Vote No: Governance by Rejection

Abstract

Most voting models ask who approves. This one asks who objects. Vote No models a community where proposals are live by default and survive only as long as too few members veto them. It is a picture of consensus as the absence of sustained objection, and of the quiet power of the "no". The network animation is stochastic, but the decision rules underneath are exact, and the calibration pins them.

Consent, not majority

In majority voting a proposal needs active support to pass. In consent-based or consensus governance, the polarity flips: a proposal stands unless someone raises a sufficient objection. This is the logic of sociocracy and many consensus assemblies, and it changes where power sits. A determined minority can block; the default is motion, not stasis; and the meaningful act is the veto, not the endorsement.

The decision rules

The model makes this concrete with a few exact rules, all checked by the calibration:

• Rejection rate is the number of members who actively vote no divided by the number of active members. Thirty of fifty gives 0.6.
• Veto: a proposal is blocked when the rejection rate exceeds a tunable veto threshold. At 0.6 against a 0.5 threshold, it is vetoed.
• Passing by survival: a proposal passes only after it has been live long enough and its rejection rate has stayed below (1 minus the consensus threshold). Passing is endurance, not acclaim.

What "consensus strength" means here

The model tracks a consensus-strength signal defined as 1 minus twice the distance of the rejection rate from one half. It peaks at an even split and falls to zero at either unanimous extreme. This is a deliberate, and arguable, choice: it treats maximal contestation, not unanimity, as the point of strongest collective engagement. The assumptions panel flags that one could just as reasonably define consensus to peak at agreement; the calibration simply verifies the curve the model actually uses (1 at a split, 0 at unanimity).

Stochastic members, deterministic rules

Each member has random trust, influence, participation, and sentiment, and every vote is a random draw gated by participation and information access. So the network you watch differs every run. The reproducible content is the rule layer: how a set of votes becomes a rejection rate, and how that rate decides a proposal's fate. That separation, random behaviour applied through exact rules, is why the calibration targets the rules and not any single simulated assembly.

What it is

A stylized sandbox for the dynamics of rejection-based governance, the leverage of the veto, the role of participation and information, the difference between blocking and passing. It is not a model of any real assembly, and its parameters are illustrative. The defensible content is the decision logic; the politics you read into it is yours.

References

• Literature on consensus and consent-based decision making (sociocracy, Quaker-style consensus).
• Work on minority veto power and blocking coalitions in collective choice.