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 transcripts over here.
In this video, I’d like to discuss the meaning and implications of blockchain decentralization.
Let me start outside blockchain with a definition: “decentralization” measures the autonomy enjoyed by a given subject in defining its competence, put differently, in defining what it can decide for itself. To be concrete, you are part of a centralized ecosystem when you can take action only if another person has been granted the right to do so. If, on the contrary, you can decide without being formally granted the right to do so, you are part of a decentralized ecosystem. Evidently, almost no system is fully centralized or decentralized; I will return to that.
OK. It appears that the virtues of decentralization have been debated for centuries. Two virtues are generally recognized.
First, decentralized systems are democratic in nature. The absence of a strong central authority makes exclusion more difficult. Of course, the majority can be tyrannic toward the minority, but this is deemed more acceptable than the tyranny of the minority.
Second, decentralized systems are more flexible. These systems allow for more informed decision-making than centralized systems because they empower individuals to protect their own interests. Also, the impact of erroneous decisions is lower in decentralized systems than centralized ones, as in theory, they tend to affect only those who have taken the decision.
Now, one may favor decentralization over centralization as a matter of principle, but we must address the elephant in the room: decentralization toward which objective? For what? When it comes to blockchain, decentralization is solutionist: it is seen as a way to achieve more efficiency. To quote Satoshi Nakamoto, Bitcoin “requires minimal structure,” and yet, it stays “robust”. In the Ethereum white paper, it is said that only “two lines of code” can give rise to “a different set of tradeoffs that we believe will be very useful for a large class of decentralized applications”. Again, decentralization rimes with efficiency.
So, one question remains: how do blockchain ecosystems achieve and ensure the decentralization of transactions? Well, answering this question requires entering blockchain architecture and understanding how the different layers interact. Sometimes, decentralizing one layer does not impact the others. Some other times, decentralizing one layer immediately decentralizes another one — or immediately centralizes another one. The relationship is complex; this is what I call “complex decentralization”. It requires understanding casual relationships and measuring decentralization over time.
OK, let’s enter blockchain ecosystems. The first level is the “real space.” If you control Internet physical architecture, you control everything digital. The second level is the Internet. Here again, if you control Transmission Control Protocol and Internet Protocol, you control the use made of it. The third level is the “blockchain” layer or, put differently, the ledger, the database. If the blockchain is public, anyone can see it and make decisions on that basis. The system is decentralized. If the blockchain is private, only selected users can. The system is more centralized.
The fourth level is the blockchain infrastructure. I refer to this as the blockchain “constitutional level”. This level comprises the blockchain core code that runs on top of the third level. That code defines, among other things, how to use the blockchain, enter new transactions, and validate them. This fourth level also contains what is often referred to as the “layer 2” (i.e., technical solutions increasing blockchain scalability, such as Bitcoin Lightning). It defines if a blockchain is permissionless, meaning that anyone can validate transactions, or permissioned, if only selected users can. The fifth level is the blockchain governance mechanism. It results from all the constraints imposed on blockchain’s participants when they interact among themselves. We will explore that together in a forthcoming video. Let me just say for now that, contrary to popular belief, the link between this level and the sixth level (blockchain applications) is not necessarily strong. For example, the centralization of blockchain miners does not rhyme with the centralization of blockchain uses. It means that one cannot say that a blockchain is centralized because mining power (and, more generally that level) is centralized. This is too simplistic in the absence of casual relationships with other levels.
The sixth level is comprised of blockchain applications. As I have explained, three types of applications can be distinguished: cryptocurrency (blockchain 1.0), smart contracts (blockchain 2.0), and the decentralized version of any kind of program a computer can run (blockchain 3.0). The seventh level is comprised of blockchain users. Its degree of (de)centralization depends on concepts that will be familiar to antitrust-savvy readers, such as network effects, lock-in, and switching costs. Please note that ensuring the seventh level’s decentralization is not the aim of blockchain or antitrust law. Should a blockchain application prove more useful or better designed than others, all users would benefit from using it. It would create a network effect, increasing the utility they can derive from this application.
In a nutshell, decentralization is a multi-level and dynamic concept that defines one’s ability to choose its own spheres of action. Following the Darwinian perspective we have explored together in a previous video, maintaining decentralized appears central to blockchain survival. We will explore how far we can push this logic in a forthcoming video.
That is all for today. Thank you very much for listening. Take care of yourself, and, if you can, someone else too. Cheers.