Video: “Collusion using blockchain”

I am delighted to publish a 15-video series dedicated to my book, “Blockchain + Antitrust: The Decentralization formula”. You can access all the chapters over here, and all the video transcripts over here.



In this video, I’d like to discuss how companies could, and most certainly are, using blockchain to collude in the “real space”. To be clear, I am not saying that companies should, but I think it is important to identify that risk if we want to detect it and provide a satisfying answer.

Let me start with the easy part. Companies may use blockchain to enter into collusive agreements. For example, they could use a blockchain to share their products’ prices and ensure the information always remains accessible. Blockchain immutability could be attractive to companies, but they could also fear to give this kind of information an immutable aspect. I am curious to see how this will play out.

Now, let me address the real potential to use blockchain for colluders, or, should I say, the real risk for consumers. It implies smart contracts. As I already explained in the third video, a smart contract is a script that records a potential transaction to be triggered if and when the conditions are met.

Let me pause here a minute. You may have heard that smart contracts are not smart; and that they are not contracts. The second point is over the top. Smart contracts can be legally binding contracts. Should you be interested in this issue, I suggest you read a report I’ve written for the European Commission on the subject. The first point — the idea that smart contracts are not smart — is wrong. The word “smart” comes from the Latin “intelligere,” which means “to choose between.” Precisely because smart contracts automate the choice according to pre-defined conditions, they are “smart” in the term’s original meaning.

OK, back to our subject. Smart contracts are relevant to antitrust because companies may set up a few to automate an illegal agreement and make it more predictable and transparent between them. For instance, a smart contract could automate transfers between colluders and make side payments. It could also set a collusive price for sharing markets. You get the idea.

What is interesting for companies here is blockchain’s double effects. Let me explain. The stability of illegal collusion depends on economic and social elements. We see from empirical work that collusion succeeds when participants have more to gain from cooperation than deviation. That’s the “economic perspective” of cartel stability. And also, colluders must trust that none of them will denounce the others by applying for leniency. That’s the “social perspective” of cartel stability.

Well, blockchains can be helpful on both fronts. It provides colluders with more information than they get using another medium — I call this the “visibility effect.” At the same time, blockchains also ensure that agreements are opaque to outsiders, such as non-colluding competitors and agencies — I call this the “opacity effect.”

I say blockchains are visible for colluders because they know, coordinating in the real space, who is behind the transactions, and the nature of these transactions. On top of that, they can use smart contracts to increase visibility. Smart contracts may ensure the execution of agreements between participants. They could regulate the price that colluders charge and allow the automatic distribution of earnings according to pre-defined criteria. In a sense, smart contracts help to transform collusion into a cooperative game. Collusions are non-cooperative in nature because none of the colluders can force the others to follow their illegal agreement. If there is a contract between them, they cannot enforce it before the court, so, they cannot force the other player to reach the optimal outcome. But smart contract changes that. Of course, there are non-enforceable by courts if they are illegal in nature, but… they self-execute. In a nutshell, immutable and self-executing code makes collusions cooperative when colluders can predict how it will be triggered.

I also said blockchains are opaque to outsiders because only meta-information is made public. When antitrust agencies and competitors ignore the correspondence between blockchain identities and real-life identities, all they see is a flow between two pseudonymous users, among millions. And they do not know the reason behind transactions. They can access the smart contract, but, depending on blockchain, only the bytecode — 0 and 1s — will be made available. This means they won’t be able to detect collusion by scanning blockchains — more on that later.

For now, let me stress that, fortunately, blockchain and smart contracts won’t make collusion eternal. There are two main reasons for that. First, blockchains and smart contracts are not really flexible. This could push collusion to crack. Alternatively, this could lead colluders to include flexibility, for example, by entering information manually into the blockchain. If they do that, their cartel goes back to being non-cooperative. Second, most collusion will require using oracles. Oracles connect the blockchain space with the real space. They can be physical or intangible devices. For example, should companies collude about the prices of oranges when it rains, they will need to record if it rains into the blockchain. They can use a humidity sensor or an algorithm that scrapes information from a weather channel website, but they reinsert a single point of failure within their collusion when they do so. Indeed, if one colluder controls the oracle, it can manipulate it. So, most likely, colluders will use decentralized networks of oracles, meaning that they will combine the information coming from several of them, but implementing such a decentralized system is, of course, more complex.

With that in mind, what could antitrust agencies do? They have two options. First, they can rely on reactive methods, that is, leniency applications, complaints, etc. Unfortunately, we know, thanks to empirical work, that these methods’ efficiency is declining over the years. Blockchain will accelerate this decline, as cooperation will be more cooperative. Second option, they could be proactive. There, as I already mentioned, it will be very difficult for them and expensive to start their investigation from the blockchain space. On the contrary, it will show somewhere if the collusion is effective. This is especially true for B2C markets because final prices are generally public. So, I recommend that agencies start from real-life data, order dawn raids, go to the companies’ premises, seize computers, and enter blockchains. Doing so will require two things: (1) the expertise to conduct such market investigations and know how to operationalize computational antitrust; (2) the budget to hire the right experts and develop or acquire the right tools.

Now, to my very last point, I want to mention that although blockchains and smart contracts will make collusion more stable, I believe they will also make them explode more rapidly when the conditions are met. Colluders could indeed use smart contracts to exclude deviant colluders from their cartel, or they could use them to eject themselves when other colluders start deviating. This is what I call a natural death to collusion, as opposed to death by antitrust. Both are effective, but the first as one major inconvenient: although it’s not costly to agencies, it proves costly to all the companies who suffered from the collusion because they cannot take the agency decision and easily sue the colluders in follow-on enforcement without having to prove the practice in the first place.

That is all for today. Thank you very much for listening. Take care of yourself, and, if you can, someone else too. Cheers.

Thibault Schrepel

Related Posts