consumer credit being a really good example of that as well. But, Professor Misra, I just want to go back to something you said a second ago, about micro-Markets, because I feel that we're in a age now where one of the fundamental premises of how a market achieves fair pricing is being completely turned on its head, right?

Everyone who takes Econ 101 knows about that supply and demand curve, and it's an elegant idea that where those two curves meet, that is the fair market price, but it involves having to have a market. When you talk about a market of one, how do we even define what fairness is in pricing?

MISRA: That is the question.

So just a couple of things is to get to this. But just very quickly, this idea that personalized pricing is a relatively new idea. ... Talked about this in 1938, this idea of a market of one. It's not my idea. It goes back to the '30s. Let me just flip this on its head for a second.

I think what we've been talking about is all the examples are, here's how data can be used to make things worse for you. So recently I parked my car at a spot which I didn't realize, I shouldn't have parked it there. My car got impounded. I went to the pound in Chicago, and I could afford to just pay the $200 and get my car out. What I noticed there was a number of people trying to argue with the impound officials about that they can't afford to pay that money. They couldn't even afford to get there to that lot, to get their car out and their jobs and their livelihoods depended on that.

This is an example of a fair price, a uniform price, which causes immense amounts of hardship. Now, had you been allowed to tailor prices based on the degree to which a particular citizen is disadvantaged, you might have been able to increase welfare. Again, I'm not saying companies will do this, but I think we have to look at the consequences of the prices that companies charge.

And in those cases where laws are being violated, or we think as a society, this is bad for our citizens. We should change our policies. I'm completely for that. But I don't know.

CHAKRABARTI: In this sense, that's what the FTC, that's the path the FTC wants to travel with this investigation. Yeah.

MISRA: And that's absolutely, I support it completely. But what I would like is a balanced view of what are the downstream impact on citizens and who are being, like, think about when we say we'll charge higher prices to certain people. Companies aren't stupid. They're not going to charge higher prices to people who cannot afford them, because then the person just drops out of the market.

The higher prices, the coupons, the rebates, the discounts, they're all going to be targeted towards people who are poorer. When Procter & Gamble tried to take away coupons and come up with a fair, uniform price, citizens actually traveled to their headquarters and picketed to get their coupons back, right?

So it goes both ways. I'm advocating a balanced approach to evaluating what the technology, what AI is doing for pricing, but also about how this pricing actually impacts our citizens, our customers. Just to put a final point on it. ... It isn't just about pricing.

It's about policies around data and privacy and how those impact downstream decisions across the board.