5 to 7 is the ideal size of the Board. The optimal number of people to make a joint decision, according to any management consultant. Small enough to avoid chaos, large enough for a good discussion, and if you keep the number odd, no tedious ties. Decades of organisational psychology stand behind this rule, and it is not wrong.
The ideal size is an answer to a different question
The five-to-seven rule tells you the ideal size of a conversation. It says nothing about the ideal size of a good judgement. Political systems have quietly built themselves on the first while claiming to deliver the second – and the boardroom rule is in fact obeyed, just not where you would expect. Vallentuna’s council has 51 seats. The decisions those seats ratify are made in two rooms of roughly six people: the governing parties’ leadership group, and the executive committee of the municipal board, where the leading opposition politician and the chief executive of the administration also sit. By the time an issue reaches the full board, it is anchored, and the outcome is known; by the time it reaches the council, the vote is a formality and the debate is for the record. The Swedish Riksdag, at 349, is the same theatre with a larger stage.
The theorem the boardroom forgot
In 1785 the Marquis de Condorcet proved something that should have ended the boardroom’s political career on the spot. If a group of people each has a better-than-even chance of judging a question correctly, and they judge independently, then the majority is more likely to be right than any single member – and the larger the group, the more certain the majority becomes. Toward infinity, toward certainty.
The theorem cuts both ways, and honesty requires saying so. If the average member is worse than a coin toss, a larger group is more reliably wrong. Condorcet is not a promise that crowds are wise. It is a precise statement of the conditions under which they are – and a warning about what happens when those conditions fail.
Those conditions are four:
1. The question that has a better and a worse answer
2. Group members are, on average, competent on it
3. Judgments are independent of one another.
4. There are enough judgments to matter.
The rest of this post is about those four conditions, because a democracy that takes them seriously looks nothing like a boardroom.
“But politics isn’t about truth”
This is the standard objection, and it deserves a straight answer. Whether a town should build an ice rink or a preschool is not a fact. No amount of voting makes it true or false. So what business does a jury theorem have in politics?
Look at what a political decision actually is. Nobody votes on “ice rink, yes or no”. They vote on a model: the rink will cost this much, be used by these people, crowd out these alternatives, and the town will be like this in fifteen years if we build it and like that if we don’t. Every political decision is a forecast dressed as a choice.
Forecasts are factual. They come true or not, and someone who consistently forecasts municipal budgets better than someone else is, in the Condorcet sense, more competent. Then comes the second layer: how much the town should care about hockey players versus four-year-olds. That is normative, and no theorem settles it.
So political decisions are both. They are factual in the model and normative in the weighting. The boardroom handles both badly – it forecasts with a tiny n, and legitimizes its values by the majority vote of 51 people who happen to have been placed on a list. Peer Democracy separates the two. The factual layer is where Condorcet applies, and where scale helps. The normative layer is where legitimacy applies: the weighting belongs to everyone affected, bounded by a moral floor that is not up for a vote. With Peer Democracy comes a moral compass. Every decision must take into account and answer “no” to the question: would this cause unnecessary suffering?
Building for the four conditions
Once you take Condorcet’s conditions seriously, the design writes itself. Peer Democracy is what you get when you build a decision system around the four conditions instead of around a meeting room.
Competence. The boardroom’s answer to competence is generalists: the same few people decide on the rink, the sewage plant, and the school curriculum. Peer Democracy does the opposite. Each citizen chooses a small number of issues – two per year – and votes only on those. This is not a concession to busy lives; it is the mechanism that lifts average competence above one half. People self-select into the questions they know, and the theorem only works from there. A council member who votes on 200 items a year has, on most of them, a p closer to 0.5 than any self-selected citizen.
Independence. The party whip is a machine for destroying independence: it turns 51 judgements into perhaps four, and in practice into the two rooms described above. Anonymous voting removes the whip, and it removes the status games of the public debate where the loudest voice shapes the room. But anonymity does not make people independent of each other’s information. Everyone reads the same feeds. This is the condition a platform can improve but never fully secure, and any system that claims otherwise is lying. What Peer Democracy can do is structure deliberation before the vote so that arguments – not personalities – are what circulate. Anonymous pro-and-con debate with clear rules and an AI supervisor that helps improve arguments are means to combat groupthink and raise p over 0,5.
The provocation, stated plainly
My municipal council is a jury of 51 whose members have been selected for loyalty rather than competence on any given question, whose independence is abolished by design, and who vote on questions they did not write. Measured against Condorcet, it fails three conditions out of four before the meeting starts.
The boardroom is the right size for a conversation. Keep it for that. Then let the judgement be made by the people who know the question, on their own, in numbers – and let the values be weighed by everyone the answer will touch.
Condorcet did the mathematics in 1785. We have spent the time since perfecting the meeting room.