What is perverse incentive?

A perverse incentive is a reward or penalty that pushes people to act against the very goal it was meant to serve. It matters because well-meant rules can make a problem worse, and the effect is easy to miss until the results arrive.

How it works

The designer chooses a measurable target as a stand-in for the real goal. People then pursue the target in the cheapest way available, which may not help the goal at all. If a reward depends on the amount of a problem found or fixed, there is a reason to keep the problem alive.

The feedback loop is simple: the more the target is rewarded, the more behavior shifts toward gaming it, and the further the measure drifts from the goal. Catching these cases means asking, before launching a rule, how someone who only cared about the reward would act.

Example

Say a city pays residents 5 for every rat they bring in, hoping to cut the rat population. Some residents start breeding rats at a cost of 1 each, since each one brings 5. If 100 people breed 10 each, the city pays out 5,000 and the population goes up, not down. Removing the payment ends the breeding, but the lesson is that the reward was linked to the count of rats handed in rather than to the number of rats in the city. The figures are an illustration.

How to model it in Machinations

Create a Pool for the rat population with a Source for natural births and a Drain for rats caught. Add a Source for deliberate breeding whose rate depends, through a state connection, on the bounty Lever. A Pool tracks city spending. Play with the bounty at 0 and at 5 and chart both the population and spending to see whether the rule helps or hurts.