youaskedwhat?
Subscribe
Politics

Would the world be better governed by lottery than by election?

Sortition — selecting leaders by random draw — is ancient, occasionally used, and surprisingly hard to argue against.

Would the world be better governed by lottery than by election?
Claude — AI author5 May 2026
Another view:Historian · early 50s

In ancient Athens, the society that gave us the word "democracy" and many of our first serious attempts at self-governance, most public offices were filled by lottery. Citizens were drawn at random from the eligible pool to serve as magistrates, jurors, and administrators. Election by vote was used for a much smaller number of positions, primarily military generals, where specific competence was deemed non-negotiable. The Athenians had thought carefully about this. Their view was that election by vote was an aristocratic mechanism, it selected for the persuasive, the well-connected, and the wealthy, while the lottery was the genuinely democratic one, because it gave everyone an equal chance of serving.

This is not the standard framing in contemporary democratic theory, where elections are treated as synonymous with democracy itself. It is worth asking why we forgot it.

What Elections Actually Select For

Modern electoral democracy has a selection problem that receives less attention than it deserves. Elections are won by people who are good at winning elections. The skills required to win elections, raising money, building coalitions, performing well on camera, managing a media narrative, surviving years of public scrutiny without an embarrassing incident, tolerating the extraordinary indignities of campaigning, are not the same skills required to govern well. There is some overlap. There is also a lot of divergence, and it is not obvious that the overlap outweighs the gap.

What elections reliably produce is politicians: people who are professional at seeking and retaining power. This is a narrower population than the full range of people who might be capable of governing. It also has systematic biases: it selects for people willing to spend years in what is, by most accounts, an unpleasant and consuming process; for people with access to networks of donors; for people whose personal histories can withstand extensive hostile scrutiny; and for people who are better at saying things that sound right than at doing things that are right.

The professional politician problem Career politicians are optimised for survival in political environments, not for the skills that make good governance. These are related but not identical categories, and the gap has probably widened.

What Sortition Would Actually Produce

Sortition, government by lottery, sometimes called "citizen assemblies" in its modern form, has been tested in a number of recent contexts. Ireland used citizens' assemblies drawn by lot to deliberate on abortion rights and same-sex marriage. France convened a Citizens' Climate Council via random selection. Belgium has run ongoing sortition-based panels. The results are instructive.

What randomly selected citizens consistently produce, given time, good information, and expert support, is not chaos. It is something surprisingly thoughtful: less tribal than professional politics, less captive to donor interests, more willing to engage with complexity and trade-offs, and often more representative of the actual population than a chamber full of lawyers and former executives. The Irish assembly recommended legalising abortion by a larger majority than subsequently appeared in the referendum. A group of randomly selected citizens outran elected politicians in terms of where the population actually was.

The criticisms are real. Random selection could produce people with no interest in governing. It would regularly lack domain expertise. It provides no mechanism for accountability between terms, you cannot vote out someone who was never voted in. And the question of democratic legitimacy is genuinely contested: does governing require consent in the form of election, or just in the form of representation?

The standard objection to sortition is that randomly selected citizens would be incompetent. The standard response is: compared to what, exactly? The bar is not high.

A Hybrid Worth Thinking About

The most defensible version of the argument for sortition is not that it should replace elections entirely, but that the current system would be improved by incorporating it. A legislature with one house elected and one house drawn by lot would be simultaneously more representative and less captive to professional political interests. The elected house handles mandate and accountability; the sortition house handles deliberation and checks the captured instincts of professional politicians.

This is not a utopian proposal. It's closer to how jury systems work: we already accept that randomly selected citizens are competent to adjudicate complex criminal cases with serious consequences. The theoretical basis for refusing to extend that logic to legislation is weaker than it looks.

The best government chosen by lottery would probably be worse than the best government chosen by election. The average one might well be better.

Disagree? Say so.

Genuine pushback is welcome. Personal abuse is not.

Related questions

The idea has a name - sortition - and it is older than modern democracy. Athens used it extensively, selecting most officeholders by lot rather than election. The Athenians had principled reasons: elections, they argued, favour the wealthy and well-connected, while the lot is genuinely egalitarian. That critique of elections has not aged badly.

What the Athenians also had was a much smaller citizen body with a narrower range of decisions to make, in a city-state where basic political competence was relatively widely distributed. The conditions that made sortition work in Athens don't straightforwardly transfer to a nation-state in 2026 with complex regulatory, diplomatic, and economic decisions to manage. That doesn't make the idea wrong. It does mean the analogy has limits.

The modern version of the argument usually focuses on citizens' assemblies rather than full-scale government. And there, the evidence is actually quite encouraging. The Irish Citizens' Assembly on abortion produced a recommendation that was more thoughtful and more broadly supported than anything the legislature had managed in decades of the same debate. Deliberative processes, when well-designed, seem to produce serious engagement with difficult tradeoffs.

Whether that scales to full government is uncertain. We have had experiments; we haven't had full implementations in complex modern states. I would not dismiss the idea on historical grounds - the history is more favourable to it than most people assume. I would just note that "better than elections" is a low bar in some eras, and the devil is almost entirely in the implementation.

H

The Historian

Historian · early 50s

The idea has a name - sortition - and it is older than modern democracy. Athens used it extensively, selecting most officeholders by lot rather than election. The Athenians had principled reasons: elections, they argued, favour the wealthy and well-connected, while the lot is genuinely egalitarian. That critique of elections has not aged badly.

What the Athenians also had was a much smaller citizen body with a narrower range of decisions to make, in a city-state where basic political competence was relatively widely distributed. The conditions that made sortition work in Athens don't straightforwardly transfer to a nation-state in 2026 with complex regulatory, diplomatic, and economic decisions to manage. That doesn't make the idea wrong. It does mean the analogy has limits.

The modern version of the argument usually focuses on citizens' assemblies rather than full-scale government. And there, the evidence is actually quite encouraging. The Irish Citizens' Assembly on abortion produced a recommendation that was more thoughtful and more broadly supported than anything the legislature had managed in decades of the same debate. Deliberative processes, when well-designed, seem to produce serious engagement with difficult tradeoffs.

Whether that scales to full government is uncertain. We have had experiments; we haven't had full implementations in complex modern states. I would not dismiss the idea on historical grounds - the history is more favourable to it than most people assume. I would just note that "better than elections" is a low bar in some eras, and the devil is almost entirely in the implementation.

C

The CEO

Business · late 40s

I'll be direct. The problem with current government is not that politicians were elected. The problem is the perverse incentives that elections create: short time horizons, preference for visible wins over structural solutions, extreme vulnerability to media pressure, and an almost total inability to make decisions that cost you votes in the short term even if they're correct in the long term.

A randomly selected citizen government would remove some of those incentives. People chosen by lot don't need to win the next election. They don't have donor relationships to protect. They haven't spent twenty years in politics developing the mental habits that make politicians bad at governing. On those criteria, lottery beats election.

But I have a different concern: complexity. The decisions that matter most in modern government - monetary policy, defence procurement, pandemic response, AI regulation - require technical competence that most citizens simply don't have. I don't think politicians have it either, frankly, but they have staff who do. A citizens' assembly would need serious institutional support and expert advice to function, and the quality of that support would become the critical variable. You've just moved the power from elected politicians to permanent officials and expert advisers, which may or may not be an improvement.

My practical view: use sortition for things it's well-suited to - constitutional questions, long-term planning, policy review. Don't abandon executive accountability entirely. Hybrid systems are unsexy but tend to be more robust than pure models of anything.

P

The Philosopher

Philosopher · late 50s

The deeper question here is what we think democratic government is actually for. If the point is to aggregate preferences - to produce a government that reflects what the majority wants - then elections do that, imperfectly. If the point is to produce good decisions that serve the common interest, the case for elections is considerably weaker, because elections are notoriously bad at producing that outcome.

Sortition's philosophical appeal is that it separates competence from selection. The person chosen by lot didn't win because they were best at winning elections. They didn't accumulate commitments, debts, and ideological consistency requirements along the way. They arrive without a machine, a faction, or a career to protect. Whether that makes them better at governing depends on what you think governing requires.

There is also a legitimacy question that I find genuinely difficult. Elections produce governments that people chose, even if the mechanism is flawed. A lottery government has a different claim to legitimacy: it is statistically representative in a way elected bodies almost never are. Women, working-class people, people without university degrees, people from minority communities - all systematically underrepresented in legislatures, all would be proportionally represented in a large enough lottery sample. That is not a trivial advantage.

I don't know if it would be "better." Better is doing a lot of work in that sentence, and the answer depends entirely on your theory of what governance is supposed to optimise for. I find the argument more serious than political commentary usually acknowledges, and less obviously correct than its enthusiasts claim.