Jul 07, 2026
On July 7, The Capitol Forum held a conference call with Derek Kravitz, investigative journalist at Consumer Reports, to discuss his recent article, “Different Prices for the Same Ride: How Uber and Lyft Use AI to Get More Money Out of You.” The full transcript, which has been modified slightly for accuracy, can be found below.
TEDDY DOWNEY: Hello, everyone, and welcome. I’m Teddy Downey, Executive Editor here at The Capitol Forum. And I’m pleased to be joined by Derek Kravitz, Investigative Journalist at Consumer Reports. We’ll be discussing his recent article, “Different Prices for the Same Ride: How Uber and Lyft Use AI to Get More Money Out of You,” which examines pricing differences for similar rides, the evolution of upfront pricing, and broader questions surrounding algorithmic pricing and consumer protection. Derek, thank you so much for doing this today.
DEREK KRAVITZ: Yeah, thanks, Teddy.
TEDDY DOWNEY: So, I love this experiment that you did. It just seems incredibly rigorous, thoughtful, well-designed. How did you think about doing this and then structuring the investigation the way that you did?
DEREK KRAVITZ: Yeah. So, we’ve been doing pricing investigations for about 18 months, and all of them have this cross-civic sort of testing, married with investigative journalism, married with policy components to it. CR is a pretty unique organization in that sense. It’s 90 years old, and it has all these different arms to it.
And we started with an investigation into Kroger, the grocery chain. And we were looking at their loyalty program, which is free. They pretty much push everyone into the loyalty program through Kroger. Whenever you check out, they even sort of swipe you in or push you to register there, and they collect base information when you register. They create a lot of insights from that, a lot of inferences.
And we requested, through Oregon’s new law, getting access to your DSAR information, your personal dossier information. Several volunteers got their Kroger profiles basically. And we were able to see the inferences that they created, how likely you are to buy a car, new or used, how likely you are to take a vacation by plane, by cruise, et cetera, your household income, your race, your gender, et cetera.
And with that, we were able to see that a lot of these inferences were inaccurate in many cases. And a lot of them were being monetized in a clean room and then sold to one of 52 different third parties. A lot of different ad tech companies like that, but also two of the largest tobacco companies in the United States, some health startup companies that look at Medicare and pharmacy claims, a mixture of places.
And that sort of opened a door for us. We thought, well, we have a membership of 5 million subscribers. Why don’t we leverage some of them to request this type of information, collect it, try to explain it to people, with all the caveats and limitations that goes into that. And so, participatory action research, but with a corporate focus, especially with tech and business and finance.
So, we started with Kroger and then we did Instacart, which we looked at basically the same thing, differential pricing, variable pricing. Many of the nation’s largest grocery chains, we saw basically a very souped up A/B testing. There were, like, the same bunch of bananas cost five different prices at the same location at the same time. So, we tried to control for as much as we could, but we designed these very tightly controlled experiments across the country with hundreds of volunteers to isolate that. We published that story. They immediately stopped those tests and a lot of legislation came from that.
So, we thought, well, this is a worthy area of exploration. Let’s look at the pioneer of some of this tech, Uber, and then also Lyft. And we spent about six months designing, with the help of outside researchers, our own statisticians, community organizers, those folks to sort of try to explain what differential variable pricing looked like on Uber and Lyft, how base pricing might not be personalized, but discounts and promotions are, how these companies are using promotions and discounts to really sort of reset what a final price looks like and what that all means for consumers.
TEDDY DOWNEY: Yeah. So, tell us about what you found and what was surprising about it.
DEREK KRAVITZ: Yeah. So, we went across the country. We did three virtual tests where we had anywhere between 33 and 65 people all look for the same ride at the same time. Some with their geolocation settings on, some off, some with Uber One, some not.
And we basically priced out their rides and had them monitor and screenshot offer screens and also wait times and other things. And then submit that along with their demographic information, their usage information, how often they use Uber and Lyft and various other things. And then we basically analyzed all the data. We did also an in-person test because we wanted to—it’s one thing to do virtual testing and to see things in that world, but we wanted to see it realized, actualized.
So, we actually went out to Portland, Oregon, where they have a per minute per mile sort of base fare, a traditional taxi model. And we then matched volunteer riders with volunteer drivers from a pool of union-affiliated drivers in that city, all in the same place. And we all took the same rides to the same locations and then compared receipts, the driver receipts, which are pretty detailed at times, and then the rider receipts. And we mostly found the same thing.
So, that validated. We did beta testing to sort of figure out is there anything here? We did see something. We saw that 42 percent difference between the low and the high price groups, the median, which is the most conservative way of looking at that. And then we also saw that fake discounting. So, the idea of a baseline price, an algorithmic baseline price, and say an $82 price being offered, and everyone else seeing $62, and then that $82 price being discounted to $62. And we saw in about 12 percent of cases, there was some sort of fictitious pricing or fake discounting going on.
So, both of those things combined along with the take rate, how much the drivers take, all three of those things sort of added up in our minds to something, right? To something that consumers either don’t really understand or they don’t like. And when we took all this information to them, what do you think about this, the consensus, uniform response, we don’t like this. We want to know what goes into the price at the very least. And if it’s a huge disparity in price, why? And it shouldn’t be that way, right? It should be something closer to an algorithmic mean.
TEDDY DOWNEY: So, you touched on a lot of things in the report and what you just said in terms of the results. I would like to spend a little bit more time on why people think it’s unfair.
DEREK KRAVITZ: Yeah.
TEDDY DOWNEY: What is it about it that they find unfair? Which is kind of like how I think a discussion about how should our economy work? It should be fair, right? At some basic level. And can we just talk about that for a little bit? Because then I want to go into is it breaking the law?
I think the fairness concept—because when we did a similar investigation in Instacart before your thing came out, we got a similar vehement response from the company to the point where it deterred us from looking into it more, which I’m deeply embarrassed about and I probably shouldn’t admit here. But you guys went ahead and did it and proved it and they stopped doing it. And when we were asking people, they said, look, this is unfair. When we went around asking people that. Now you have more evidence around that. You’ve spent more time on it. I would love to hear about why people thought it was unfair.
DEREK KRAVITZ: I remember when The Capitol Forum published that and we looked at that. Really the only difference is we had some statisticians and also some subject matter experts just weigh in on the methodology. But beyond that, yeah, we just published only because we decided to alter our thinking from is this illegal to is this fair in the minds of consumers?
So, I do think that’s a good framing. I think that’s maybe the overarching framing you’re thinking about. Is this fair in the minds of consumers? And I think if you ask that question—and we do, we ask that question right up front, even in the beta and the pilot testing—does this make sense and is this fair?
And generally, if consumers can’t make sense of it and they don’t really understand the gears, the levers that are being pulled or the black box, then, yeah, they don’t find it fair. Especially, if it’s a few cents or within 5 percent was our metric for Uber or Lyft. They understand that. They understand GPS network signal or supply demand shifts or things like that. They understand basic Economics 101.
What they don’t understand is personal data being used and how that is sometimes used against you, right? Or your ride history, your account history. That is where they draw a line mentally. And for us, we do also a lot of surveys, statistically significant surveys, where we poll 2,000 Americans, U.S. adults, and they say the same thing. And that’s why we always include that in the stories because it’s not just our framing, right? It’s what, when we ask people, this is what they say.
And then when we go to regulators too, they largely say the same thing. They say that in many cases, consumer protection laws FTC Section 5 or the FTC guidelines on fictitious pricing or state laws haven’t kept up with where tech companies are going or have already gone for years now.
So, one thing I think it’s worth flagging about, both Instacart and Uber Lyft, pretty much everyone is okay with when we speak to them and also regulators, supply/demand level fluctuations in price or to some degree surge pricing. What they’re not okay with is—especially personal data, mainly your account history, but also your behavioral context data, how you use the device and in what way—that is where they draw a line.
And we saw that with Instacart, that people didn’t like being grouped into A/B testing cohorts, even if they were winning, if they’re on the winning side of it, because they just didn’t know. And same thing with Uber/Lyft. They just want to know what the difference is going into it. And I think if they saw that clearly, they would have a more informed sort of choice.
TEDDY DOWNEY: And what about the drivers?
DEREK KRAVITZ: That’s a whole other ball of wax. There’s been, as you know, going back years—I mean, it started in Portland, Oregon, actually with Greyball, this Uber initiative where they were operating illegally for two weeks and they wanted to sort of keep regulators off their backs. So, they created this sort of almost like fake mirror image app and sent it out just to the city officers, the regulators there, and basically denied them access to the Uber platform in order to forestall any sort of regulatory push there.
And since then, ridesharecompanies have waged very large, intense battles in cities, states, even the federal government, to push back against labor laws that would turn them from basically an agent to a principal in the SEC reporting parlance, but like basically treat them as an employer of drivers, as opposed to an agent taking a commission and hiring independent contractors.
And Proposition 22 in California, one of the most expensive ballot measures in U.S. history, sort of codified that in California law. And Uber and Lyft were on the front lines on that. And since then, Minneapolis, New York, Seattle, you’ve seen a bunch of cities and states really sort of try to tackle this and find either a pay floor for drivers or try to determine a take rate that they would get from each ride. But at the end of the day, the platforms are taking more and more of each trip. Multiple peer-reviewed studies have looked at this and seen the same things.
Other researchers have—Princeton and Columbia, Len Sherman, who’s a great researcher in this space—upwards of 50 percent of the ride now goes to the platform. And that’s a change. That’s a change from a few years ago when it was 25, 30 percent. The platforms will dispute this. They say commercial auto insurance is the large factor here and that it’s increasing for them. And that that shouldn’t count as a line item or they should line item it out. They shouldn’t count against them as a piece of gross income.
But it’s an expense. It’s a cost of doing business, right? And so, when we calculated, when unions calculated, and researchers too, we defined it as the percentage that the platform takes and the percentage that the driver takes and then government fees as a third part. And that’s how we sort of separate it out.
TEDDY DOWNEY: Yeah. What did people think about that take rate? Because I’ve been doing that myself. I don’t take Uber and Lyft that often, but when I do, I try to ask the driver how much of that did you get? And when I compare it to when I take a cab through Curb—which is not exactly like taking a cab originally, because it still has a bit of a weird pricing up front. But in general, they’re keeping 85 percent or something very high, the driver.
Now they have other costs and fees. But it’s not even anywhere in the same ballpark as what you’re talking about when you ask the Lyft or Uber driver. I’m curious to get your thoughts on what the people in person were seeing and then also what the consensus around like is this fair or not?
DEREK KRAVITZ: Yeah, it was really interesting to compare the receipts between the drivers and the riders, because there is a little bit of information asymmetry going on. The riders obviously see the price that they’re paying and a little bit of a line item breakout, but they don’t really see what the driver is taking home personally from each ride.
Whereas, the driver also has a little bit of missing information, doesn’t really see—sometimes doesn’t see fully what the passenger is paying before a promotion or discount is applied. And one thing to keep in mind, both of these companies are increasingly using promotions and discounts. A lot of tech companies are doing this. They’re relying on promotions and discounts to not just personalize prices, but also to sort of reset what the final price is, what the net price is, and in a way that really does capture willingness to pay in a way that they haven’t been able to before.
And the margins, I mean, the data speaks for itself. I mean, if you can increase your margins and your sales two to three percent or five to six percent on either end, I mean, that’s a huge plus, right? That’s a huge win. So, why wouldn’t companies do that?
But yeah, for the drivers we spoke to and when we compared receipts, they are deeply frustrated. They say it’s unsustainable, that when they first started with Uber Lyft five, six years ago in many cases—or even longer, 10 years ago for some people we spoke to—they were taking home 75 percent of fares. And they got sign-on bonuses too. And they got perks.
And a lot of these apps are now gamified. So, you get blue tier and gold tier and platinum tier. And it resets every month or every three months. And they’re gamified to constantly accept so they can balance the riders and drivers in different locations and make sure that everything is copacetic in terms of supply and demand, but also make sure that wait times are low and that riders aren’t frustrated.
But also, if there’s a weird ride from like the example that we got from one driver from Portland to Cannon Beach, which is on Oregon coast, an hour and a half drive through hilly terrain with like very little possibility of a return trip from that place back to Portland, that they’ll get a driver to do that trip. And even having to eat the cost on the way back. And that’s frustrating, right? Like drivers don’t want that obviously, for many different reasons.
So, they’re finding themselves working longer hours, making less pay. They just find it unsustainable. And their eyes are set to Waymo and AVs and them coming up and basically taking their jobs and making them obsolete. And they think that this job, this industry, will not exist for them in a few years.
TEDDY DOWNEY: And I think you also asked the citizens what they thought of the take rate too, correct?
DEREK KRAVITZ: Yeah, riders are like empathetic people, right? I mean, they want—when they pay a $30 fare, in their minds, a lot of people think, oh, $20 goes to the driver, right? Or $25. Maybe the cut for the company is $5 or $10. Many—actually, everyone we spoke to—no one appreciated how big the cut was for Uber and Lyft. And they wanted the drivers to make more.
And we spoke to one writer in Austin, Texas, who—he’s quoted in the story. He said I actually now prefer Waymo because at least I know that a real life human driver isn’t getting sort of the shaft there, in his terms, by not getting a decent cut of this fare. It’s just sort of this autonomous company vehicle doing it. And maybe it’s a little counter to what regulators and labor would want. But yeah, that’s the thinking, that people don’t like drivers not getting their fair share.
TEDDY DOWNEY: And a lot of this also comes out of, you mentioned this, it’s not transparent to the citizen, to the driver, to the regulator in the piece. You even talk about the lack of transparency around the algorithm and just the constant manipulation of all these numbers, right?
And that seems to be where the companies want to live when they’re pushing back on you. Oh, you’re not accounting for this, or you’re not accounting for this. And there’s slight changes in the data. But the reality is these experts looked at it and they concluded that, no, there’s something going on here. There’s something problematic in the fees and the sort of lack of transparency around all of this.
Can you talk a little bit about that? Because that seemed to come up, that it was the black box nature, how they get to these prices. And then parse out what you were able to sort of distill out of that black box.
DEREK KRAVITZ: Yeah, that’s a great question. So, one thing I’ll just flag just to frame that or set that up, when we do this type of testing, we obviously have to go back to the company for a request to comment, but also like a fairly robust one, like much more than any—I’ve been doing this 20 years. This is a different level of corporate engagement or request to comment.
So, I think I answered two or three rounds of questions from Uber before I asked a single question, if that makes sense, about what we did. And we shared our underlying data. I don’t necessarily share all of my underlying data with companies before I publish something. But in this case, you have to share as much as you can in order for them to meaningfully respond. And I get that. And I think that’s actually a fair place to be.
We stripped out PII. So, we stripped out, of course, names, phone numbers, some route information, in order to protect the anonymity of our volunteers, both the riders and the drivers. There’s good reason for that. I mean, no one wants their privacy being violated. And for drivers, there’s a bigger risk, which is being deactivated. So, we did all that stripping out of personal information.
Companies want as much as they possibly can get, obviously. So, they pushed back on that. But they were still able to meaningfully respond from the data we provided them weeks before publication. After publication, Uber was able to use—what I presume to be God View—to reverse engineer everything and to re-identify all of our volunteers. And there’s a correction on our story that notes a handful of user error issues that we now subsequently pulled out of the story.
And for the virtual testing, these were not executed transactions. So, there are no receipts to go on, right? They were able to see your behavior, your usage, your keystrokes, what you viewed, as opposed to actually purchased or executed. And they were able to reverse engineer all of that for all 174 volunteers.
So, all that’s to say, these companies – and Instacart did a similar thing when we went to them. And it sort of blew our mind. I mean, we instinctually—and when you do tech reporting and you look at patents, you know this. You know they have this capability. And when you talk to engineers, including Uber’s former chief economics guy, they all say the same thing. They have this capability. But to see this in action, I mean, the level of granular detail that they can pull out of rides and then group and insights from really small groups of customers is extraordinary. The tech is extraordinary. The capabilities are extraordinary. So, like that’s a really important place to sort of start off.
TEDDY DOWNEY: In terms of its ability to surveil.
DEREK KRAVITZ: Yes. Like it’s not really surveil. I mean, God View is probably the best way of describing that. That’s a tech term of art, but they all, they all have this. But some of them employed it in ways that—I mean, they have R&D teams made up of some of the brightest minds to come out of Ivy League institutions in the last five years, like they hoover up really, really smart people to do this type of work. And when you see it in action, it’s, at times, mind blowing.
So, I say all this because then when companies respond, they typically say some variation of, well, we don’t calculate it that way, or we don’t see it that way, or that’s illegal under consumer protection law. But then when you take that to regulators, like, for example, the fictitious discounting or pricing element, they might say, look, to a reasonable consumer, they’re not going to appreciate the difference of what Uber said to be historical comparison messaging. That’s when you see an offer screen and you see prices are lower right now, and you see an original fare crossed out, a strikethrough, and then a new fare, that little phrase, prices are lower right now, reasonable consumers might not, or definitely don’t understand that is not a discount that is historical comparison messaging in the eyes of Uber., they will treat that as a discount. They will think that the strikethrough, that is an original price, not a historical price, and that the new price is the discount price.
And regulators largely agree. And we took that to Truth in Advertising, which is sort of the gold standard, they track all of this. They said that would be misleading too. We took that to law professors. We took that to regulators. They are all sort of in agreement there.
So, there’s a gray area, right? Some state laws haven’t caught up to this type of tech. And there’s, when we looked and we dug into some of the state consumer protection laws, they don’t clearly state that this is illegal or this is against guidelines or rules. So, that’s where again, legislators, we’ve already seen in response to the Uber Lyft story, Pennsylvania has introduced legislation and passed it through one of their chambers, that would regulate this to some degree.
And then after Instacart, we saw bans now, personalized grocery store bans, pricing bans, in Maryland, Connecticut. A law passed in New York. It hasn’t been signed by the governor. A law passed in New Jersey, hasn’t been signed by the governor yet, vetoed in Colorado. But you’re seeing states already start to sort of think about this level up.
TEDDY DOWNEY: And in terms of what is actually illegal about it, I mean, the fake discount seems patently illegal under most advertising or unfair, deceptive acts and practices, laws at the state or federal level. I get it that some would say, well, hasn’t been enforced in forever, or is not enforced robustly or whatever. But that doesn’t mean it’s not a violation of the letter of the law or the spirit of the law. Will it be enforced?
You mentioned in the article, a handful of states where it could be enforced or would likely be enforced. That was my interpretation, at least, or maybe it was this is where it’s clearly counter to the law. I don’t know how you see it. But why did the experts point you to those states?
DEREK KRAVITZ: Yeah, that’s a good question. We literally shared screenshots with experts. Because we can only state so much. And a lot of this is in sort of the theoretical as opposed to the actual because you can see class actions being filed. But class actions aren’t a regulator standing up and saying, no, this is against our state consumer protection law. We’re going to file a civil suit against you.
So, we took this question to the Truth in Advertising, which tracks more than 300 different class actions across the U.S., related to pricing. But also, Veena Dubal, who’s a Professor of Law at UC Irvine, and a few others, regulators too, including in New York City, does this violate what you consider to be—does this constitute misleading, or like a UDAP violation, unfair or deceptive pricing?
And there was uniform agreement that, yes, it did. But, in only a handful of states, is it so clearly laid out or stated that a reasonable case could be brought right that there’s something on the books that clearly delineates that—California, for example, prevailing price within 90 days, there is a definition, right? And you can apply that definition to this particular case.
But algorithmic pricing is a whole ball of wax. Like, what is the prevailing price? The companies will say themselves—and do say—there is no true price for any single ride. There’s no baseline price. When we calculate it—and we shared that with experts, including one of Uber’s former chief engineers—they said, no, that there does appear to be an algorithmic baseline that is sort of the support system for all the other pricing strategies that go on top of it. But at the end of the day, a lot of these laws do not account for algorithmic pricing, and don’t account for this rapid of a change in pricing from not just hour-to-hour, minute-to-minute, second-to-second. It just wasn’t built for that, if that makes sense.
TEDDY DOWNEY: Did you talk to anyone about this being an unfair method of competition? That’s another way of, hey, you’re lying about the discount. You’re obfuscating the real price. To someone who’s being honest about the price—let’s say a cab company—you are harming the sort of method of competition. I know that’s a close cousin of sort of unfair, deceptive acts and practices. Did that come up as—you mentioned Section 5 earlier. I’m just curious if that came up as— and also for the drivers. You’re sort of obviously misleading them, to the extent they seem surprised at the price.
DEREK KRAVITZ: I think you’re hitting on something that regulators are thinking about too, especially when you have two companies that operate as a duopoly. They control 95 percent of the U.S. market. For new entrants, it’s hard for them to compete. But if you can bring up pricing strategies that may prevent both the drivers and the riders from fully understanding what the true net price is—or being quoted an upfront price that is markedly different than the final price that they see when they actually exit the vehicle—that represents an issue for some regulators.
In terms of how to combat that, a lot of regulators are now (1) pushing back against—trying to bring up—trying to push legislatures to consider algorithmic pricing bills that would try to combat most of this, if not all. But in addition, look at new competitors like TADA or Obi, places that try to level the playing field in terms of showing different prices across different apps and sort of supplying that information to both drivers and riders.
Or in the case of TADA, which is a Singaporean company, they already have 30,000 driver signups in New York City, I was told and are pilot testing in New York City. They operate off of a flat rate structure. So, you would do a $30 ride, but $28 would go to the driver and there would be a flat $2 or $3 fee to the platform. Drivers, of course, are ecstatic about that idea. And that’s why you see such a flood of drivers pre-registering for that. And they’ve been licensed by TLC in New York City, unlike Empower, which NYC says is operating illegally. So, all that’s to say that there is sort of a ground swell of sort of action on both sides there.
TEDDY DOWNEY: It sounds like the laws and regulations that we had when there were taxis, which is pretty interesting. Now, if we have any listener questions, please submit them either in the chat app or the Q&A. We’ll get to them shortly.
So, you mentioned the reaction from policymakers to both the Instacart and the Uber and Lyft reports. Can you talk a little bit more about that and what kind of response and outreach and activity you’re seeing around this type of journalism? Because to me, that is some of the most fulfilling type of reporting:when someone not only acknowledges that you did something, but wants to come up with a solution.
DEREK KRAVITZ: Yeah, and from my vantage point, the data is terribly interesting and the work we do and the result. But in terms of solutions, I’m sometimes lost. There are a lot of different policy ideas that I can quote and refer to. But in terms of dreaming up the big picture, it’s fascinating to see how different states are responding in short.
Maryland was the first state to pass an algorithmic pricing ban for grocery stores. But when we spoke to—we did a follow-up story on it. We spoke to the governor’s office. Wes Moore championed that bill, both in the state senate there and the state house. And it passed pretty quickly. I think they have a 90-day session. So, they have to move very quickly to get things going there. And the bill had a lot of gaps and loopholes in it when we spoke to experts about it.
And I actually spoke to the Chief of the Maryland Retailers Association, which is sort of an offshoot of the National Retailers Association or Federation. And she was extremely candid and sort of walked me through sort of how they got to the end product bill. She worked very closely with the governor’s office and with the bill co-sponsors to keep promotions and discounts personalized and to have all these loopholes and gaps that basically allow for them to continue to personalize pricing and do so without repercussion, without consumers being able to sue or meaningful fines or enforcement from that state’s attorney general.
Their view, of course, is that personalized promotions and discounts are important to consumers. They want to get a buy one/get one, or something off on their birthday or something for veterans or what have you. And this type of legislation could kill that off. That’s the industry line. But advocates will say that’s not the case. You can construct a bill that would allow for that type of downward personalized discount promotion and not upward or not stasis.
And it was just fascinating to see that play out. And the governor’s office, and I think others, pointed to, well, bills improve over time. Go back to the next session, we improve upon it. And two or three sessions down the road, you’re going to have something really to be proud of. But as of now, you have a bill, you have a law on the books, that everyone, I think, is in agreement is pretty toothless and sort of meaningless. So, that’s just how it plays out sometimes. And now you’re seeing other states take some of those lessons.
TEDDY DOWNEY: Yeah. Do you have any examples where the governor is not corrupt or whatever? It sounded pretty underhanded how it played out.
DEREK KRAVITZ: Yeah, I don’t think Wes Moore was corrupt or anything. But I do think that office made a calculation, made a choice, right? That we want something to be passed. This is a priority for us, but we want it to actually pass. We want to see it.
TEDDY DOWNEY: Isn’t it worse, though, to pass something that doesn’t work? Because then it didn’t work. I mean, that seems like a very cynical—it’s not a solution. They knew it wasn’t going to work. They knew it wasn’t a solution. They knew it was industry friendly. Like the whole point is you would be stopping the conduct, that is otherwise illegal, seems so very perverse.
DEREK KRAVITZ: Yeah, that was a good point that co-sponsors and others brought up is, well, this presents a line in the sand. So, we can at least point to this as being the first line in the sand. And then we will improve upon it. But I hear your point.
So, in other states, New Jersey, for example, very strong bill there that the governor had indicated—even before the bill passed their chambers, Mikie Sherrill indicated that she would sign—and basically tries to account for all of these loopholes and gaps that, because algorithmic pricing is so complex at times, because prices can shift so quickly intraday within minutes, within seconds, it sort of tries to account for that and also tries to protect promotions and discounts. But again, downward, right. Not upward.
Companies and tech companies will argue, well, there is no true price. And then regulators will say, well, there should be a true price. There should be sort of a baseline price that people sort of expect when they’re shopping for things. Because otherwise, how do you make an informed choice? Companies will push back and say comparison shop. So, there’s a lot of different arguments to this, to all of this. But you are starting to see states like, again, Connecticut, New Jersey, New York, take a stronger approach.
TEDDY DOWNEY: And your investigations have seemed to go along in line with Consumer Reports track record, I think pocketbook issue. Really where you’re looking at, hey, we have an affordability crisis. Where are these prices going up? The margins going up, the personalized pricing going up, the algorithms extracting more from the citizen where it really hurts. Some people take a lot of Ubers, a lot of Lyfts. Obviously, everyone eats.
We’ve done an investigation into algorithmic price fixing in a host of areas, including gas station pricing. Are there other areas that you think are ripe for investigation that you are going to look into or that you see, hey, these are some other areas where there is a lot of algorithmic pricing going on, going on where it really is a pocketbook issue?
DEREK KRAVITZ: We have a project list that probably is unsurprising to you in The Capitol Forum because you do similar work. But we have 50 companies on a list in areas where we would like to dig into and just not enough resources and time to do it. And it is such a rich area to focus on. And that’s both a gift, in terms of the journalism part, but also just speaks to our e-commerce and tech and how it’s sort of invaded our lives, our financial lives, day-to-day.
Just a few examples. I mean, when you think about food, Food Environment Reporting Network and The Guardian are doing—we partner with them on Kroger and they do great work. They’ve been looking at beef prices and that industry, including meatpackers, in a way that I think is thoughtful and nuanced.
But separate from that, I mean, you also see food delivery come up quite a bit as an area that you do see a lot of algorithmic pricing and personalized pricing going on. Buy Now, Pay Later is an area that I think people are starting to now appreciate as it becomes a bigger part of, again, our loan sort of system and how people buy a certain type of big ticket items, not just a Peloton from a few years ago, but increasingly like furniture and things like that. And so, those companies use a pretty sophisticated algorithms to set prices.
Affirm wants to become the new FICO score. Their CEO is very on record saying how many variables they use to help set those loan APR terms. And when you put in your information to get a new Affirm loan term, it’ll differ from person-to-person, right down to a few decimal points.
So, it’s all to say that these things are everywhere. When you look at earned wage access or payday cash advance apps, Dave, for example, or others, this is an area that the CFPB was really looking at closely before the current administration. And, I mean, they’ve moved away from a tip model and now to a model where it’s a percentage of fee for what you borrow.
But if you’re borrowing really low amounts, $25, and you’re paying $5 on that, and you have to repay it in two weeks, it’s a really high APR, under California’s definition. And companies, tech companies, don’t like those definitions, right? Sothere’s just a ton, is the short answer. There’s a ton of different things to look at. And the question is, to your larger point, what’s hurting people the most, right? How are consumers impacted the most by these pricing strategies?
TEDDY DOWNEY: Got a quick question here from the audience. What do you think of the argument that laws that allow protections on discounts will just lead to baseline pricing being inflated and then discounts being applied selectively, kind of like we see argued with PBMs leading to drug price increases so they can tell their clients they negotiated a discount?
DEREK KRAVITZ: Very good, deep question there. And the PBMs, I think, not just us, The Capitol Forum, but the New York Times has done some great work on PBMs and this issue in particular. Yeah this is a problem, right? This is something that you could see a pivot towards. I mean, you’ve already seen with personalized promotions and discounts becoming a greater share of overall pricing. What is a true baseline? And if companies are arguing that there is no such thing as a true baseline price, that it fluctuates so quickly outside of supply demand shifts, do these prevailing price calculations, do they hold any water?
I think when you speak to experts, no, is the short answer. And how to police that, a lot of states have Department of Agriculture’s weights and measures divisions that do some version of this within grocery stores, for example, and try to—even with digital price tags—try to sort of approximate what a true price should be at a given point in time and then at checkout, make sure it matches, make sure the weights match as well. I think you’re starting to see regulators take that same approach to other types of retail, consumer facing retail. But, I mean, the short answer is this is a really tough, sticky area to regulate. And I don’t think people have the full answers yet on how to do it.
TEDDY DOWNEY: I kind of disagree just in that, if you have a law that says you can’t do things that are unfair, and then everyone agrees that it’s unfair, that seems like a fairly easy law to enforce. It’s really just that they don’t actually enforce the law. It’s more a question of like prosecutorial discretion away from this.
Obviously, you’re not expecting the FTC—with Andrew Ferguson being close personal friends with David Sachs—to go after tech companies. But what is stopping a California, a Colorado, or whatever, where you mentioned that the law is fairly defined and straightforward? Someone just needs to bring a case.
To me, it’s more—I don’t know. I mean, maybe you’re right. Obviously, if you get into the intricacies of exactly how it works, but that’s exactly what Uber wanted you to do.
DEREK KRAVITZ: I know, yeah.
TEDDY DOWNEY: Like that’s exactly—they were pushing you to engage with them. What about this? What about this? What about this little thing that you can’t—you don’t have access to? What about this change in price? What about this other little tweak that we do?
And if you get in the weeds, you’re never actually going to get the crux of it, which is that the whole thing is unfair. This whole game that you’ve set up for the citizen to be have wealth extracted from them in an opaque way and have wealth extracted from the worker in an opaque way. This whole system is unfair. To me, I don’t know. I think you’re right. It’s going to be a losing battle to get into the weeds.
DEREK KRAVITZ: I mean, you bring up a great point. And it’s funny. As you were speaking, I was thinking CR has policy folks including a team that works with state lobbyists and tries to pass bills. And so, we have to disclose that in our stories and sort of separate a little bit of a firewall between the two. And you sound verbatim like a lot of our policy folks, which is a completely fair—like, so when I speak, I’m thinking of it from a data perspective, from almost like a journalistic sort of view from a neutral perspective.
The data can be noisy and messy, fascinating, but extremely difficult to parse at times. And with all these caveats and limitations. Your first order conversation, your first order framing, is 100 percent valid. And it brings up, we did it—this is for the Instacart investigation—but we did a video with More Perfect Union, which is a very interesting video journalism startup that just had six Emmy nominations recently, won one, very new, very progressive, admittedly advocacy journalism.
And so, we partnered with them. But they did an accompanying video and we interviewed Lina Khan, formerly of the FTC. And we asked her that basic question, that framing. How should we think about this? How should we be thinking about it in this disclosure way? Should there be more transparency? Should there be more of this sort of data discovery, this adversarial audits, this participatory action research?
And she’s like stepped back for a second. Have a first order conversation. Do you, do we, Royal We, want our data being used in this way? Just think about that. If we don’t want our data to be used to personalize prices, we need to state that and we need to push legislators and regulators to codify that, to make that reality. I think only now are regulators starting to think in those terms.
And on the enforcement side, a hundred percent, I think you see certain AGs, California, Colorado, New York, Oregon, they’ve hired up all these CFPB attorneys. They have these really giant consumer protection units. They are thinking about this a hundred percent and they are trying to fill the gaps that the CFPB currently has. So, yeah, no, I think that’s very valid.
TEDDY DOWNEY: Last call for questions. Otherwise, I’m going to get my last question in here. You mentioned the brightest minds are being brought into companies like this. I mean, does that seem like a good use of the brightest minds to effectively trick people out of money? I mean, I find that sad really at some fundamental level, that giving a glorified taxi roll-up company, God-like views of consumer behavior and surveilling people. That’s one question.
My second part of that is did they catch the drivers? Because like, if they found who the customers were, did they also catch and punish the drivers for sharing the receipts?
DEREK KRAVITZ: Yeah, I can answer the second part first. So, I mean, without knowing for certain if they’ve identified every single driver, they – I mean, first of all, we visually and in the story identified some of the drivers, with consent, of course, and with permission from the driver’s union there. But, in addition, I mean, yeah, the short answer is yes. They can identify everyone. And we haven’t gotten any reports that they’ve deactivated or punished any of the drivers in response. Which is, journalistically, one of my biggest concerns. Of course, it’s like a source relationship. You don’t want that.
So, but then, yeah, separately to your other point, the brightest minds thing is, like, terribly interesting. I mean, you see this with other industries too, right? I mean, unfortunately, our economy is predicated on the idea that compensation, whether it’s big law or whether it’s tech or in some cases medicine, can be tied to who’s paying and how lucrative it is. And the salaries for Uber and Lyft are strong. When you speak to formers who left these companies, they will tell you I got training there that I wouldn’t have gotten elsewhere. I saw it from the inside. Plus, I was able to run teams and do projects that were only theoretical at MIT or Penn or elsewhere. And I saw it play out in real life. I got to play with the real toys.
In terms of the moral aspect or the ethical aspect of it, I do think that’s not lost on many formers. And there are books written by formers on this, whistleblowers. I can point you to Uber Files and “Uberland,” which is a great book by a former. We quote a former in the story. So, that’s all to say, like, I think that there is that sort of realization. But money talks. Compensation is a big part of this.
TEDDY DOWNEY: And I guess I have one follow-up question there. How concerned are you about sharing data with the companies going forward to the level that you did, just given that they were able to identify the drivers?
DEREK KRAVITZ: Yeah, I think you’ve honed-in on my biggest sensitivity there. Yeah. Like it’s this huge balancing act, right? You want to be fair to companies. You want to give them a meaningful opportunity to respond. You want them to actually give you substantive answers. To have a company shut down and not engage actually is a disservice to the reader, to the audience. You want to have, in essence, a meaningful debate of ideas and also of the data itself. You want room for interpretation.
That’s why we published a GitHub data repository with the underlying data and a full methodology and all of that. But if companies can go ahead and reverse engineer everything, and I have to explain to our volunteers that this happened to them, I don’t know. It’s not an enviable position to be in. And so, it’s hard. It’s hard being a journalist for many reasons. And now that’s just an added one.
TEDDY DOWNEY: Yeah, I mean, it kind of goes to do we want to give companies the ability to punish people for being honest, right? For telling the truth, for doing journalism, or talking to journalists. It’s scary. It’s pretty scary, to be honest.
But Derek, you’ve done incredible work, as always, as being a fan of you over the years. And I can’t thank you enough for doing this. This was a really great conversation.
DEREK KRAVITZ: Well, thank you, Teddy. Really appreciate it. And thanks, everyone, for joining in.
TEDDY DOWNEY: Yeah. Thanks, everyone, for joining us today. This concludes the call. Bye-bye.
