The Sequence · Part II · What To Do
06. The Hook
What if wrongdoers paid for their own busts?
Part I was philosophy, and I am aware of how it reads. Several thousand words on solipsism and moral patienthood and whether there is anybody home in the universe, ending in a promise that all of it cashes out in something as dull as policy design. This is the policy design. It starts with garbage.
I want to build the machine somewhere the stakes are far too low to distort your judgment, get it standing, walk around it, and only then point it at something that matters. So for the length of this page nothing is at stake except a soda can on the sidewalk.
A thousand dollars for a soda can
Imagine a jurisdiction that makes exactly one change to its litter law.
Littering carries a fine of $1,000. That part is not exotic — American highways are lined with signs threatening penalties in that range, and have been for decades. The unusual part is where the money goes. It does not go to the city. It goes to whoever brought the evidence that secured the conviction.
That is the entire statute. No new agency. No inspectors. No enforcement budget line, no hiring, no department that has to justify itself at the next council meeting. The legislature defines the offense, sets the number, and the courts go on doing what courts already do.
Now ask why litter was ever cheap to drop in the first place. Not because the penalty was small. The penalty was already $1,000. It was cheap because the expected cost of an offense is not the fine, it is — the probability of being caught, times the fine — and for littering has always been approximately zero. You can drop a cup on an empty street with total confidence. A fine that is never collected is not a deterrent. It is a decoration on a signpost.
Everything that follows is about .
What you would actually do
Watch what happens to a person who has no particular feelings about litter.
The camera. A decent outdoor camera costs about $50. If your street sees a few careless people a week, that camera is not a civic gesture, it is a piece of capital equipment with an obvious payback period. You are not buying it because you care about the appearance of your neighborhood. You are buying it because the sidewalk outside your window has just become a revenue opportunity and the equipment is cheap.
Everyone else’s cameras. Your neighbors can do arithmetic too. Within a year the block is covered, not by design and not by anyone’s plan, but because the same calculation was available to everybody who could see a sidewalk. Nobody coordinated this. Nobody had to.
The fingerprinting kit. Cameras have a specific failure mode: they show you a person, and a person is not yet a name. Footage of an unidentifiable hand releasing an unidentifiable wrapper does not pay. Identification is now the bottleneck, which means identification is now a market, and a $25 kit shows up to serve it. This step is worth pausing on, because it is the first one that is genuinely uncomfortable, and the discomfort is information — hold onto it for page 12.
The premium. Here is where it turns. Once the cameras are dense enough that detection is close to certain, you are not only a potential bounty hunter. You are a potential defendant. You drop things. Everyone drops things. Sooner or later you will drop something in front of a lens, and there is a four-figure liability attached to it. So when someone offers you coverage for a few dollars a year — call it $5, priced off a risk that is real but small — you take it, for precisely the reason you take any insurance. Not virtue. Not civics. You would rather pay a small certain amount than face an unlikely large one.
At no point in that chain did anyone act from public spirit. Every step was an obvious, slightly grubby response to the step before it, taken by someone pursuing their own interest with no interest whatsoever in the outcome the chain produces.
What just got built
Stand back and name the thing.
Detection is privatized. It is done by whoever found it worth their while to own a camera, and it costs the public nothing.
Prosecution is privatized. It is done by whoever holds the evidence and wants the money, and it costs the public nothing.
Punishment is a fine, and the fine is paid by the offender to the person who caught them. The enforcement system is funded, end to end, by the people it is enforcing against. There is no budget to cut, no agency to capture, no line item that competes with schools.
The state has retained exactly three jobs: defining the offense, setting the numbers, and adjudicating disputes. Those are not small jobs and I do not want to pretend they are — most of the hard parts of this proposal live in the first two, and pages 10 through 12 are largely about how badly they can go wrong. But notice what is not on the list. Nobody has to be hired to go looking.
And now the part that matters most, which is easy to miss because it is the absence of something. In equilibrium, almost nobody collects a bounty. Once detection is near-certain and the price of being detected is four figures, people simply stop dropping things. The machine’s success case is that it sits there, fully assembled, idle, and expensive to trip over. A regime like this working looks almost identical to a regime like this being unnecessary, and that is a genuine problem for anyone trying to evaluate one — but it is the target, and it will matter enormously when we get to what this is actually for.
Why start with garbage
Because you have no position on litter.
If I had opened Part II with the real subject, you would have started arguing with the conclusion before you had seen the mechanism, and you would have been right to. The stakes there are enormous, the political valence is loud, and almost everyone arrives with something to defend. Under those conditions nobody evaluates a mechanism; they evaluate whose side it seems to be on.
Litter is useful precisely because it is beneath your notice. Your intuitions about soda cans are not load-bearing for anything you believe about yourself, so you can watch the cameras and the kits and the premiums appear without having to decide in advance whether you approve.
I should be straight about the cost of this trick, though. Toy cases are toys because the objections are weak. As the stakes rise, the fingerprinting-kit discomfort you felt three paragraphs ago grows into something considerably worse — collusion between hunters and offenders, manufactured evidence, the chilling of legitimate behavior, and a culture of denunciation with an ugly history behind it. None of that shows up when the crime is a wrapper. All of it shows up later. I am not smuggling those past you by opening here; I am deferring them to the page that is entirely about them .
The littering case is not only a toy
One more thing before the handoff, because it is easy to read this page as pure allegory and it is not.
Offenses that are cheap to commit, hard to detect, and diffuse in their harm are exactly the class this mechanism suits, and quite a lot of real law falls into it — illegal dumping, fly-tipping, wage theft, small-scale environmental violations. Page 13 argues that something in that family is the correct first real-world test of the whole idea, and that anyone who wanted to find out whether this works should be running it there before running it anywhere consequential.
The soda can is where the argument starts. It is also, plausibly, where the argument should start in practice.
What is missing
Everything general. This page is a single worked example with the numbers picked to make it come out nicely, and an example is not a mechanism. What makes the chain run in the litter case, what makes it stall in other cases, why the fine must be insured rather than merely large, how you would size such a fine when the harm is not a soda can — none of that is here.
It is also not mine. The general form is Robin Hanson’s, from Privately Enforced & Punished Crime , and he states it far more carefully than a story about a sidewalk can.
That, and the arithmetic that makes silence expensive, is the next page .