Governing Ourselves: A Blueprint for Manifold Futarchy
When @bens suggested a Manifold Government, many users worried that it would become an oligarchy, an elite club, old-timers and whales calling all the shots. Futarchy lets us avoid that.
A veteran's reputation doesn't carry more weight than a newcomer. Decisions get made based on what will actually produce the best outcomes, because everyone has financial incentives to identify the policies that will truly work.
How It Works First, we agree on our values (see roadmap below); what we want the government to achieve. For example: fair resolutions for subjective markets.
Imagine someone creates a market: "Will AI be friendly in 2026?" It has no resolution criteria and the creator is inactive. We don't argue endlessly in comments. Instead, we let the market guide us:
People propose resolution criteria, like:
Proposal 1: "Resolve based on a poll of 10 recognized AI safety researchers."
Proposal 2: "Resolve YES if no one is killed by AI in 2026."
We open prediction markets:
"What will the satisfaction score on resolution criteria be if we adopt Proposal 1?"
"What will the score be if we adopt Proposal 2?"
The proposal with the higher forecasted satisfaction wins. Everyone can bet; decisions are made based on expected value, not those with reputation, loud voices, account age or manna.
Tentative Roadmap
Legitimacy Vote (simple headcount): We need to decide whether to pursue this through Ben's constitutional convention or with our own poll asking "Should we explore Manifold futarchy governance?" Either path needs 60%+ support to proceed.
Values Survey (simple headcount): Multiple choice poll (with user additions open) on core values - any option getting 50%+ makes our official goals list. Anything at this stage (e.g. consolidating 2 goals into one) is voted on in a poll with a simple headcount.
Constitutional Convention: Elect 7-11 drafters via approval voting to write our constitution. The constitution should include 3 articles.
Our values as decided by stage 2
Inalienable rights (e.g. Active market creators retain full control of their markets - barring mod actions)
Mechanisms and Processes: Weighting of the Futarchy Meta-Index, when are polls created, minimum volume / liquidity on decision markets etc.
Public Drafting Period: 2-3 weeks of open drafting with community input
Constitution Ratification (simple headcount): Final up/down vote - needs 67% to pass
We elect The Stewards: our initial moderation team through a process the community defines, potentially using Futarchy forecasts about their future performance.
We build the basic infrastructure (posts page, dashboard, market templates) to run prediction markets on government decisions.
Example Goals (from @crowlsyong Constitution draft)
Establish fair resolution criteria for subjective markets
Provide advice for controversial markets
Manage a community treasury (liquidity, prizes, bailouts)
Award meaningful honors like Market of the Month
The Hybrid Model We combine quick community feedback with longer-term metrics to guide both immediate and lasting success:
Weekly Pulse: Each goal will have its own satisfaction metric and poll, so we can measure how well we're meeting specific objectives. For example, a poll on whether we're establishing fair resolution criteria for subjective markets. This provides fast community feedback.
Futarchy Meta-Index: A composite of (for example):
Community Satisfaction (40%)
Resolution Quality (20%)
Treasury Impact (20%)
Engagement Growth (20%)
Example Decision Process Every policy triggers prediction markets with fixed timing:
Proposal Window: First week of each month
Market Period: 1 week of betting on "What will the Futarchy Meta-Index be if we adopt this?" vs. "What will it be if we don't?"
Decision: Automatic implementation based on higher predicted outcome
Satisfaction Polling: 1 week after implementation
(Might not need the following due to the hardworking mods)
Enforcement: The Stewards A small team - The Stewards - keeps the system clean, preventing spam and low-quality proposals. But they're accountable too: if markets forecast that replacing them would improve our metrics, they're replaced automatically. No entrenched bureaucracy, just forecasts and outcomes.
This is governance designed for a people that believes in prediction markets. It rewards what works, not the landed gentry.
JOHN HANCOCK
Here's the problem: whoever has the most mana can change the market outcome to whatever they want. If you're willing to burn mana you can move the percentage to whatever you want.
There's no current way on Manifold to set a limit as to how much mana a user can spend on a market. Without that in place, one user can manipulate the market however they want. Someone with very small amounts of mana cannot meaningfully contribute and their voice is muted or silenced by whoever has more currency.
@BlackCrusade But then they lose their money when the poll comes out!
Say the question is "Will the treasurers approval rating go up if he adds liquidity to X market" Whale A dumps 100k on YES, the treasurer adds the liquidity, the approval rating poll is launched and no one liked the idea, Whale A loses 100k!
You have to predict correctly to make any manna, there's no room for whalebait
@BlackCrusade I have to assume manifold the business has some counter measures. They give out free mana to whoever joins so if it's easy to make 100 sock puppets you have a money (mana) printing machine
@FergusArgyll the problem is that assuming it won't be a problem because of a feature you assume exists does not make a convincing argument for someone like me who is skeptical and convinced this will result in rampant abuse.
@BlackCrusade we can ask @ian but I'll say I haven't noticed any accounts that managram themselves from 100 new accounts. I think the only login is with google - that makes it pretty hard.
What are some abuse scenarios you're nervous about?
@FergusArgyll my main issue is creating a dichotomy of less-than users and important users.
What you are proposing seems to work better as a system implemented byanifold staff rather than a user run government.
@BlackCrusade Yeah the reason I thought you'd like it is because no one has any title or position of power. It basically runs itself after it's set up
@FergusArgyll that's far preferable to people deciding they are representing others. I do agree with you there.
I think your ideas will be unpopular with other framers precisely for that reason though, as it seems that some people envision themselves as in charge and are hoping this provides them power over others.
@FergusArgyll We do have countermeasures for people signing up lots of sock puppets, and you can also check the global payments page here to report users using alts to harvest bonuses