Persistent Voting

Whenever any group of people organizes a vote for any issue, they unavoidably make a critical assumption: elections are discrete events with an arbitrary deadline and a final outcome. However there's nothing inherent to collective decision making that makes this necessary.

The discreteness assumption has made voting theory quite obsessed with strategic voting, the degree to which a voting method is sensitive to the outcome predictions of voters. This pulls attention away from the more important question of finding outcomes that make the electorate happy, or that maximize welfare.

Changing our election processes to be continuous rather than discrete could allow election outcomes to depend less on "horse race" guessing games, and more on the real push/pull of an underlying social negotiation. Removing the discreteness assumption is the core insight at the heart of Persistent Voting, which itself forms the basis for a broader Persistent Democracy.


At any time a polity has some set of positions decided by elections. These positions can be for anything, even questions traditionally decided by representatives such as legislation or budgets, but we'll leave discussion of other applications for future chapters.

Rather than having scheduled election events, citizens could update their votes for any of the positions at any time, and candidates could exit a race or be nominated to enter a race at any time. I'll discuss nomination in a future chapter.

To prevent rapid switches of power with a narrow margin between two candidates, there would be a "stable period" for which a newly winning candidate would have to maintain their advantage before they officially took power. The length of the stable period would be inversely proportional to the size of the winning margin, with some low number acting as a floor. The stable period would reset if a newly winning candidate lost their advantage before the period had elapsed.

For example, if some candidate overtook another in an election by a tiny amount, they would have to maintain that advantage for a long time (perhaps several months), whereas a massive advantage would change the outcome as soon as possible (perhaps the next day or week, depending on the lower floor number).

It would probably also make sense to only publish all vote changes on some reasonable schedule, such as daily or weekly. This would save voters from anxiously updating their votes too often.


I'm excited about this idea because it unravels many limiting assumptions about elections and social coordination.

  • A shorter feedback loop between decisions and their consequences would help groups move more nimbly towards socially optimal decisions.
  • Elections would be less sensitive to noise, politicking, campaigning, and predictions about the outcome.
  • Allows "auto-pilot democracy." With Persistent Voting, voter turnout becomes essentially a nonissue. If some concerned minority makes most of the electoral choices, and everyone else is perfectly fine with those choices, then the majority can simply not participate without worry. If the concerned minority begins making harmful or foolish choices, the rest of society can quickly override them before harm is done.
  • Dramatically mitigates the cost of "surprises" in elections, since a surprised electorate can merely counteract a negative outcome before it does real harm. Positive surprises are also much more possible, since their possibility can be detected and reinforced more easily.
  • Potentially allows more honest and expressive voting methods to be used, since strategic voting might matter less.
  • Could allow more complex voting methods, since voters have unlimited time to think about and understand how they wish to vote.
  • Could allow a more direct democracy. Since voters can change their vote in any election at any time, voters can participate in only the elections they're actually concerned about while leaving other decisions for other voters. Many more decisions could be made democratically without overwhelming the electorate.

Potential Objections

  • Could potentially be unstable and confusing. I think this would depend very much on the exact parameters of the system, such as the equation for stable periods or the frequency of vote publication.
  • Might lead to short term thinking, electorates who change outcomes too frequently for any system to fully develop and work.
  • Much more logistically and infrastructurally complex. A future chapter discusses some ideas to make Persistent Voting tractable.

Open Questions

  • What kind of equilibrium outcomes will this method converge towards, if any?
  • What honesty incentives does this method give to voters?
  • Will elections converge towards the optimal utility outcome?
  • How does this method change the incentives of candidates? What kind of candidates will enter/exit races and under what conditions? How will this change how society as a whole changes over time?
  • How much does the trend of elections depend on the inner voting method used (Plurality, Quadratic, Approval, etc)?

I have some conjectures about these questions, and I intend to formalize and attempt to prove them with the Coq proof assistant.

  • Most of the inefficiency in existing voting methods stems from the voting method giving the whole electorate an inaccurate picture of itself. Each voter is misinformed about the true preferences of their fellow voters when a voting method punishes a voter for expressing their honest views. This distortion manifests itself when candidates who actually have broad support aren't able to demonstrate that support and therefore are "buried" by the voting mechanism and forgotten. This gap between perception and reality is what really makes strategic voting so poisonous.
  • Persistent voting will cause all voters to trend toward the most strategically optimal vote, since the mechanism will reveal the preferences of other voters and cause each to change their vote to prevent/support imminent outcomes. However, because the mechanism progressively reveals the real preferences of the electorate, the distance between the honest and strategic votes will get smaller, and only depend on the inner voting mechanism rather than various pathologies of the surrounding society. So for inner voting methods that allow strategically powerful votes that still expressively reveal a large amount of honest information, the votes will trend toward both the strategic and honest votes.
  • Enabling gradual electorate discovery that closes the gap between electorate perception and reality is the real crux of what makes persistent voting so powerful. The stable period mechanism allows the electorate to discover the possibility of surprising negative outcomes and counteract them before they actually take place.
  • The trend of elections will depend very strongly on what inner voting method is used, but all voting methods will gradually converge towards the optimal utility outcome as the true preferences of the electorate are gradually discovered. Very inexpressive or strategic methods however will take much longer to do so, and might spiral in strategic loops as different groups oscillate between honestly expressing their preferences and defending themselves with a more strategic vote.
  • The shorter feedback loop between decisions and consequences will also enable the electorate to converge towards higher experienced utility, rather than the more imaginary projected utility voters use to make decisions in elections.

Table of Contents

In Defense of Pure Logic
Persistent Voting
Quadratic Range Voting
Persistent Documents
Persistent Prioritization
Persistent Endorsement
Persistent Commitments
Persistent Funding
A Theory of Minimum Necessary Rights
Markets and Rights
Common Resource Taxes
The Crowdsell Mechanism and Intellectual Property
Democratic Districts
Free Borders
Persistent Logistics
Democratic Adjudication
Misinformation Trials
Member Cooperatives and Economic Activism
Make a contribution