On Fairness Concerns in the Blockchain Ecosystem

TL;DR

This dissertation empirically investigates three fairness dimensions in blockchains: (i) transaction ordering fairness, (ii) contention/priority transparency, and (iii) governance voting power fairness. Across Bitcoin, Ethereum, and Compound, it finds that while miners generally follow fee-rate based norms, significant violations occur via self‑interest, private relays, and dark fees, including collusion among large mining pools. In Ethereum, private relays such as Flashbots dominate MEV capture, enabling bundling of private and public transactions with major implications for market transparency. In governance, Compound exhibits pronounced token-decision centralization, with a small set of holders and delegates shaping outcomes, raising concerns about decentralization and suggesting the need for redesigned voting mechanisms. Overall, the work demonstrates tangible fairness risks in current blockchain ecosystems and provides datasets, metrics, and analyses to guide more transparent, decentralized, and robust future designs.

Abstract

Blockchains revolutionized centralized sectors like banking and finance by promoting decentralization and transparency. In a blockchain, information is transmitted through transactions issued by participants or applications. Miners crucially select, order, and validate pending transactions for block inclusion, prioritizing those with higher incentives or fees. The order in which transactions are included can impact the blockchain final state. Moreover, applications running on top of a blockchain often rely on governance protocols to decentralize the decision-making power to make changes to their core functionality. These changes can affect how participants interact with these applications. Since one token equals one vote, participants holding multiple tokens have a higher voting power to support or reject the proposed changes. The extent to which this voting power is distributed is questionable and if highly concentrated among a few holders can lead to governance attacks. In this thesis, we audit the Bitcoin and Ethereum blockchains to investigate the norms followed by miners in determining the transaction prioritization. We also audit decentralized governance protocols such as Compound to evaluate whether the voting power is fairly distributed among the participants. Our findings have significant implications for future developments of blockchains and decentralized applications.

Figures (52)

• Figure 1: CDF of the error in predicting where a transaction would be positioned or ordered within a block according to the greedy fee-rate-based norm. Bitcoin Core code shifted completely to the fee-rate-based norm starting April 2016: Transaction ordering in Bitcoin closely tracks the fee-rate-based norm from April 2016, but differs significantly from it prior to April 2016 when a different norm was in place.
• Figure 2: The lifecycle of a Compound proposal lasts 7 days. After a proposal is created, it waits for 2 days before the 3-day voting period begins. Once the outcome of the election is decided, it takes 2 more days for the proposal to be executed and become part of the Compound Governance protocol. Proposals can also be cancelled at any time before they are executed.
• Figure 3: Illustration of a blockchain consisting of three blocks, with an extimation generation and inclusion time approximately 10 minutes. It is important to note that every block, except for the Genesis block which is the initial block in the blockchain, also uses the hash of its preceding block to compute its own hash.
• Figure 4: Volume of transactions issued and blocks mined as a function of time, showing that transactions have been issued at high rates for both (a) Bitcoin and (b) Ethereum blockchains.
• Figure 5: Distribution of blocks mined and transactions confirmed by the top-20 MPOs in data sets $\mathcal{A}$, $\mathcal{B}$, and $\mathcal{C}$. Their combined normalized hash-rates account for 94.97%, 93.52%, and 98.08% of all blocks mined in data set $\mathcal{A}$, $\mathcal{B}$, and $\mathcal{C}$, respectively.