This paper, available here , argues that important competition concerns arise from the use of consensus mechanisms in blockchains. Under such mechanisms, new information is only added to the database if the majority of network participants, the ‘nodes’, agree to do so. This requires coordination between the various network participants, which raises questions regarding whether and to what extent this voting behaviour is anticompetitive.
The paper also discusses what type of measures may be adopted to ensure that a blockchain complains with competition law by design. It is structured as follows:
Its second section provides the legal background for concerns about the functioning of consensus mechanisms.
Information may be exchanged between competitors in a blockchain that would otherwise have remained undisclosed to the participating companies or the public. In some cases, public disclosure or selective disclosure of certain information may have procompetitive effects, e.g. when information is aggregated and contributes to greater price transparency so that customers can make more informed decisions, thus reducing search costs. On the other hand, the exchange of sensitive commercial information could allow companies to anticipate other companies’ strategic decisions and to adjust their own behaviour accordingly. As a result, mere participation in an inadmissible information exchange system can lead to antitrust liability in Europe, including when the information is not only exchanged bilaterally but also centrally via a platform infrastructure in a manner akin to a hub-and-spoke. This was recently analysed in the Etruras decision by the European Court of Justice.
Another form of anti-competitive coordination may arise when the companies involved agree to use a certain technology, and thereby exclude or displace competing technologies from the market. A vote on the use or specifications of a particular technology can be legitimate if it is an expression of well-functioning competition between technologies, or if it leads to standardisation leading to significant improvements in product or distribution channels. As a rule, standardisation decisions by means of an open and transparent voting procedure do not raise competition concerns as long as any relevant party has access to this procedure.
A third section provides a description of the fundamental elements of blockchain technology.
Blockchains create distributed network structures consisting of several individual components on different computers that are able to pursue common goals without a central platform or intermediary. A blockchain is comparable to a P2P structure, in which communication or information exchange takes place directly between individual participants. Like in other distributed networks, each individual node in a blockchain is also the owner of a complete image of the database stored inside the network, which is constantly updated. The nodes are then responsible for the validation and inclusion of new information in the database. This is the quintessence of blockchain technology: new information can only added with the consent of the majority of nodes without requiring a central authority to monitor the validity of the information.
The term “blockchain” is derived from the concept of the storage process. The complex validation process, the so-called proof-of-work procedure, is not carried out for each individual piece of information. Rather, several individual pieces of data are combined in one bloc, and each block contains a summary of the information of its predecessor. The irreversibility of the database results from the fact that the proof of work must be carried out every time the information stored within a block is changed. If the information of a block changes, this affects the immediate successor block in the chain, so that the proof of work must also be resolved again for the successor.
To assess the risks posed by blockchain to competition law and policy, one must distinguish between public and private blockchains. With private blockchains, access to the network is limited from the outset by a central instance similar to a platform operator which determines the fate of the blockchain and its users. However, coordination occurs at a different technical level than in a platform. In effect, this central instance does not have to be a node of the network itself; rather, it may take any form required to set up the organisation of the Private Blockchain, including by agreement between various blockchain operators. In public blockchains, there is no centralised restriction mechanism that excludes user groups from the outset. Instead, the nodes coordinate in accordance with a previously defined algorithm, which determines which syntactic requirements information must fulfil, and which requirements must be fulfilled for the validation of new blockchain information. This algorithm sets up a consensus mechanism that makes coordination possible between the various nodes.
A fourth section identifies a number of blockchain implications for antitrust.
A first challenge concerns the irreversibility of blockchains. Companies that infringe competition law may be required by the authorities to stop engaging in such practices. This may extend to anticompetitive consensus mechanisms in a blockchain that are deemed either abusive or collusive. However, the strong technical interdependence of parties involved in a blockchain and the blockchain’s irreversibility may mean that complying with such orders may jeopardise the existence of a blockchain and associated (legal) business models and activities.
In the light of this, one may want to focus on compliance, and in ensuring that blockchains comply with competition law by design. Compliance by design may focus on the type of information which is exchanged (to ensure that information available to blockchain users is not commercially sensitive or likely to give rise to anticompetitive effects) or on how the blockchain is formed (to ensure it does not exclude competitors or alternative technologies). Furthermore, blockchain creators may assess whether the blockchain falls within existing safe harbours (which presumes the blockchain or its member companies are not dominant nor have substantial market power).
The main practical burden of compliance must be borne by the users of the respective technology. To avoid infringing competition law, blockchain users must check that the algorithm complies with competition law and, if necessary, refrain from participating in the blockchain or ask the developer to make improvements. A possible technical solution may consist of adding a second hash function to the blockchain algorithm – a so-called chameleon hash, which forms a trapdoor to the information stored inside the blocks. This makes the alteration of stored information possible and creates a redactable blockchain.
This paper provides yet another assessment of possible blockchain arrangements under EU law.
I found the discussion of the issues raised by competition enforcement in practice, and concomitant need to focus on compliance, compelling. At the same time, the piece seems to presuppose that parties will not pursue anticompetitive practices other than by accident, and therefore does not address the need for competition enforcement in those situations where competition infringements are pursued intentionally. This can be problematic from a practical perspective, as outlined in this paper, but also conceptually, as outlined in the papers reviewed over the past two weeks.
In any event, I really enjoyed the (very technical) explanation of the blockchain provided by the authors of this paper, and would recommend it to anyone interested in understanding how blockchains work and what distinguishes them from other types of distributed databases.