Every Policy Is an Incentive System Wearing a Statute

In government you learn quickly that the text of a law and the behavior it produces are two different documents. The first is written by people describing a world they want. The second is written by everyone who has to live inside it, and it is written in decisions rather than sentences.

This is not cynicism. It is mechanics. A statute allocates money, risk, time, and permission. Those four things constitute a payoff structure, and payoff structures are read fluently by anyone whose livelihood depends on them.

Legislatures debate intent. Markets, agencies, and households respond to payoff. Only one of those is binding.

Three ways good policy goes wrong

Almost every failed program I have reviewed failed in one of three predictable ways, and none of them required bad actors.

• —The measured outcome is easier to produce than the intended outcome — so it gets produced instead.
• —The compliance cost falls on the party with the least capacity, so participation collapses among the intended beneficiaries.
• —The reward arrives after the political cycle that would have punished failure, so nobody owns the result.

Designing policy the way you would design a mechanism

Before a policy moves, run it through the same interrogation an engineer runs on a system: who benefits fastest, who pays first, what is the cheapest way to appear compliant, and what happens at the boundary conditions.

That review takes days. Recovering from a policy that pays for the wrong behavior takes years, and the people who absorb the cost of the error are rarely the people who wrote it.