Numbers are not power until they are organized.

Mosca's argument, made in 1896 and not seriously dislodged since, is that a minority governs not because it is wiser or better born, but because it is organized against a majority that is not. Numbers alone are inert. Coordination is what converts them into power.

Schattschneider sharpened it sixty years later. The decisive power is not the vote but the prior question of which alternatives get considered at all, and the chorus of organized interests, he observed, sings with an upper-class accent. Whoever frames the ballot has won most of the argument before anyone votes.

This is why a ballot explainer is a political intervention and not a public service.

Tocqueville supplies the last piece, and it is the reason this site is organized around counties rather than states. Local institutions are where citizens actually learn the practice of liberty. The county is not a design convenience. It is the scale at which the practice is available to an ordinary person.

The response is not to hand government to experts, and not to exhort citizens to care more. Both have been tried for a century. The response is to change the conditions under which citizens act: lower the cost of knowing what is on the ballot, who is paying for it, and what can be done about it, at the scale where one person's participation is not negligible.

Paine is the ancestor here, more than any of the others. He wrote in deliberately plain language because comprehensibility was itself the political act; the pamphlet was simultaneously the argument and the distribution method. That is the standard this project holds itself to.

That third move is the product. The first two are why it matters.

Anything built on the premise that better civic information matters has to answer the argument that it changes nothing. The best version of that argument is Anthony Downs's.

Rational ignorance. A single vote decides essentially nothing. The expected benefit of becoming informed is therefore near zero, while the cost in time and attention is entirely real. On this account, remaining ignorant is not a character failure. It is the correct decision, arrived at rationally. Publishing more information changes nobody's behavior, because cost was never the binding constraint. Pointlessness was.

That argument is serious and it cannot be waved away. But notice its shape. Downs's conclusion rests on a product of two terms: the cost of knowing, multiplied by how little one's participation matters. Most civic technology attacks only the first term, which is why most civic technology fails on exactly Downs's prediction.

The county scale attacks the second term, and attacks it harder.

One signature among the 5,074 that compel Nevada County's board of supervisors is not a lottery ticket. It is roughly one part in five thousand of a binding outcome, several orders of magnitude removed from one vote among eleven million in a statewide race. Downs is right about the state. He is much weaker about the county, and the threshold for every one of California’s 58 counties is published, sourced, and checkable.

This is why the local threshold, rather than a statewide target, is the organizing device of this entire project. It is not a presentation choice. It is the load-bearing response to the strongest objection anyone has made to us.

California is where we start, not the limit of the argument. Every state has bodies whose decisions turn on the attention of very few people, and most set the numbers that bind them by statute. The figures differ, the statutes differ, and in some states the citizen initiative does not exist at all. The method survives all of that: find the real number in your jurisdiction, publish it, and work at that scale.

Lippmann and, in our own decade, Jason Brennan press a different line: that ordinary citizens cannot know enough to govern, so power should sit with those who can. Both are worth engaging rather than dismissing, and both are answered at length, along with Downs, in the philosophical foundation.