An Empirical Study on Governance in Bitcoin's Consensus Evolution

TL;DR

The paper addresses how governance operates in Bitcoin's consensus evolution, challenging the view that centralised developers or pools unilaterally control rule changes. It uses an empirical grounded theory approach, analyzing 34 consensus forks in BTC and BCH to trace roles, decision processes, and power distribution both on and off the chain. It contributes a formalization of governance processes and demonstrates that deployment techniques are tightly coupled with consensus changes, shaping activation and network outcomes. The findings reveal that, despite some centralised actors, miners, full nodes, and users collectively constrain and direct evolution, thereby supporting a decentralized governance narrative. This work informs practitioners and researchers by outlining governance dynamics and highlighting the need for better transparency and quality indicators in proposal evaluation and deployment.

Abstract

Blockchain systems run consensus rules as code to agree on the state of the distributed ledger and secure the network. Changing these rules can be risky and challenging. In addition, it can often be controversial and take much effort to make all the necessary participants agree to adopt a change. Arguably, Bitcoin has seen centralisation tendencies in pools and in development. However, how these tendencies influence blockchain governance has received minimal community and academic attention. Our study analyses the governmental structures in a blockchain by looking into the history of Bitcoin. We investigate the process of changing consensus rules through a grounded theory analysis comprising quantitative and qualitative data from 34 consensus forks in Bitcoin and Bitcoin Cash. The results reveal the decentralised behaviour in Bitcoin and blockchain. Our results are in contrast to related work, emphasising centralisation among miners and developers. Furthermore, our results show how the consensus-driven deployment techniques and governance of consensus rules are intertwined.

Figures (16)

• Figure 1: Bitcoin blocks in a blockchain showing the content of the header and the body. A new block must comply with the consensus rules, e.g., only include valid transactions and find a nonce, creating a blockhash that meets the target difficulty. The figure was first presented in self.
• Figure 2: Choosing the longest chain during a collision. Miners choose to extend one of the competing blocks at height (n + 2). As soon as another block (n + 3) is found, that chain becomes the longest chain. The other block becomes orphaned. The figure was first presented in self.
• Figure 3: BIP-2: Figure from "BIP process, revised" bips. The BIP process describes how contributors can formalise proposals and the proposal lifecycle before potential deployment and activation in the Bitcoin network.
• Figure 4: In the research method implementation, oval shapes represent processes, lines indicate the flow of these processes, rectangles denote data objects, and cylinders signify archives. Different colours highlight whether the concepts relate to data sampling or grounded theory analysis.
• Figure 5: Analytical codes and categories. The relation between the categories indicates a pattern in the cycle of consensus evolution. In the interest of space, we include codes relevant to consensus evolution governance, excluding codes for technical issues, software development and human errors.