The Governance of Decentralized Autonomous Organizations: A Study of Contributors' Influence, Networks, and Shifts in Voting Power

TL;DR

This paper empirically examines the governance of Decentralized Autonomous Organizations (DAOs) by analyzing vested contributors—owners, administrators, and developers—and their influence on voting. Using Snapshot off-chain data, Ethereum on-chain records, ENS, and The Graph, the authors construct a multi-layer dataset to quantify contributor involvement, map co-voting networks, and detect pre-vote power shifts. They find that contributors, while not universally dominant, can hold majority voting power in a non-trivial fraction of DAOs, occupy central positions in voting networks, and create inner circles through co-voting patterns; they also observe last-minute shifts in voting power preceding polls. These findings have significant implications for accountability and regulatory efforts aimed at enhancing governance transparency in DeFi ecosystems. The study provides robust methodological steps, including data validation against on-chain records and a reproducible framework, and paves the way for expanding analysis to on-chain voting and additional data sources.

Abstract

We present a study analyzing the voting behavior of contributors, or vested users, in Decentralized Autonomous Organizations (DAOs). We evaluate their involvement in decision-making processes, discovering that in at least 7.54% of all DAOs, contributors, on average, held the necessary majority to control governance decisions. Furthermore, contributors have singularly decided at least one proposal in 20.41% of DAOs. Notably, contributors tend to be centrally positioned within the DAO governance ecosystem, suggesting the presence of inner power circles. Additionally, we observed a tendency for shifts in governance token ownership shortly before governance polls take place in 1202 (14.81%) of 8116 evaluated proposals. Our findings highlight the central role of contributors across a spectrum of DAOs, including Decentralized Finance protocols. Our research also offers important empirical insights pertinent to ongoing regulatory activities aimed at increasing transparency to DAO governance frameworks.

Figures (15)

• Figure 1: Conceptualization of DAO voting. A proposal $p$ introduces potential changes to a DAO space $s$, and users $u$ can exert their decision-making power on them with their vote $v$ (green,238;blue,238, -latex, line width = 0.5mm](0,0.05) -- (0.5,0.05);) and voting power $w$ indicated by the arrow thickness. Governance users can be vested by their contribution $c$ as owner, administrator or developer to a space (). We denote their vested vote as $V^{P}_{SS}$ when they are contributors of the same-space (green,0;blue,255, -latex, line width = 0.5mm](0,0.05) -- (0.5,0.05);) they are voting on, and $V^{P}_{OS}$ when they are contributors of an other-space (green,144;blue,255, -latex, line width = 0.5mm](0,0.05) -- (0.5,0.05);).
• Figure 2: Contributor involvement across DAO spaces. The DAOs are ranked by contributor involvement $\bar{w}^{s}_C$ () from highest (left) to lowest (right). Some high-TVL dApps () are annotated for illustrative purposes and contributor involvement of more than 50.0% is colored ().
• Figure 3: Contributor self-decisions across DAO spaces. The 178.0 DAOs are ranked by contributor self-decisions $\delta^s$ () in descending order, with a threshold of 0.1%. The y-axis represents the fraction of proposals in which DAO contributors voted and decided their outcome with dominant voting power. Some high-TVL dApps () are annotated for illustrative purposes.
• Figure 4: Pagerank and k-core statistics in the four co-voting networks. Contributors tend to have higher pagerank and k-core across networks. All centrality measures make use of edge weights and are applied to the giant component; k-core statistics use geometric mean to limit the effect of outliers. Error bars are 95% confidence intervals of the means.
• Figure 5: Concentration of contributors across network communities. The bar plots show the Herfindahl-Hirschman concentration index for the distribution of contributors (green,144;blue,255,line width=1pt] (0,0) rectangle ++(0.2,0.2);) and non-contributors (green,238;blue,238,line width=1pt] (0,0) rectangle ++(0.2,0.2);) to communities assigned by the Louvain community detection algorithm. The inset donut plots show the share of communities with at least one contributor; in all networks, contributors are concentrated in a few of them.