Democracy on Rugged Landscapes: Phase Transitions in Optimal Voting Rules

Joshua Nunley

arxiv: 2606.02813 · v1 · pith:DAHJBH4Enew · submitted 2026-06-01 · 💻 cs.GT · cs.MA· cs.SI· physics.soc-ph

Democracy on Rugged Landscapes: Phase Transitions in Optimal Voting Rules

Joshua Nunley This is my paper

Pith reviewed 2026-06-28 11:44 UTC · model grok-4.3

classification 💻 cs.GT cs.MAcs.SIphysics.soc-ph

keywords voting methodsNK landscapesphase transitionsdirect democracyrepresentative democracycollective decision makingfitness optimization

0 comments

The pith

As governance landscapes grow more rugged the best voting method switches from cardinal scoring to Borda count to STAR voting.

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

The paper treats collective law-making as an NK optimization process in which shared law bits are chosen by vote while individual citizen traits stay fixed and a single parameter alpha sets how strongly laws affect those traits. Eight common voting rules plus a tunable scoring family are run on thousands of landscapes that vary in ruggedness K and in alpha. The simulations show sharp switches in which rule produces the highest average fitness: cardinal scoring on smooth landscapes, a low-p ordinal scorer at modest ruggedness, Borda count over a broad central band, and STAR voting at the highest K. Borda also records the best mean fitness together with the smallest variance across most of the tested space. When the model is extended to representative democracy the pattern changes again, with cardinal scoring dominant and plurality rising only at high identity weighting and low candidate self-interest.

Core claim

Under direct democracy on NK landscapes the optimal voting method undergoes sharp phase transitions as a function of landscape complexity: cardinal score voting dominates on smooth landscapes, ordinal scoring with p=0.35 at low-to-moderate ruggedness, Borda count across a wide middle range, and STAR voting at the highest complexity. Borda count achieves the highest mean fitness and lowest variance across most of the parameter space. Representation reshapes the structure so that cardinal score voting dominates across most regimes while plurality emerges at high identity weight and low-to-moderate candidate self-interest.

What carries the argument

NK fitness landscapes in which shared law bits are updated by voting while individual bits remain fixed, with cross-dependency strength controlled by the scalar alpha.

If this is right

Borda count supplies the highest average outcome quality and the smallest outcome variance over most values of ruggedness and dependency strength.
No single voting rule is optimal; the winner depends on the measured complexity of law-citizen interactions.
Representative institutions with strong identity weighting and low candidate self-interest favor plurality over the other tested rules.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

If policy domains can be ranked by their effective K, election designers could select the matching rule rather than defaulting to one method for all decisions.
The two-parameter reduction of the (K, alpha) plane to a single axis offers a practical way to classify real legislative problems by complexity before choosing a voting rule.

Load-bearing premise

That real collective governance can be faithfully captured by NK optimization on shared-law bits controlled by a single cross-dependency scalar.

What would settle it

Running the same voting rules on landscapes whose ruggedness is measured directly from real policy-outcome data and checking whether the predicted switches in best rule appear.

Figures

Figures reproduced from arXiv: 2606.02813 by Joshua Nunley.

**Figure 2.** Figure 2: Mean fitness rank (top) and variance rank (bottom) [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗

**Figure 3.** Figure 3: Mean rank vs. complexity for all nine methods across the [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗

read the original abstract

Laws and institutions shape individual outcomes through complex interactions with citizens' diverse circumstances, yet how different voting methods navigate this coupled landscape remains poorly understood. We model collective governance as optimization on NK fitness landscapes, where shared bits (laws) are updated by voting while individual bits (personal traits) remain fixed. A cross-dependency parameter $\alpha$ controls how legislation's effects depend on individual circumstances. We compare eight standard voting methods and a generalized scoring family across landscape ruggedness $K \in \{1,\ldots,20\}$ and $\alpha \in [0,1]$ with 1000 runs per configuration. Under direct democracy, the optimal voting method undergoes sharp phase transitions as a function of landscape complexity: cardinal score voting dominates on smooth landscapes, ordinal scoring with $p=0.35$ at low-to-moderate ruggedness, Borda count across a wide middle range, and STAR voting at the highest complexity. A two-parameter empirical formula reduces the $(K, \alpha)$ plane to a single complexity axis for visualization. Borda count achieves the highest mean fitness and lowest variance across most of the parameter space. We further introduce a representative democracy model parameterized by identity weight $\beta$ and candidate self-interest $p_{\mathrm{self}}$. Representation reshapes the complexity-dependent structure even under favorable conditions: cardinal score voting dominates across most regimes, with plurality emerging as the top method at high $\beta$ and low-to-moderate $p_{\mathrm{self}}$.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

The paper maps voting rules onto NK landscapes and reports phase transitions in the best-performing rule as ruggedness K and cross-dependency alpha change.

read the letter

The main takeaway is that cardinal scoring wins on smooth landscapes, a tuned ordinal scorer at low-to-moderate K, Borda across a broad middle band, and STAR at the highest complexity, with Borda also showing the best average fitness and lowest variance over most of the (K, alpha) grid. A two-parameter fit collapses the plane for visualization. The representative-democracy extension shifts the picture toward score voting in most regimes.

What is actually new is the direct NK mapping for shared-law bits plus the empirical identification of those ordering changes and the collapse formula. The work does a clean job of running 1000 replicates per cell across eight standard methods plus the parameterized scorer, which gives a systematic picture rather than isolated examples.

The soft spots are the missing details on how phase boundaries were determined—no mention of statistical tests, confidence intervals, or sensitivity to random seeds and landscape generation. The two-parameter formula looks post-hoc, and it is not obvious whether it generalizes or simply summarizes the observed grid. The modeling premise that governance reduces to NK optimization on shared bits with a scalar alpha is a strong simplification; the paper treats it as given rather than testing alternatives.

This is for readers already comfortable with NK models or computational approaches to social choice who want to see how ruggedness might reorder institutional design choices. It is not yet a finished theory but a set of simulation patterns worth checking.

I would send it to peer review so referees can examine the code, the exact transition statistics, and whether the representative model adds much beyond the direct case.

Referee Report

3 major / 1 minor

Summary. The paper models collective governance as optimization on NK fitness landscapes, with shared law bits updated by voting and individual bits fixed, controlled by a cross-dependency parameter α. Simulations of eight voting methods plus a parameterized ordinal scoring family across K ∈ {1,…,20} and α ∈ [0,1] (1000 runs per cell) report sharp phase transitions in the optimal method under direct democracy: cardinal score voting on smooth landscapes, ordinal scoring (p=0.35) at low-to-moderate ruggedness, Borda count over a wide middle range, and STAR voting at highest complexity. Borda achieves highest mean fitness and lowest variance across most of the space. A two-parameter empirical formula collapses the (K,α) plane to one axis. Representative democracy (parameters β, p_self) shifts the structure, with cardinal score dominating and plurality emerging at high β and low-to-moderate p_self.

Significance. If the NK model with scalar α captures essential features of law-citizen interactions, the work supplies a quantitative, simulation-driven taxonomy of voting-rule performance across complexity regimes. The scale of 1000 runs per configuration and the explicit ordering of methods provide a reproducible empirical baseline for comparing voting systems on rugged landscapes; this could inform mechanism-design questions in computational social choice.

major comments (3)

[Abstract] Abstract, model description paragraph: the central claim of sharp phase transitions rests on forward simulation; the manuscript supplies no statistical tests (e.g., bootstrap or permutation tests on dominance boundaries) to establish that observed switches are distinguishable from sampling noise or gradual transitions.
[Abstract] Abstract, results paragraph: the two-parameter empirical formula is used to reduce the (K,α) plane to a single complexity axis for visualization, yet the text gives no indication whether the parameters were fitted to the same simulation output they summarize or derived independently; post-hoc fitting would render the reported dominance regions partly circular.
[Abstract] Abstract, experimental protocol sentence: the protocol states 1000 runs per (K,α) pair but omits random-seed handling, exact NK landscape generation procedure (e.g., how the dependency matrix is sampled), and any pre-registration or cross-validation steps; these omissions are load-bearing for assessing whether the reported ordering of methods (Borda, STAR, etc.) is robust.

minor comments (1)

[Abstract] Abstract: the notation p=0.35 for the ordinal scoring exponent and the interval α ∈ [0,1] are introduced without a forward reference to the precise functional form used in the generalized scoring family.

Simulated Author's Rebuttal

3 responses · 0 unresolved

We thank the referee for the careful review and the recommendation of minor revision. We respond point by point to the three major comments, indicating revisions where appropriate.

read point-by-point responses

Referee: [Abstract] Abstract, model description paragraph: the central claim of sharp phase transitions rests on forward simulation; the manuscript supplies no statistical tests (e.g., bootstrap or permutation tests on dominance boundaries) to establish that observed switches are distinguishable from sampling noise or gradual transitions.

Authors: We agree that formal statistical tests would strengthen the claim of sharp transitions. With 1000 runs per cell the observed dominance separations are large, but we will add bootstrap resampling to compute confidence intervals on mean fitness differences between methods and thereby quantify the transition boundaries more rigorously. revision: yes
Referee: [Abstract] Abstract, results paragraph: the two-parameter empirical formula is used to reduce the (K,α) plane to a single complexity axis for visualization, yet the text gives no indication whether the parameters were fitted to the same simulation output they summarize or derived independently; post-hoc fitting would render the reported dominance regions partly circular.

Authors: The two-parameter formula was obtained by fitting the simulation output to collapse the plane for visualization. We will revise the text to state this explicitly and to emphasize that the dominance regions themselves are determined directly from the raw per-cell simulation results, not from the fitted formula, so the reported ordering remains non-circular. revision: yes
Referee: [Abstract] Abstract, experimental protocol sentence: the protocol states 1000 runs per (K,α) pair but omits random-seed handling, exact NK landscape generation procedure (e.g., how the dependency matrix is sampled), and any pre-registration or cross-validation steps; these omissions are load-bearing for assessing whether the reported ordering of methods (Borda, STAR, etc.) is robust.

Authors: We will expand the methods section (and the abstract if space permits) to specify the random-seed protocol, the standard Kauffman NK generation procedure including uniform sampling of the dependency matrix, and the absence of pre-registration or cross-validation (neither applies to an exhaustive simulation sweep). The complete simulation code and data will be released to permit independent verification of the method ordering. revision: yes

Circularity Check

0 steps flagged

No significant circularity; results are direct simulation outputs

full rationale

The paper reports phase transitions and dominance regions among voting methods as empirical findings obtained by forward simulation on NK landscapes whose fitness functions are defined independently of the voting rules being tested. The abstract states that results come from 1000 runs per (K, α) configuration with no equations shown that would make any reported optimum a function of the same simulation outputs. The two-parameter empirical formula is used only for visualization of the (K, α) plane and is not load-bearing for the ordering of methods. No self-citations or uniqueness theorems are invoked to justify the central claims, so the derivation chain remains self-contained against external benchmarks.

Axiom & Free-Parameter Ledger

2 free parameters · 1 axioms · 0 invented entities

Ledger constructed from abstract only; full paper may list additional parameters or background results.

free parameters (2)

ordinal scoring exponent p = 0.35
Value 0.35 reported as optimal in low-to-moderate ruggedness regime; appears chosen to maximize fitness in the simulations.
two parameters of empirical complexity formula
Formula that reduces the (K, α) plane to a single axis; parameters must be fitted to the simulation output.

axioms (1)

domain assumption NK landscapes with tunable K and α faithfully represent the interaction structure of laws and individual circumstances
Core modeling choice stated in the abstract's description of the governance model.

pith-pipeline@v0.9.1-grok · 5800 in / 1302 out tokens · 28470 ms · 2026-06-28T11:44:09.702176+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

13 extracted references

[1]

Achen, C. H. and Bartels, L. M. (2016).Democracy for Real- ists: Why Elections Do Not Produce Responsive Government. Princeton University Press

2016
[2]

Arrow, K. J. (1951).Social Choice and Individual Values. Wiley

1951
[3]

D., editors (2016).Handbook of Computational Social Choice

Brandt, F., Conitzer, V ., Endriss, U., Lang, J., and Procaccia, A. D., editors (2016).Handbook of Computational Social Choice. Cambridge University Press

2016
[4]

Gibbard, A. (1973). Manipulation of voting schemes: A general result.Econometrica, 41(4):587–601

1973
[5]

Kauffman, S. A. (1993).The Origins of Order: Self-Organization and Selection in Evolution. Oxford University Press

1993
[6]

Levinthal, D. A. (1997). Adaptation on rugged landscapes.Man- agement Science, 43(7):934–950

1997
[7]

Mansbridge, J. (1999). Should blacks represent blacks and women represent women? a contingent “yes”.The Journal of Poli- tics, 61(3):628–657

1999
[8]

Rivkin, J. W. (2000). Imitation of complex strategies.Management Science, 46(6):824–844

2000
[9]

Ruhl, J. B. (1996). Fitness of law: Using complexity theory to de- scribe the evolution of law and society and its practical mean- ing for democracy.Vanderbilt Law Review, 49:1407–1490

1996
[10]

Ruhl, J. B. (2008). Law’s complexity: A primer.Georgia State University Law Review, 24:885–911

2008
[11]

Saari, D. G. (1995).Basic Geometry of Voting. Springer

1995
[12]

Satterthwaite, M. A. (1975). Strategy-proofness and arrow’s con- ditions: Existence and correspondence theorems for voting procedures and social welfare functions.Journal of Economic Theory, 10(2):187–217

1975
[13]

Young, H. P. (1974). An axiomatization of borda’s rule.Journal of Economic Theory, 9(1):43–52

1974

[1] [1]

Achen, C. H. and Bartels, L. M. (2016).Democracy for Real- ists: Why Elections Do Not Produce Responsive Government. Princeton University Press

2016

[2] [2]

Arrow, K. J. (1951).Social Choice and Individual Values. Wiley

1951

[3] [3]

D., editors (2016).Handbook of Computational Social Choice

Brandt, F., Conitzer, V ., Endriss, U., Lang, J., and Procaccia, A. D., editors (2016).Handbook of Computational Social Choice. Cambridge University Press

2016

[4] [4]

Gibbard, A. (1973). Manipulation of voting schemes: A general result.Econometrica, 41(4):587–601

1973

[5] [5]

Kauffman, S. A. (1993).The Origins of Order: Self-Organization and Selection in Evolution. Oxford University Press

1993

[6] [6]

Levinthal, D. A. (1997). Adaptation on rugged landscapes.Man- agement Science, 43(7):934–950

1997

[7] [7]

Mansbridge, J. (1999). Should blacks represent blacks and women represent women? a contingent “yes”.The Journal of Poli- tics, 61(3):628–657

1999

[8] [8]

Rivkin, J. W. (2000). Imitation of complex strategies.Management Science, 46(6):824–844

2000

[9] [9]

Ruhl, J. B. (1996). Fitness of law: Using complexity theory to de- scribe the evolution of law and society and its practical mean- ing for democracy.Vanderbilt Law Review, 49:1407–1490

1996

[10] [10]

Ruhl, J. B. (2008). Law’s complexity: A primer.Georgia State University Law Review, 24:885–911

2008

[11] [11]

Saari, D. G. (1995).Basic Geometry of Voting. Springer

1995

[12] [12]

Satterthwaite, M. A. (1975). Strategy-proofness and arrow’s con- ditions: Existence and correspondence theorems for voting procedures and social welfare functions.Journal of Economic Theory, 10(2):187–217

1975

[13] [13]

Young, H. P. (1974). An axiomatization of borda’s rule.Journal of Economic Theory, 9(1):43–52

1974