Algorithmic Software-Facilitated Price Fixing with Jonathan Rubin
Jul 02, 2024
Tom Hagy
Everyone knows that price fixing is against the law, chiefly Section 1 of the Sherman Act.
Competitors may not collude, i.e., agree, to keep prices where they want them, but there are relatively new pricing platforms that some companies maintain take them out of the equation, so they do not have to share private information directly with competitors. Instead, they claim, they feed their data to a third-party which uses algorithms to come up with pricing for these competitors based on data they all contribute. The subject has been getting a lot of attention as cases mount against a company called RealPage, a firm that provides shared pricing services for landlords. The company faces dozens of suits in multidistrict litigation and has also captured the attention of federal antitrust law enforcers. But they are not the only company finding themselves in litigation.
As our guest recently wrote: “When pricing algorithms are used by individual firms, such as airlines, e-commerce platforms, ride-share and room-share companies, stock traders, and others, there are unlikely to be anti-competitive consequences. It is when market competitors avail themselves of the same algorithmic program or service that the specter of unlawful collusion arises.” That risk increases as markets become more concentrated, he says.
He is Jonathan Rubin, Partner and Co-Founder of MoginRubin LLP, a widely recognized competition law attorney, economist, and commentator who has presented at antitrust conferences in the United States and Europe, testified before several congressional committees, and before the Directorate General for Competition of the European Commission.
“The fact that these services employ an algorithm is not central to what's going on in this scenario,” he told me, “because what's important is the conduct of the businesspeople involved.”
Listen to my interview with Jonathan Rubin as we discusswhat algorithmic or software-facilitated pricing is, what the law says about price collusion, how this new pricing mechanism violates that law, and recent developments in litigation.
Everyone knows that price fixing is against the law, chiefly Section 1 of the Sherman Act.
Competitors may not collude, i.e., agree, to keep prices where they want them, but there are relatively new pricing platforms that some companies maintain take them out of the equation, so they do not have to share private information directly with competitors. Instead, they claim, they feed their data to a third-party which uses algorithms to come up with pricing for these competitors based on data they all contribute. The subject has been getting a lot of attention as cases mount against a company called RealPage, a firm that provides shared pricing services for landlords. The company faces dozens of suits in multidistrict litigation and has also captured the attention of federal antitrust law enforcers. But they are not the only company finding themselves in litigation.
As our guest recently wrote: “When pricing algorithms are used by individual firms, such as airlines, e-commerce platforms, ride-share and room-share companies, stock traders, and others, there are unlikely to be anti-competitive consequences. It is when market competitors avail themselves of the same algorithmic program or service that the specter of unlawful collusion arises.” That risk increases as markets become more concentrated, he says.
He is Jonathan Rubin, Partner and Co-Founder of MoginRubin LLP, a widely recognized competition law attorney, economist, and commentator who has presented at antitrust conferences in the United States and Europe, testified before several congressional committees, and before the Directorate General for Competition of the European Commission.
“The fact that these services employ an algorithm is not central to what's going on in this scenario,” he told me, “because what's important is the conduct of the businesspeople involved.”
Listen to my interview with Jonathan Rubin as we discusswhat algorithmic or software-facilitated pricing is, what the law says about price collusion, how this new pricing mechanism violates that law, and recent developments in litigation.
Tom Hagy Litigation Enthusiast and Host of the Emerging Litigation Podcast Home Page Follow us on LinkedIn Subscribe on your favorite platform.
Tom Hagy:
Jonathan Rubin, thank you very much for talking to me today.
Jonathan Rubin:
Thanks for inviting me. It's a pleasure to be here, Tom.
Tom Hagy:
So we're talking about today is algorithmic pricing, so-called, but I thought first of all we would get the definition right. What is it and what else might we call it?
Jonathan Rubin:
Well, algorithmic pricing is a phrase that has sort of grown up around the recent interest in determining whether or not the common use of software by competitors might be a violation of the antitrust laws. This is a fairly new technology that involves the application of sophisticated computer software to create price recommendations and companies run the software and sell it as a service to competitors who receive those recommendations and rely on them. So really the fact that it employs an algorithm is not central to what's going on in this scenario, because what's important is the conduct of the business people involved. So you might look at it as software-aided price fixing, because in certain circumstances price fixing is really the conduct that's being engaged in and the focus should be on that rather than the mechanism to accomplish it.
Tom Hagy:
Well, what type of mechanisms were used previously in history?
Jonathan Rubin:
A complaint was recently filed in this area that pointed out that price fixing as a human activity has been intermediated by, you know, the Pony Express and the US mails and the telephone, then by emails and text messages and now by the common use of software programs.
Tom Hagy:
What types of industries or companies are most often using this software-aided pricing platform?
Jonathan Rubin:
In recent years and by that I mean the last two, maybe three years it's come to the attention of antitrust enforcers and antitrust plaintiffs' attorneys that some industries are susceptible to a type of coordination that involves the use of a third party's software system, for example hotel rooms. The rate setting on hotel rooms in a competitive market is performed by the individual hotels, but there are now software services available to the operators of hotels that collect information that is otherwise not available in the public domain and then create with that information pricing recommendations to their clients and in so doing, promising them to increase their overall revenues, and in fact this software is generally referred to as revenue management software, or RMS or revenue management solutions. Hotels are particularly susceptible. So are rental housing units. In fact, the principal case that's now pending involves a company called RealPage and their software revenue management software that's used by the owners of multifamily rental units.
Tom Hagy:
Where else are you seeing it being used besides with hotels and landlords and the reimbursement of insurance payments to healthcare providers.
Jonathan Rubin:
There are third parties that offer assistance to insurance companies to help them determine the reimbursement that is owed to an out-of-network healthcare provider. So that's a healthcare provider that's not in the insurance company's network, that they don't have a continuing relationship with and they have to make reimbursements across the country. So they often use the services of a third-party software provider which has an algorithm for making recommendations for those reimbursement levels. So those are the main examples of the industries now which have come under scrutiny from the use of a third-party software vendor.
Tom Hagy:
Well for non-antitrust experts. What does the law generally say about price collusion?
Jonathan Rubin:
Well, I think the place to start is Section 1 of the Sherman Act, which was passed in 1890. So this is a long-standing federal law that we've had in our country. Section one says every contract, combination in the form of trust or otherwise, or conspiracy and restraint of trade or commerce among the several states or with foreign nations is hereby declared to be illegal. So the elements of this statute are that there has to be an agreement, which can be a contract or a combination or a conspiracy. It is a common intention to execute a common plan. It's a commitment to a common course of action. It's also called concerted action, but for shorthand in the law we call that the agreement element. And the second element is that the agreement must unreasonably restrain trade and interstate commerce.
Tom Hagy:
The Sherman Act was enacted in 1890. So how has it evolved in the courts since then?
Jonathan Rubin:
So over the years the courts have dealt with what this means and what the parameters of this are. The unreasonable requirement, for example, came out of a case that interpreted the statute to say that clearly the statute doesn't mean that all contracts are restraints of trade that are illegal, but only unreasonable restraints of trade are unreasonable.
Tom Hagy:
Are there other specific precedents that you think our listeners should know about in this context?
Jonathan Rubin:
I think the next place to go is to look at how a case in 1939 applied Section 1 of the Sherman Act called the Interstate Circuit case, and that involved a couple of theater chains in Texas and New Mexico Together. They were jointly managed, or at least very close, and they controlled about 75% of the theaters in that area of the country. And they wrote a letter to eight movie studios, movie distributors. The same letter addressed everybody. All the addressees were on one letter. All the addressees were on one letter and they said we want you to consider that if you want to continue to do business with us, the licenses that you allow for the subsequent runs of the first run movies that you are licensing to us, those must be exhibited at a minimum price of 25 cents or at a time when the price was usually less than 25 cents, and they must not be permitted to be shown in a double feature. What they actually said was that they invited the distributors to agree to limit subsequent runs of first-run movies to a minimum admission price which was above current prices and to prohibit second-run theaters from showing double features for those first-run movies that they were exhibiting.
Jonathan Rubin:
The Supreme Court had to review a decision of the district court it was a direct appeal in those days it was an equity because there was an injunction issued evaluated the evidence supporting the district court's conclusion that, even though each of the distributors had responded on its own affirmatively to the restrictions that the theater chains were demanding, the Supreme Court says in its opinion the government is without the aid of direct testimony that the distributors entered into any agreement with each other to impose the restrictions upon subsequent run exhibitors and so in order to establish the agreement element of Section 1 in that case, the government was compelled to rely on inferences drawn from the course of conduct of the alleged conspirators. And one should note that this is not an unusual situation. Plaintiffs or the government enforcers who are plaintiffs do not usually have the benefit of direct evidence of an agreement. Folks do not engage in illegal conspiracies out in the open.
Tom Hagy:
Then where did the court come down on this?
Jonathan Rubin:
The Supreme Court observed that the circumstantial evidence was such that, well they say, from the beginning of the whole incident, each of the distributors knew that the same proposals were under consideration by the others and each was aware that they were all in active competition and that without substantially unanimous action by all of them with respect to the restrictions for any given territory, there was a risk of a substantial loss of business and goodwill of the subsequent run theaters and independent exhibitors, but that if they were all together on it, there was the prospect of increased profits.
Jonathan Rubin:
So there was a common awareness of the risk of saying no and the benefits of saying yes, and so therefore there was a very strong motive for an agreement, concerted action, and of course the theater chains took advantage of this very strong motive in presenting their demands.
Jonathan Rubin:
But what's interesting is that what the case very much turned on was the state of mind of the defendants in the case.
Jonathan Rubin:
They pointed out that they knew the competitors, that each of the competitors received the same proposal, they were aware of the risk of no agreement and they were aware of the prospect of increased profit with agreement, and there was therefore a very strong motive to engage in joint action.
Jonathan Rubin:
All of this is pretty much about the intentions and the state of mind, although there was also empirical, circumstantial evidence that the court used, in particular that by complying with the proposals, the distributors would be departing from a very customary previous business practice. There was also observed a drastic increase in the admission prices in most subsequent-run theaters, as you would expect. So therefore, the holding was that the evidence supported the trial court's inference that there was an agreement and note that this is even in the absence of communication, direct communication on the subject shown between the distributors. There was nonetheless an agreement, and in holding that what the court said, and I'll quote here it taxes credulity to believe that the several distributors would, in the circumstances, have accepted and put into operation with substantial unanimity such far-reaching changes in their business methods without some understanding that they were all going to join, and we reject as beyond the range of probability that it was a result of mere chance.
Tom Hagy:
So that decision, the interstate circuit case, was in 1939. Bring us forward to its present-day importance and how the DOJ relied on it in filing a statement of interest in the RealPage rental software antitrust litigation case which is pending in the Middle District in Tennessee.
Jonathan Rubin:
First, the DOJ said that under Section 1, it's personally unlawful for competitors to join together their independent decision-making power to raise, depress, fix, peg or stabilize prices, and they also say that the machinery employed by a combination for price fixing is immaterial. The question in this case, the DOJ stated, is whether the defendants have violated Section 1 of the Sherman Act by allegedly knowingly combining their sensitive, non-public pricing and supply information in an algorithm that they rely upon in making pricing decisions, with the knowledge and expectation that other competitors will do the same and, importantly, in establishing concerted action. The DOJ claimed in their memorandum, the law does not require a showing of simultaneous action or even action that is close in time. Quote from Interstate Circuit. It is elementary that an unlawful conspiracy may be, and often is, formed without simultaneous action or agreement on the part of the conspirators.
Tom Hagy:
From there, the DOJ analyzed the facts alleged in the case against RealPage, a service used by landlords to price multifamily housing units. So, based on that, what did the DOJ argue in its statement of interest?
Jonathan Rubin:
The DOJ said that the factual allegations in those complaints against the landlords and against RealPage point to evidence of an invitation to act in concert, followed by acceptance, evidence that is sufficient to plead concerted action under interstate circuit. So under interstate circuit, it suffices to show that RealPage proposed the price fixing scheme to competing landlords who were each aware that its competitors were also being invited to participate in the scheme and the competitors adhered to it, generating a common understanding among the competitors that they would increase prices collectively by using RealPage. So this statement by the DOJ is almost pulled directly from the reasoning and holding of interstate circuit. And what's really interesting is that interstate circuit 1939, and this case is 2023, and one involves software and the other involves a letter. It doesn't matter. It has to do with the intentions of the parties and what the true nature of their behavior is what do you think?
Tom Hagy:
Has the DOJ made its case and is interstate circuit going to be enough?
Jonathan Rubin:
And I'm going to say no, but there is more.
Tom Hagy:
How are the companies defending themselves? I mean, it sounds like they're saying well, there's no specific, explicit agreement to collude, but who does that? So how are they defending themselves?
Jonathan Rubin:
a piece recently where he said no legal statute or precedent that people who use the same software to inform their decision-making represents anything even remotely similar to collusion. Another point is that there's no intent to agree, there's no making of an agreement, because there is no direct interaction between the users. And then finally, the evidence about the intentions and all of that stuff. That is speculation and we shouldn't be speculating about what's you know are the intentions and what is the what are the motivations of the defendants. I think all of those are a little bit desperate and a little bit misguided, because there's a little more to the story and it ends up being that there's a number of considerations or kinds of circumstantial evidence, all of which, when they're present, makes a pretty overwhelming case.
Tom Hagy:
But when I asked you if the DOJ had made its case and whether Interstate Circuit was going to be enough, you said no. Why is that?
Jonathan Rubin:
Because of some of these criticisms in Interstate Circuit. That was an agreement that expressly required the distributors to, you know, require minimum pricing by second-run theaters and restrict, you know, output, meaning no double features where, as in RealPage, there's a third party who purports to be offering a service to competing clients and they're not jointly solicited the way they were in interstate circuit. They go to the client separately and say sign up for our services.
Tom Hagy:
So what else is needed besides the offer and acceptance?
Jonathan Rubin:
setup of interstate circuit, the centralization of decision-making, as expressed in the Supreme Court case from 2010 of American Needle. The Supreme Court there recognized how important independent decision making is in the market to the competitive process. And in fact, when you think about it, the economics of section one is a prohibition against competitors engaging in what economists call joint profit maximization. And that's important, because economic decisions that you make in the best interests of a group, so all the competitors together, for example, are not the same decisions that are made by independent actors making decisions in their own self-interest, and that is the problem, that is the harm, that's the evil that Section 1 seeks to prevent. So, rather than you know, have an evidentiary argument about, is there sufficient evidence of an agreement, as between the group, it's really a question of whether or not the group is jointly profit maximizing or they're individually profit maximizing, and for that, the American needle point of view regarding centralization of decision making is very important.
Jonathan Rubin:
In citing American Needle, the DOJ said in its statement of interest in RealPage that section one applies to collaborations that eliminate independent decision-making. However they've been brought about, realpage has themselves stated that the ability to quote, outsource daily pricing and ongoing revenue oversight close quote to RealPage Allows RealPage to set prices for its clients' property quote as though we, realpage, own them ourselves. Close quote. So this is an advertisement for precisely what American Needle says should not happen, which is the centralization of decision-making.
Tom Hagy:
Then what did the district court rule?
Jonathan Rubin:
The district court in RealPage issued an order where it upheld the multifamily housing complaint and they observed that the RealPage revenue management solution clients provide RealPage with independent, commercially sensitive pricing and supply data. To allow RealPage to use this data to set prices for not only their own properties but also the properties of their horizontal competitors who use RMS, the clients must be willing to quote, outsource their daily pricing and ongoing revenue oversight by accepting price recommendations 80 to 90% of the time. The relevant question, as it is whenever you're considering circumstantial evidence of conspiracy, is whether or not we're talking about parallel conduct of the sort that's the product of independent action, or is there a plausible inference of unlawful agreement, that is, are the competitors jointly profit maximizing?
Tom Hagy:
So give us the litigation landscape about how many cases are out there and you've got one as well.
Jonathan Rubin:
There are I've counted depending on how broadly you cast the net, but between 20 and 30 pending pieces of litigation now involving software-aided price fixing. There are cases involving a nationwide class of the multifamily housing units. That's the real page we've been talking about. There are specialty software cases involving a different software Yardi Systems cases. That's also for multifamily housing renters for a different set of landlords but a significant proportion of landlords in any local market. There's specialty software that is built for casino hotels which have a different demand profile than ordinary hotels, and there's cases against the Sendin Group and their software Rainmaker.
Tom Hagy:
But what was different about the Gibson case? Why did the Gibson court not allow it to proceed, while RealPage is allowed to proceed?
Jonathan Rubin:
proceed while real page is allowed to proceed. The Gibson Court made the observation that a hub and spoke theory of the Sherman Act liability, where you know, in fact you've got horizontal competitors connected by a wheel rim and in the middle is a third party with spokes going out to each of them, that is fulfilling some function having to do with coordination or creating the opportunity or enforcing the cartel. Whatever the case may be, we call that a hub-and-spoke theory of sermon act liability and the gibson court said that if you're going to use that theory based on the use of algorithmic pricing, it would depend in part on the exchange of non-public information between competitors through the algorithm. The court found that it was quote unclear whether the pricing recommendations generated to hotel operators included competitors' confidential information fed in. Perhaps they only get their own confidential information back mixed with public information from other sources.
Tom Hagy:
Then how does that decision inform future cases?
Jonathan Rubin:
This highlights some of the additional types of elements of evidence that we might look for in order to have greater security that we have identified unlawful conduct, and that is the question of whether non-public, proprietary, competitively sensitive information of the kind that you would not and should not share with your competitors is being sent to the software, and whether the software is sending back information that in some degree employs or relies on or depends on competitively sensitive information that your rival has sent in. So this takes the exchange of information, which is evidence of possible collusion. It's not of itself illegal, but it's very damning evidence of a conspiracy in the context of other evidence. This is not merely an exchange of information, no-transcript in some way. If the recommendation that you're getting back from the third party would not have been able to exist but for the input by your rivals of competitively sensitive information, then I think that's a very important indicia that the software is aiding price fixing.
Tom Hagy:
Okay, you've got a case of your own. What can you tell us about that one?
Jonathan Rubin:
We at my firm, at Mogan Rubin are counsel for plaintiffs that have sued SAS, who have a program called the IDEAS program, which is used by hotels. We have identified 17 US metropolitan statistical areas where there are a significant number of hotels that are using the SAS IDEAS software and essentially we believe that the various indicia that we've been talking about today of when there is an agreement through the use of a common software platform are present in that case. The case has recently been filed, so there's nothing to report as yet.
Tom Hagy:
So how does your case stack up or line up with all of various requirements and precedents?
Jonathan Rubin:
we've discussed already that case fulfills the various criteria that we've discussed.
Jonathan Rubin:
There's a change in conduct.
Jonathan Rubin:
There is proprietary information going into the software and information coming back which would not have been feasible or possible without the rival's proprietary information.
Jonathan Rubin:
There's a very important element about conduct that's against your self-interest. Generally, raising prices without some understanding from your co-conspirators is against your self-interest because you'll lose business and certainly providing confidential, non-public, competitively sensitive information to a third party is against your own interests unless you know that your rivals will also provide the same information. There's various evidence and allegations that the participants in certainly in the SAS in our case, and in RealPage, knew very well that not only were they providing their non-public information, but so were their rivals, and it's not a huge leap to believe that they would not have provided that information if they were, if they did not know that their rivals were also doing the same. So there are many of the boxes that there's some, a combination of the sort of classic antitrust plus factors that go into supporting the inference of an agreement, together with some more modern ones, and that creates a cluster of circumstantial evidence which makes it very difficult to imagine what else, beside price fixing, the purpose of the whole enterprise is.
Tom Hagy:
So take this back to how interstate circuit applies in this case. What's kind of?
Jonathan Rubin:
fun about interstate circuit is that the Supreme Court made an observation that there was no upper level executive presented a trial to explain why all of the distributors decided to do the same thing. There was no legitimate justification presented. All they had were sort of local managers who did not have visibility into what was going on on a sort of industry level, come in and say, oh, we decided on our own to do this and, as the judge said, it just created strange credulity. But they used that absence of evidence to create an adverse inference. Where is the upper level executive giving us the real non-price fixing reason that they're doing this? I think you get a similar dynamic in the real page and the SAS and the you know what we're now calling the software aided price fixing.
Tom Hagy:
I guess some would argue that one person's coincidence is another person's conspiracy.
Jonathan Rubin:
You're getting close to that third point that the commentator made, that evidence bearing on incentives is just speculation about what their motives are, what their intent is. Well, that's not true, you know. There's a certain sense in which, come on, we all know what the distributors were up so, so that that would be in the group's best interest rather than in the individual's self-interest. So, therefore, making a joint profit.
Tom Hagy:
Shifting gears a little bit. Going on to legislation, senator Amy Klobuchar of Minnesota has been banging the drum on updating, modernizing antitrust law. She recently proposed the Algorithmic Pricing Collusion Act. Right now it only has Democratic support, so who knows about its prospects? But regardless, what do you think of it?
Jonathan Rubin:
Well, I'm not a legislative expert or a lobbyist, but my sense is that I think it's good to get a bill out there, if for nothing else but for discussion purposes the bill does have. It has a number of definitions. For example, it defines algorithmic pricing, it defines various things and it tries to capture the nature of certain behavior, and I think that it probably should sort of sit there for a while until there are some more judicial decisions that hone in on precisely what the essence of the conduct is. That's sufficient in order to infer price fixing. You know I'm a pro-enforcement person.
Jonathan Rubin:
I think the antitrust laws are very important and should be enforced, but in order to retain their legitimacy, it's very important that they avoid false positives. We need to be very rigorous about what the requirements are before we can find a violation of Section 1 and whether or not there's been sufficient experience with this type of price fixing to allow us to draw up legislation that sort of puts these things in stone and sets things in certain definitions. It might be a little bit early for that. I appreciate the effort, but I think it might be just a bit premature. I think we ought to let the antitrust courts get an opportunity to analyze the various flavors of this behavior before we attempt to make a statute.
Tom Hagy:
We've covered a lot of ground. Do you have any final thoughts?
Jonathan Rubin:
Well, I just think it's a very interesting area and is a testament to the ingenuity of business folks who have an understandable and constant motivation, particularly in highly competitive environments, to engage in strategies that try to avoid or minimize competition, and their revenues are enhanced by that kind of behavior, but consumer surplus is lessened.
Jonathan Rubin:
The whole point of the antitrust enterprise is to try to make sure that the rules are clear and to deter, you know what is a you know, a kind of a natural desire for business people to want to sort of cut corners and maybe blunt the rigors of competition a little bit.
Jonathan Rubin:
And so this is another step on that journey and I think it's important to do. And so far the decisions made in the judiciary have been pretty good. Even in the Nevada case in Las Vegas, if in fact, the pleading was not sufficiently specific about how RILO's information was used in order to come up with output that was provided to the clients, then maybe the court was right about that. But that's an easily correctable pleading thing if those aren't in fact facts, if that's the problem, I think that's an important legal benchmark, shall I say, for whether or not we're talking about illegality and that's that sort of. Are you getting back information that wouldn't have been possible to get without the inclusion, in whole or in part, of your rival's confidential information? If so, I think then you're talking about unlawful conduct.
Tom Hagy:
Well, jonathan Rubin, thank you very much for speaking with me about this important and interesting issue. Thank you for having me, tom, it's been a pleasure.
