The Tenth of a Percent

By Ricardo Vega

There is a room. It's usually beige, usually too cold, and there are usually more empty chairs than full ones. Nine people speak. A vote is taken. Forty thousand people are now governed by what happened in that room, and thirty-nine thousand nine hundred and ninety-one of them will never know it took place.

This is the part of the story where I stop describing a problem you already suspected and put a number on it, because the number is worse than you think.

Researchers collected and transcribed a decade of city council meetings across a hundred and fifteen California cities — not agendas, not minutes, the actual spoken record. What they found is that roughly one in a thousand registered voters speaks at a council meeting in a given year. A tenth of one percent. And most of those are repeat participants: the same faces, meeting after meeting. So the true number of distinct citizens whose voices enter the formal record is smaller still.

Broaden the definition and it doesn't rescue you. Count anyone who attended any local public meeting about any issue in the past year and you get to about eleven percent nationally. That's the generous figure. Eighty-nine out of a hundred people had no contact with local governance at all.

Now, a tenth of a percent would be survivable if it were a random tenth of a percent. Statistics is built on the idea that a small sample can speak for a large population — provided it's drawn fairly. This one isn't.

The people who speak at council meetings are older than the voting population. Whiter. More male. More likely to own their home rather than rent it. Every one of those skews points the same direction: toward people with time, with stability, and with an asset whose value is affected by what the council decides.

And then the finding that should stop you cold. When researchers looked specifically at meetings about affordable housing — hearings on actual proposed developments — the people who turned out were disproportionately the ones opposing the project. Not split. Tilted.

Sit with what that means. The single formal channel we have for reading what a neighbourhood wants is not merely quiet. It is biased, consistently, in one direction, against precisely the class of thing that helps people who are falling. It's a thermometer that reads cold.

This explains something that otherwise looks like hypocrisy.

You know the pattern. A city announces a supportive housing project, or a shelter, or a service centre. There's a hearing. Forty people show up furious. The story writes itself: the neighbourhood doesn't want it, people say they care about homelessness until it's on their street. And there's some truth in that — but it's not the whole truth, and treating it as the whole truth has cost us a decade.

Because forty furious people out of a neighbourhood of nine thousand is not a neighbourhood's opinion. It's a sample of the people for whom this issue crossed the threshold of showing up on a Tuesday night. Opposition clears that threshold easily; fear and property values are excellent motivators. Mild support does not clear it at all. Nobody has ever taken an evening off work to say "that seems fine to me."

So we build systems that read only the strongest negative signal in a neighbourhood and call the result democracy. Then we act surprised when the people who did want it — who were at work, or with their kids, or simply never heard — turn out later to have existed all along.

Here's where I have to be honest about a temptation, because I had it. If the wrong people are being sampled, the obvious move is to go get a better sample — reach the ninety-nine point nine percent where they actually are, on their phones, with the same machinery that sells them everything else. And what makes people respond on their phones is outrage. I got far enough down that road to design the campaign before I saw what was wrong with it, and the anatomy of that mistake deserves its own piece — it's the next one in this series. The short version: if you provoke with disgust, you measure disgust. You manufacture the very signal you claimed to be reading.

The honest instrument is available, and it's stranger than the dishonest one: tell people the truth about the waste.

Within a mile of where you're standing, tonight, this much edible food goes into a dumpster. Here's what it costs to throw away. Here's what could be done with it. What should happen to it? That provokes without lying. It provokes about a fact rather than about a population. And the anger it produces points at a system rather than at a person sleeping in a doorway.

And it can reach people the beige room never will. The parent at work Tuesday night. The renter who doesn't know what a zoning variance is. The twenty-six-year-old who has never in their life heard of a council meeting. You will not get everyone. But if you can get five percent to answer one honest question, you have a sample fifty times larger and far less tilted than the room, and for the first time you'd have something worth calling the will of a neighbourhood.

Which matters enormously, because this series is heading toward moving things — and a delivery system with no legitimate address is just trespassing with extra steps. You cannot route resources into a community that never asked for them. The routing question and the participation question are the same question, and this is the one we've never bothered to answer.