Pith. sign in

REVIEW 3 major objections 6 minor 71 references

In DAO ballots, the option the proposal author picks gets about 59 points more voting-power share than similar options; approval stance and first-list position trail it.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

T0 review · grok-4.5

2026-07-13 03:00 UTC pith:SMR2FO6U

load-bearing objection Solid large-scale Snapshot decomposition: author-selected choices dominate voting-power share, residual after author-VP removal is real but not cleanly causal. the 3 major comments →

arxiv 2607.09435 v1 pith:SMR2FO6U submitted 2026-07-10 cs.CY cs.HCecon.GNq-fin.EC

Voting Biases in Decentralized Autonomous Organization (DAO) Governance

classification cs.CY cs.HCecon.GNq-fin.EC
keywords Decentralized Autonomous OrganizationDAOGovernanceBiasBlockchainVotingCollective decision-makingDeFi
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

Decentralized Autonomous Organizations vote with tokens on Snapshot, yet support often piles onto one side. This paper asks what features of the ballot itself track that concentration. At the level of each choice on a proposal, it links voting-power share to three observables: whether the choice is approval-oriented, where it sits in the ordered list, and whether the proposal author selected it. In the main single-preference sample, author-selected choices show the largest conditional association—about 58.8 percentage points more share than otherwise similar non-author choices—followed by approval-oriented stance (about 27.1 points) and first-versus-second list position (about 7.7 points). The author-choice pattern remains large after many robustness checks, including recomputing shares after removing the author’s own voting power. The authors treat “bias” as a descriptive label for systematic association, not proven causal distortion, and argue that choice order, author signals, and vote visibility are institutional design choices rather than neutral interface details.

Core claim

On Snapshot single-preference proposals, conditional on stance, position, space fixed effects, and related covariates, an author-selected choice is associated with roughly 58.8 percentage points higher voting-power share than a comparable non-author choice; approving stance adds about 27.1 points and first-listed versus second-listed about 7.7 points. On the author-voted subsample the author-choice average marginal effect falls from about 48 to about 33 points after subtracting the author’s own voting power but does not disappear. The ordering of effects holds across a wide robustness grid.

What carries the argument

Choice-level fractional-logit models of voting-power share (with space-clustered errors and average marginal effects on the response scale), using three indicators—author selection, approval stance (from a multi-stage heuristic classifier validated against LLMs and crowdworkers), and list position—plus diagnostics that recompute shares after removing author voting power.

Load-bearing premise

The analysis only sees proposals that already reached a formal Snapshot ballot, so pre-vote screening and unmeasured proposal quality can still drive part of the author-choice association even after author votes are subtracted.

What would settle it

If, within proposals and after removing author voting power, randomizing choice order and masking which option the author selected erased the residual author-choice and first-position share advantages while leaving stance effects intact, the claimed interface and author-signal associations would not hold as design-relevant ballot features.

Watch this falsifier — get emailed when new claim-graph text bears on it.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

3 major / 6 minor

Summary. The paper studies Snapshot DAO governance at the proposal-choice level, linking each choice’s voting-power share to three observable features: list position, approval-oriented stance, and whether the proposal author selected the choice. Using cleaned data on ~127k proposals and fractional-logit models with space clustering and extensive robustness checks, it reports average marginal effects of about 58.8 pp for author-selected choices, 27.1 pp for approving stance, and 7.7 pp for first-versus-second position in the main single-preference sample. A residual author-choice association remains after subtracting the author’s own voting power (AME falls from 47.92 to 32.56 pp on the author-voted subsample). Stance labels are validated against LLMs and crowdworkers. The authors treat “bias” as descriptive association, not proven causal distortion, and argue that ordering, author signals, and vote visibility should be treated as institutional design choices.

Significance. If the reported associations hold, the paper supplies large-scale, choice-level evidence that concentrated support in DAO voting is systematically related to ballot presentation and author selection, not only to token-weighted aggregation. That is a genuine contribution relative to prior work that documents high approval rates and token concentration without decomposing within-ballot features. Strengths include the proposal-choice unit of analysis, multi-method stance validation (heuristics vs Llama/GPT/humans, κ≈0.61–0.71), favorite-choice vs support-fraction mappings, nested fractional-logit specifications, space clustering, and the author-VP removal diagnostic. The design discussion is useful for platform and DAO practitioners even if causal channels remain open. The contribution is empirical and institutional rather than theoretical, but it is well placed in cs.CY / digital-governance research.

major comments (3)
  1. [§5.2 Table 6; §6 Implications] Table 6 and §5.2–§6: The residual author-choice AME after removing author voting power (47.92→32.56 pp) is the load-bearing result for treating author selection as more than mechanical self-voting. The residual can still reflect selection of already-favored options (proposal quality, pre-vote screening, coordination, expertise) that stance, rank, and space FE do not absorb (Limitations; §2 positive-outcome skew; Table 3 overlap). Please tighten the abstract and implications so design recommendations about “author signals” are explicitly conditional on this residual not being pure quality selection, and add a clearer mechanism discussion of what the residual can and cannot identify.
  2. [§5 Table 3; Table 5] Table 3 and Table 5: First-position choices are also disproportionately approving and author-selected (e.g., author-choice mean 0.73 at rank 1). Nested fractional logits separate conditional associations, but collinearity limits sharp attribution of distinct “channels.” Report variance-inflation or joint-exclusion diagnostics, and state more explicitly which contrasts remain identified when author×position and author×stance splits (Table 6) are the preferred evidence rather than the pooled three-way decomposition.
  3. [Figure 5; Table 17; §6] Figure 5 / Table 17 TVL-25/50/100 rows: On top DeFi protocols the author AME falls sharply (≈27–33 pp) and can sit below the approve-stance AME, with wide CIs. The text notes this but the headline “author strongest” claim is sample-dependent. Either qualify the main claim as holding primarily outside the high-TVL subsample, or provide a short analysis of why high-TVL spaces differ (participation, professionalization, visibility) so the ordering is not over-generalized.
minor comments (6)
  1. [Abstract] Abstract reports “58.8% increase” while the body reports 58.8 percentage-point AMEs; use consistent percentage-point language throughout.
  2. [§4 Fig. 1] Figure 1 pools proposals with different choice counts so shares need not sum to one; the caption notes this, but a brief reminder in the main text would help readers.
  3. [§1–§2] Clarify early that “bias” is used only descriptively (as in the abstract) whenever the triad is introduced in §1–§2, to avoid causal reading of “author bias.”
  4. [Table 4; Fig. 5] Table 4 lists “Compat full” / “Expanded” naming that does not fully match the short names used in Figure 5; align labels across table and figure.
  5. [§3.1; Appendix C] Appendix stance registries are thorough; a short main-text note on multilingual coverage and the Approve+Reject joint-label rule would help non-appendix readers assess measurement risk.
  6. [§6; Appendix] Minor typos: “a a 58.8” in §6; “ressults” in Appendix D.4; check “againist/aganist” only appear as intentional keyword variants.

Circularity Check

0 steps flagged

No circular derivation: headline AMEs are observational associations estimated from Snapshot ballots, not quantities forced by definition, fitted constants, or self-citation.

full rationale

The paper’s load-bearing claims are response-scale average marginal effects from fractional-logit models of choice-level voting-power share on three observed indicators (author selection, approval stance, list rank), with space clustering and extensive sample/specification robustness (Tables 5–6, Fig. 5, Table 17). Voting-power share is constructed from token-weighted votes; author choice, stance, and position are separately coded features of the ballot—not algebraic rewrites of the outcome. Stance labels come from a documented heuristic pipeline validated against LLMs and crowdworkers (Cohen’s κ ≈ 0.61–0.71), not from the vote totals being explained. The residual author-choice AME after subtracting the author’s own voting power (Table 6: 47.92 → 32.56 pp) is an empirical sensitivity, not a tautology: the indicator remains favorite-choice author selection while the outcome is recomputed without author mass. Self-citations (e.g., prior DAO/governance work by overlapping authors) supply background and data-cleaning conventions, not uniqueness theorems or ansatzes that force the 58.8/27.1/7.7 pp ordering. The authors explicitly treat “bias” as descriptive association, not causal distortion. Selection/endogeneity concerns (pre-vote screening, unmeasured quality) affect causal interpretation, not circularity of the reported associations. No step reduces a claimed prediction or first-principles result to its own inputs by construction.

Axiom & Free-Parameter Ledger

4 free parameters · 4 axioms · 1 invented entities

The paper is empirical social science: load-bearing content is measurement and modeling choices on observational Snapshot ballots, not free physical constants or new particles. The central claim rests on treating voting-power share as the outcome, on heuristic/LLM/human stance labels, on favorite-choice expansion of multi-preference ballots, and on fractional-logit conditional associations with space clustering—not on a closed-form derivation.

free parameters (4)
  • Stance keyword registry and tier rules (exact/prefix/suffix/contains)
    Curated multi-language keyword lists and conflict rules define Approve/Reject/Abstain/Other; validated but still a hand-built measurement system that can shift the approval indicator.
  • Favorite-choice vs support-fraction ballot-mass mapping
    How multi-preference votes are expanded into choice-level mass affects author-choice definition and shares; authors report small overall mass differences but fix favorite-choice for the author indicator.
  • Maturity and cleaning thresholds (e.g., ≥5 followers, >2 voters, >1 proposal; final/flagged filters)
    Sample construction rules determine which spaces/proposals enter the analytic set (~41–43% reduction from raw).
  • Author-choice visibility window of first six interface positions
    UI-constrained covariate for whether author choice is among the first six listed options; interacts with author choice in nested models.
axioms (4)
  • domain assumption Voting-power share (and excess vs uniform-within-proposal) is the appropriate outcome for comparing choices across heterogeneous DAO weighting rules.
    Stated in §4 Measures; justified by high correlation of votes and voting power but still a modeling choice.
  • domain assumption Fractional logit with quasi-binomial link and DAO-space clustered SEs (plus space or proposal FE in robustness) separates conditional associations of author, stance, and rank despite collinearity.
    §5.1 Model Design; standard for fractional outcomes but does not identify causal channels.
  • ad hoc to paper A proposal must contain both Approve and Reject for those labels to apply; one-sided proposals are recoded Other.
    Explicit classification rule in §3 and Appendix C; affects approval-bias measurement.
  • domain assumption “Bias” denotes systematic association with voting-power share, not proven causal distortion of preferences.
    Abstract and Introduction; scopes all headline percentages.
invented entities (1)
  • Descriptive triad of position / approval / author “bias” indicators at the proposal-choice level independent evidence
    purpose: Operationalize presentation, stance, and author-signal channels for Snapshot ballots
    Not new physical entities; constructed binary/rank features from list order, stance labels, and reconstructed author selection. Independent evidence is the external literature on order effects, approval skew, and reputation, not a new object.

pith-pipeline@v1.1.0-grok45 · 39812 in / 3077 out tokens · 41175 ms · 2026-07-13T03:00:25.776935+00:00 · methodology

0 comments
read the original abstract

Decentralized Autonomous Organizations (DAOs) use token-weighted voting to allocate resources, set protocol rules, and legitimate collective decisions. Yet, support in DAO voting is strikingly concentrated. What happens inside the ballot that produces this concentration? We study DAOs' governance at the proposal-choice level, linking each choice's voting-power share to three observable features: whether it expresses an approval-oriented stance, where it appears in the choice list, and whether it is selected by the proposal author. We find that (i) author-selected choices show the strongest and most robust association with voting-power share, with a 58.8% increase relative to non-author choices; (ii) approval-oriented choices retain a positive but slightly less consistent advantage (27.1%); and (iii) first-listed choices also attract systematically higher shares, consistent with position and order effects (7.7%). Results are robust across several specifications, which include subtracting an author's own voting power from computations. We use bias descriptively, to denote systematic associations rather than proven causal distortion. The results shift attention from proposal outcomes alone to the interface and social signals through which choices are presented. In DAO governance, ordering, author signals, and vote visibility should be treated as institutional design choices, not neutral implementation details.

Figures

Figures reproduced from arXiv: 2607.09435 by Markus Strohmaier, Pietro Saggese, Stefano Balietti.

Figure 1
Figure 1. Figure 1: Share and excess of choice-level voting power for the three indicators under study. A. Outcome: mean voting-power share for choices that are first-listed, approval-oriented, or selected by the proposal author. B. Outcome: mean excess voting power relative to a uniform￾within-proposal benchmark. Error bars show 95% confidence intervals for mean estimates. Across both measures, the choice associated with eac… view at source ↗
Figure 2
Figure 2. Figure 2: A shows how the share of voting power changes across proposals with different numbers of choices. The association is strongest on proposals with two and three choices; in the latter, the first choice captures around 80% of total voting power. These proposals often follow an approve/reject pattern, as in the Basic voting system, which always has exactly three choices. Fig. 2B shows the excess voting power f… view at source ↗
Figure 3
Figure 3. Figure 3: Choice stance, list position, and excess voting power. Excess voting power is plotted separately for each stance and by choice position. Excess is relative to a uniform-within-proposal benchmark. Error bars show 95% confidence intervals for mean estimates. All stances obtain a boost in voting power when placed in position 1, but the Approve stance obtains the largest gain and has limited deficit at other p… view at source ↗
Figure 4
Figure 4. Figure 4: B compares approval stance and author selection, highlighting two descriptive patterns. First, even when disaggregated by stance, author-selected choices show positive excess voting power across positions, although the size varies. Second, the author boost is stable across all choice positions only for the Approve stance, while it is decreasing for Reject and Other stances (see also Fig. S12); interestingl… view at source ↗
Figure 5
Figure 5. Figure 5: Three-bias average marginal effects (AMEs) across main sample and robustness specifications. Points show response-scale AMEs in percentage points; horizontal intervals are Bonferroni FWER-adjusted across the 36 shared AMEs. Main corresponds to the primary speci￾fications; others are robustness checks. Across all ten robustness specifications, the author-choice AME exceeds both the position effect (first vs… view at source ↗
Figure 6
Figure 6. Figure 6: Timeline of a DAO governance process. The decision-making process can be divided into three time windows: the pre-voting period, the voting period, and the post-voting period. A Decentralized Autonomous Organizations [PITH_FULL_IMAGE:figures/full_fig_p025_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Agreement across stance classification methods. Bars count source-specific non￾missing choice-level stance labels for the heuristic, Llama 3.1, Llama 3.3, OAI, and ChatGPT sources. Labels are counted within each source-specific denominator and are limited to Approve, Reject, Abstain, and Other. posal or as a no-suggestion decision. When an automated model does detect a positive suggested stance, the stance… view at source ↗
Figure 8
Figure 8. Figure 8: Suggested-stance labels by source. Bars show the share of proposals for which the source found no suggested voting direction or found a textual cue suggesting Approve, Reject, Abstain, or Other. Details about human labels are available in Sec. C.3, but anticipate here the relevant result [PITH_FULL_IMAGE:figures/full_fig_p037_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: Choice stance counts by number of choices in a proposal (full sample) [PITH_FULL_IMAGE:figures/full_fig_p049_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: Voting power and position by stance (full sample). A. Mean choice-level voting￾power share by stance. B. Share of choices with a given stance that are in position 1, by number of choices in the proposal. These shares are conditional on stance and therefore do not describe the overall distribution of stances. Confidence intervals are 95% confidence intervals of the means [PITH_FULL_IMAGE:figures/full_fig_… view at source ↗
Figure 11
Figure 11. Figure 11: Excess voting power by stance, choice position, and number of choices (full sample). Only stance-position-num-choice combinations with more than 10 observations are kept in the plot. Error bars are 95% confidence intervals of the means [PITH_FULL_IMAGE:figures/full_fig_p051_11.png] view at source ↗
Figure 12
Figure 12. Figure 12: Author boost by stance and choice position. Each dot in the plot is the difference between the point estimate of the average excess voting power of a choice of an author vs other choices. Error bars are 95% confidence intervals of the means [PITH_FULL_IMAGE:figures/full_fig_p052_12.png] view at source ↗

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

71 extracted references · 7 canonical work pages · 1 internal anchor

  1. [1]

    Taming the unpredictability of cultural markets with social influence

    Andrés Abeliuk, Gerardo Berbeglia, Pascal Van Hentenryck, Tad Hogg, and Kristina Lerman. Taming the unpredictability of cultural markets with social influence. In Proceedings of the 26th international conference on world wide web , pages 745–754, 2017

  2. [2]

    Uniswap v3 core

    Hayden Adams, Noah Zinsmeister, Moody Salem, River Keefer, and Dan Robinson. Uniswap v3 core. Technical report, Uniswap, 2021. A vailable at https://uniswap.org/whitepaper-v3.pdf

  3. [3]

    Addressing trust bias for unbiased learning-to-rank

    Aman Agarwal, Xuanhui Wang, Cheng Li, Michael Bendersky, and Marc Najork. Addressing trust bias for unbiased learning-to-rank. In The World Wide Web Conference , pages 4–14, 2019

  4. [4]

    The technology of decentralized finance (DeFi)

    Raphael Auer, Bernhard Haslhofer, Stefan Kitzler, Pietro Saggese, and Friedhelm Victor. The technology of decentralized finance (DeFi). Digital Finance , August 2023. doi:10.1007/s42521-023-00088-8

  5. [5]

    nodeGame: Real-time, synchronous, online experiments in the browser

    Stefano Balietti. nodeGame: Real-time, synchronous, online experiments in the browser. Behavior Re- search Methods, 49(5):1696–1715, 2017

  6. [6]

    From crypto to nfts: Identifying the new wave of digital investors

    Stefano Balietti, Can Celebi, and David Tercero-Lucas. From crypto to nfts: Identifying the new wave of digital investors. International Review of Financial Analysis , 104:104172, 2025

  7. [7]

    Slaying the dragon: The quest for democracy in decentralized autonomous organizations (daos)

    Stefano Balietti, Pietro Saggese, Stefan Kitzler, and Bernhard Haslhofer. Slaying the dragon: The quest for democracy in decentralized autonomous organizations (daos). arXiv preprint arXiv:2511.09263 , 2025

  8. [8]

    Banerjee

    Abhijit V. Banerjee. A simple model of herd behavior. The Quarterly Journal of Economics , 107(3):797– 817, 1992. doi:10.2307/2118364

  9. [9]

    De- centralised finance’s timocratic governance: The distribution and exercise of tokenised voting rights

    Tom Barbereau, Reilly Smethurst, Orestis Papageorgiou, Johannes Sedlmeir, and Gilbert Fridgen. De- centralised finance’s timocratic governance: The distribution and exercise of tokenised voting rights. Technology in Society , page 102251, April 2023. URL: https://www.sciencedirect.com/science/ar ticle/pii/S0160791X23000568, doi:10.1016/j.techsoc.2023.102251

  10. [10]

    Defi, not so decentralized: The measured distribution of voting rights

    Tom Josua Barbereau, Reilly Smethurst, Orestis Papageorgiou, Alexander Rieger, and Gilbert Fridgen. Defi, not so decentralized: The measured distribution of voting rights. In Proceedings of the Hawaii International Conference on System Sciences 2022 , page 10, 2022

  11. [11]

    A theory of fads, fashion, custom, and cultural change as informational cascades

    Sushil Bikhchandani, David Hirshleifer, and Ivo Welch. A theory of fads, fashion, custom, and cultural change as informational cascades. Journal of Political Economy , 100(5):992–1026, 1992. doi:10.1086/ 261849

  12. [12]

    Evaluating position bias in large language model recommenda- tions

    Ethan Bito, Yongli Ren, and Estrid He. Evaluating position bias in large language model recommenda- tions. arXiv preprint arXiv:2508.02020 , 2025

  13. [13]

    Buddensiek and Paul P

    Eric F. Buddensiek and Paul P. Momtaz. Information cascades in decentralized autonomous organizations (daos), 2025. SSRN working paper. doi:10.2139/ssrn.5258156

  14. [14]

    The myopia of crowds: Cognitive load and collective evaluation of answers on stack exchange

    Keith Burghardt, Emanuel F Alsina, Michelle Girvan, William Rand, and Kristina Lerman. The myopia of crowds: Cognitive load and collective evaluation of answers on stack exchange. PloS one , 12(3):e0173610, 2017

  15. [15]

    Origins of algorith- mic instabilities in crowdsourced ranking

    Keith Burghardt, Tad Hogg, Raissa D’Souza, Kristina Lerman, and Marton Posfai. Origins of algorith- mic instabilities in crowdsourced ranking. Proceedings of the ACM on Human-Computer Interaction , 4(CSCW2):1–20, 2020

  16. [16]

    Quantifying the impact of cognitive biases in question- answering systems

    Keith Burghardt, Tad Hogg, and Kristina Lerman. Quantifying the impact of cognitive biases in question- answering systems. In Proceedings of the International AAAI Conference on Web and Social Media , volume 12, 2018. 22 S. Balietti, P. Saggese, M. Strohmaier

  17. [17]

    Moving beyond coin voting governance, 2021

    Vitalik Buterin. Moving beyond coin voting governance, 2021. A vailable at https://vitalik.ca/gener al/2021/08/16/voting3.html

  18. [18]

    Sequential voting promotes collective discovery in social recommendation systems

    L Elisa Celis, Peter Krafft, and Nathan Kobe. Sequential voting promotes collective discovery in social recommendation systems. In Proceedings of the International AAAI Conference on Web and Social Media , volume 10, pages 42–51, 2016

  19. [19]

    A study of position bias in digital library recommender systems

    Andrew Collins, Dominika Tkaczyk, Akiko Aizawa, and Joeran Beel. A study of position bias in digital library recommender systems. arXiv preprint arXiv:1802.06565 , 2018

  20. [20]

    Improvement proposal: Compound v2 to v3 migration phase 1, 2023

    Compound. Improvement proposal: Compound v2 to v3 migration phase 1, 2023. A vailable at: https: //compound.finance/governance/proposals/152

  21. [21]

    Centralized governance in decentral- ized organizations

    Lin William Cong, Daniel Rabetti, Charles CY Wang, and Yu Yan. Centralized governance in decentral- ized organizations. Available at SSRN 5168660 , 2025

  22. [22]

    Documentation

    ConstitutionDAO. Documentation. Technical report, ConstitutionDAO, 2021. A vailable at https: //www.constitutiondao.com/

  23. [23]

    An experimental comparison of click position-bias models

    Nick Craswell, Onno Zoeter, Michael Taylor, and Bill Ramsey. An experimental comparison of click position-bias models. In Proceedings of the 2008 international conference on web search and data mining , pages 87–94, 2008

  24. [24]

    Improvement proposal: reduction to crv token inflation, 2020

    Curve. Improvement proposal: reduction to crv token inflation, 2020. A vailable at: https://gov.curv e.fi/t/discussion-reduction-to-crv-inflation/851

  25. [25]

    Quantifying voter biases in online platforms: An instrumental variable approach

    Himel Dev, Karrie Karahalios, and Hari Sundaram. Quantifying voter biases in online platforms: An instrumental variable approach. Proceedings of the ACM on Human-Computer Interaction , 3(CSCW):1– 27, 2019

  26. [26]

    The Vulnerable Nature of Decentralized Governance in DeFi

    Maya Dotan, A viv Yaish, Hsin-Chu Yin, Eytan Tsytkin, and A viv Zohar. The Vulnerable Nature of Decen- tralized Governance in DeFi. Technical report, The Hebrew University, August 2023. arXiv:2308.04267 [cs] type: article. URL: http://arxiv.org/abs/2308.04267, doi:10.48550/arXiv.2308.04267

  27. [27]

    A comparative analysis of the platforms for decentralized autonomous organizations in the ethereum blockchain

    Youssef Faqir-Rhazoui, Javier Arroyo, and Samer Hassan. A comparative analysis of the platforms for decentralized autonomous organizations in the ethereum blockchain. Journal of Internet Services and Applications, 12(1):9, 2021. doi:10.1186/s13174-021-00139-6

  28. [28]

    The hidden shortcomings of (d) aos–an empirical study of on-chain governance

    Rainer Feichtinger, Robin Fritsch, Yann Vonlanthen, and Roger Wattenhofer. The hidden shortcomings of (d) aos–an empirical study of on-chain governance. arXiv preprint arXiv:2302.12125 , 2023

  29. [29]

    Analyzing voting power in decentralized gover- nance: Who controls daos?, 2022

    Robin Fritsch, Marino Müller, and Roger Wattenhofer. Analyzing voting power in decentralized gover- nance: Who controls daos?, 2022. arXiv:2204.01176

  30. [30]

    Metaverse Governance: An Empirical Analysis of Voting Within Decentralized Autonomous Organizations

    Mitchell Goldberg and Fabian Schär. Metaverse Governance: An Empirical Analysis of Voting Within Decentralized Autonomous Organizations. Technical Report 4310845, Rochester, NY, December 2022. URL: https://papers.ssrn.com/abstract=4310845, doi:10.2139/ssrn.4310845

  31. [31]

    Social desirability bias

    Pamela Grimm. Social desirability bias. Wiley international encyclopedia of marketing , 2010

  32. [32]

    Decentralized autonomous organization

    Samer Hassan and Primavera De Filippi. Decentralized autonomous organization. Internet Policy Review, 10(2), 2021. doi:10.14763/2021.2.1556

  33. [33]

    Disentangling the effects of social signals

    Tad Hogg and Kristina Lerman. Disentangling the effects of social signals. arXiv preprint arXiv:1410.6744, 2014

  34. [34]

    How decentralized is the governance of blockchain-based finance: Empirical evidence from four governance token distributions, 2021

    Johannes Rude Jensen, Victor von Wachter, and Omri Ross. How decentralized is the governance of blockchain-based finance: Empirical evidence from four governance token distributions, 2021. arXiv: 2102.10096. Voting Biases in Decentralized Autonomous Organization (DAO) Governance 23

  35. [35]

    Accurately interpreting clickthrough data as implicit feedback

    Thorsten Joachims, Laura Granka, Bing Pan, Helene Hembrooke, and Geri Gay. Accurately interpreting clickthrough data as implicit feedback. In 28th Annual International ACM SIGIR Conference on Research and Development in Information Retrieval, SIGIR 2005 , pages 154–161, 2005

  36. [36]

    The governance of decentralized autonomous organizations: A study of contributors’ influence, networks, and shifts in voting power

    Stefan Kitzler, Stefano Balietti, Pietro Saggese, Bernhard Haslhofer, Markus Strohmaier, et al. The governance of decentralized autonomous organizations: A study of contributors’ influence, networks, and shifts in voting power. In Financial Cryptography and Data Security 2024 . IFCA, 2024

  37. [37]

    Jonathan G. S. Koppell and Jennifer A. Steen. The effects of ballot position on election outcomes. The Journal of Politics , 66(1):267–281, 2004. doi:10.1046/j.1468-2508.2004.00151.x

  38. [38]

    Krosnick and Duane F

    Jon A. Krosnick and Duane F. Alwin. An evaluation of a cognitive theory of response-order effects in survey measurement. Public Opinion Quarterly , 51(2):201–219, 1987

  39. [39]

    Compound: The money market protocol

    Robert Leshner and Geoffrey Hayes. Compound: The money market protocol. Technical report, Com- pound, 2019. A vailable at https://bit.ly/3ioWOjW

  40. [40]

    Yuyang Liang. Knowledge sharing in online discussion threads: What predicts the ratings? In Proceedings of the 2017 ACM Conference on Computer Supported Cooperative Work and Social Computing , pages 146–154, 2017

  41. [41]

    Counterfactual voting adjustment for quality assessment and fairer voting in online platforms with helpfulness evaluation

    Chang Liu, Yixin Wang, and Moontae Lee. Counterfactual voting adjustment for quality assessment and fairer voting in online platforms with helpfulness evaluation. arXiv preprint arXiv:2506.21362 , 2025

  42. [42]

    Lorenz, H

    J. Lorenz, H. Rauhut, F. Schweitzer, and D. Helbing. How social influence can undermine the wisdom of crowd effect. Proceedings of the National Academy of Sciences (PNAS) , 108(22):9020–9025, 2011

  43. [43]

    The maker protocol: Makerdao’s multi-collateral dai (mcd) system

    MakerDAO. The maker protocol: Makerdao’s multi-collateral dai (mcd) system. Technical report, Mak- erDAO, 2020. A vailable at https://makerdao.com/en/whitepaper

  44. [44]

    Influence via ethos: On the persuasive power of reputation in deliberation online

    Emaad Manzoor, George H Chen, Dokyun Lee, and Michael D Smith. Influence via ethos: On the persuasive power of reputation in deliberation online. Management Science , 70(3):1613–1634, 2024

  45. [45]

    Voting, fast and slow: ballot order and likeability effects in the five-star movement 2012 online primary election

    Francesco Marolla, Angelica Maineri, Jacopo Tagliabue, and Giovanni Cassani. Voting, fast and slow: ballot order and likeability effects in the five-star movement 2012 online primary election. Contemporary Italian Politics , 16(3):266–283, 2024

  46. [46]

    On the causes and consequences of ballot order effects

    Marc Meredith and Yuval Salant. On the causes and consequences of ballot order effects. Political Behavior, 35(1):175–197, 2013. doi:10.1007/s11109-011-9189-2

  47. [47]

    Gummadi, and Patrick Loiseau

    Johnnatan Messias, Vabuk Pahari, Balakrishnan Chandrasekaran, Krishna P. Gummadi, and Patrick Loiseau. Understanding blockchain governance: Analyzing decentralized voting to amend defi smart contracts, 2024. arXiv:2305.17655

  48. [48]

    Social influence bias: A randomized experiment

    Lev Muchnik, Sinan Aral, and Sean J Taylor. Social influence bias: A randomized experiment. Science, 341(6146):647–651, 2013

  49. [49]

    Decentralized Finance, Centralized Ownership? An Iterative Mapping Process to Measure Protocol Token Distribution

    Matthias Nadler and Fabian Schär. Decentralized Finance, Centralized Ownership? An Iterative Mapping Process to Measure Protocol Token Distribution. arXiv:2012.09306 [cs, econ, q-fin] , December 2020. arXiv: 2012.09306. URL: http://arxiv.org/abs/2012.09306

  50. [50]

    Identifying topical authorities in microblogs

    Aditya Pal and Scott Counts. Identifying topical authorities in microblogs. In Proceedings of the fourth ACM international conference on Web search and data mining , pages 45–54, 2011

  51. [51]

    Who is authoritative? understanding reputation mecha- nisms in quora

    Sharoda A Paul, Lichan Hong, and Ed H Chi. Who is authoritative? understanding reputation mecha- nisms in quora. arXiv preprint arXiv:1204.3724 , 2012

  52. [52]

    Reputation systems

    Paul Resnick, Ko Kuwabara, Richard Zeckhauser, and Eric Friedman. Reputation systems. Communi- cations of the ACM , 43(12):45–48, 2000. doi:10.1145/355112.355122. 24 S. Balietti, P. Saggese, M. Strohmaier

  53. [53]

    Blockholders and strategic voting in daos’ governance

    Romain Rossello. Blockholders and strategic voting in daos’ governance. Available at SSRN 4706759 , 2024

  54. [54]

    Voters’ preferences and the will- ingness to pay for voting rights: New evidence from decentralized finance

    Romain Rossello, Stefano Balietti, Stefan Kitzler, and Pietro Saggese. Voters’ preferences and the will- ingness to pay for voting rights: New evidence from decentralized finance. Available at SSRN 6313439 , 2026

  55. [55]

    Salganik, P.S

    M.J. Salganik, P.S. Dodds, and D.J. Watts. Experimental study of inequality and unpredictability in an artificial cultural market. Science, 311(5762):854–856, 2006

  56. [56]

    Documentation

    Snapshot. Documentation. Technical report, Snapshot, 2023. A vailable at https://docs.snapshot.or g/

  57. [57]

    Popularity dynamics and intrinsic quality in reddit and hacker news

    Greg Stoddard. Popularity dynamics and intrinsic quality in reddit and hacker news. In Proceedings of the International AAAI Conference on Web and Social Media , volume 9, pages 416–425, 2015

  58. [58]

    Decentralization illusion in decentral- ized finance: Evidence from tokenized voting in makerdao polls, 2023

    Xiaotong Sun, Charalampos Stasinakis, and Georigios Sermpinis. Decentralization illusion in decentral- ized finance: Evidence from tokenized voting in makerdao polls, 2023. arXiv:2203.16612

  59. [59]

    Documentation

    Sushiswap. Documentation. Tech. rep., Sushiswap, 2022. A vailable at https://docs.sushi.com/

  60. [60]

    Predicting the perceived quality of online mathematics contributions from users’ reputations

    Yla R Tausczik and James W Pennebaker. Predicting the perceived quality of online mathematics contributions from users’ reputations. In Proceedings of the SIGCHI conference on human factors in computing systems , pages 1885–1888, 2011

  61. [61]

    Improvement proposal: Making protocol fees operational, 202

    Uniswap. Improvement proposal: Making protocol fees operational, 202. A vailable at: https://gov.un iswap.org/t/making-protocol-fees-operational/21198

  62. [62]

    Improvement proposal: deploy uniswap v3 on bnb chain, 2022

    Uniswap. Improvement proposal: deploy uniswap v3 on bnb chain, 2022. A vailable at: https://gov.un iswap.org/t/rfc-update-deploy-uniswap-v3-1-0-3-0-05-0-01-on-bnb-chain-binance/19734

  63. [63]

    Improvement proposal: Uniswap dao can help 80k uni to turkiye to relief after the big earthquake disaster, 2023

    Uniswap. Improvement proposal: Uniswap dao can help 80k uni to turkiye to relief after the big earthquake disaster, 2023. A vailable at: https://gov.uniswap.org/t/governance-proposal-uniswap-dao-can -help-80k-uni-to-turkiye-to-relief-after-the-big-earthquake-disaster/20758

  64. [64]

    Large language models are not fair evaluators

    Peiyi Wang, Lei Li, Liang Chen, Zefan Cai, Dawei Zhu, Binghuai Lin, Yunbo Cao, Lingpeng Kong, Qi Liu, Tianyu Liu, et al. Large language models are not fair evaluators. In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) , pages 9440–9450, 2024

  65. [65]

    An empirical study on snapshot daos

    Qin Wang, Guangsheng Yu, Yilin Sai, Caijun Sun, Lam Duc Nguyen, Sherry Xu, and Shiping Chen. An empirical study on snapshot daos. ArXiv, abs/2211.15993, 2022. URL: https://api.semanticschola r.org/CorpusID:254069759

  66. [66]

    Decentralized autonomous organizations: Concept, model, and applications

    Shuai Wang, Wenwen Ding, Juanjuan Li, Yong Yuan, Liwei Ouyang, and Fei-Yue Wang. Decentralized autonomous organizations: Concept, model, and applications. IEEE Transactions on Computational Social Systems , 6(5):870–878, 2019

  67. [67]

    The decentralization of science and the emergence of decentralized science

    Lukas Weidener and Bence Lukács. The decentralization of science and the emergence of decentralized science. Available at SSRN 5094597 , 2025

  68. [68]

    Strategic vote timing in online elections with public tallies

    A viv Yaish, Svetlana Abramova, and Rainer Böhme. Strategic vote timing in online elections with public tallies. arXiv preprint arXiv:2402.09776 , 2024

  69. [69]

    Anchoring bias in online voting

    Zimo Yang, Zi-Ke Zhang, and Tao Zhou. Anchoring bias in online voting. Europhysics Letters , 100(6):68002, 2012. Voting Biases in Decentralized Autonomous Organization (DAO) Governance 25 Create a Proposal Cast Votes Execute the Proposal -2 0 2 4 6 8 10

  70. [70]

    Determine Voting Power

    Voting Period2. Determine Voting Power

  71. [71]

    Lorem ipsum dolor sit amet, consectetur adipiscing elit [...]

    Execute Proposal Fig. 6: Timeline of a DAO governance process . The decision-making process can be divided into three time windows: the pre-voting period, the voting period, and the post-voting period. A Decentralized Autonomous Organizations Figure 6 summarizes the governance sequence used throughout the empirical analysis. A proposal is created, voting ...