Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Why Ethereum Needs Fairness Mechanisms that Do Not Depend on Participant Altruism

Patrick Spiesberger, Nils Henrik Beyer, Hannes Hartenstein

TL;DR

It is suggested that relying solely on the mere presence of altruistic proposers is insufficient to ensure that proposed fairness mechanisms reestablish Ethereum's ideals, highlighting the need for additional incentive- or penalty-based mechanisms.

Abstract

Ethereum's ideals of decentralization and censorship resistance are undermined in practice, motivating ongoing efforts to reestablish these properties. Existing proposals for fairness mechanisms depend on the assumption that a sufficient fraction of block proposers adhere to Ethereum's protocols as intended. We refer to such proposers as altruistic, as this behavior may come at the cost of reduced revenue. Prior analyses indicate that a consistent share of 91 percent of proposers delegate block construction to centralized services, effectively signing externally constructed blocks blindly, and are thus not considered altruistic. To assess whether the remaining 9 percent of proposers genuinely exhibit altruistic behavior, we conducted an empirical analysis and found that an additional 6.1 percent also interact with such external services. Further, we found that less than 1.4 percent of proposers consistently acted in accordance with Ethereum's decentralization and censorship resistance objectives. These findings suggest that relying solely on the mere presence of altruistic proposers is insufficient to ensure that proposed fairness mechanisms reestablish Ethereum's ideals, highlighting the need for additional incentive- or penalty-based mechanisms.

Why Ethereum Needs Fairness Mechanisms that Do Not Depend on Participant Altruism

TL;DR

It is suggested that relying solely on the mere presence of altruistic proposers is insufficient to ensure that proposed fairness mechanisms reestablish Ethereum's ideals, highlighting the need for additional incentive- or penalty-based mechanisms.

Abstract

Ethereum's ideals of decentralization and censorship resistance are undermined in practice, motivating ongoing efforts to reestablish these properties. Existing proposals for fairness mechanisms depend on the assumption that a sufficient fraction of block proposers adhere to Ethereum's protocols as intended. We refer to such proposers as altruistic, as this behavior may come at the cost of reduced revenue. Prior analyses indicate that a consistent share of 91 percent of proposers delegate block construction to centralized services, effectively signing externally constructed blocks blindly, and are thus not considered altruistic. To assess whether the remaining 9 percent of proposers genuinely exhibit altruistic behavior, we conducted an empirical analysis and found that an additional 6.1 percent also interact with such external services. Further, we found that less than 1.4 percent of proposers consistently acted in accordance with Ethereum's decentralization and censorship resistance objectives. These findings suggest that relying solely on the mere presence of altruistic proposers is insufficient to ensure that proposed fairness mechanisms reestablish Ethereum's ideals, highlighting the need for additional incentive- or penalty-based mechanisms.
Paper Structure (26 sections, 4 figures)

This paper contains 26 sections, 4 figures.

Table of Contents

  1. Introduction
  2. Defining Non-Altruistic Behavior
  3. Extracting Maximum Extractable Value
  4. Sharing Account Governance
  5. Methodology
  6. Analyzing Relay Announcements
  7. Method I:
  8. Analyzing Shared Account Governance
  9. Method II:
  10. Method III:
  11. Method IV:
  12. Examining Block Construction
  13. Method V:
  14. Method VI:
  15. Results
  16. ...and 11 more sections

Figures (4)

  • Figure 1: Methodology for identifying altruistic proposers: starting from all observed proposers, excluding evident MEV-Boost participants (rationale in \ref{['sec:on_alt_mev']}), checking shared account governance (cf. \ref{['sec:on_alt_gov']}), and detecting block-level MEV-related block construction (cf. \ref{['sec:on_alt_mev']}). Sections 3.1 to 3.3 refer to methodology, Sections 4.1 to 4.3 to results.
  • Figure 2: Analysis results -- percentages denote shares of the observed proposer set (600 115 proposers in total).
  • Figure 3: Histogram of proposers sharing an EOA coinbase address, referred to as clusters. Grey dots denote clusters in which all proposers' blocks were announced by an MEV-Boost relay. Accordingly, these clusters are excluded by classifier I. Blue dots indicate clusters in which no proposer published a block announced by any MEV-Boost relay. Red dots indicate clusters (not necessarily all of the same size) that include a proper subset of proposers using MEV-Boost. These proper subsets are excluded under classifier II. The cumulative distribution function shows the relative share of blocks produced by clusters, highlighting that two clusters (nearly identical in size, attributable to the two largest builders heimbach_ethereums_2023) account for more than 80 % of all blocks.
  • Figure 4: Histogram of proposers classified as potentially altruistic (blue, down), together with the number of proposers that proposed at least one block during our measurement period (orange, top). Each bin corresponds to a consecutive range of 37 919 validator indices. The dashed line separates proposers registered before and after Ethereum's "The Merge" ethereum2024merge. The two series are shown on different scales.