Ask what the rule rewards

A rule or target changes what everyone does. Before you set one, ask what is the cheapest way to score on it, because that is what you will get.

Steven Kerr’s paper “On the folly of rewarding A, while hoping for B” (1975) collected cases from business, government, universities and medicine where an organisation rewarded one thing while hoping for another, and then was surprised to get what it rewarded. His point was not that people cheat. It was that the reward system, not the speeches about the goal, tells them what counts.

The economist Charles Goodhart made a similar point about monetary policy in 1975: a statistic that moves with what you care about tends to stop doing so once you start steering by it. The anthropologist Marilyn Strathern (1997) put it in the form now quoted as Goodhart’s law: when a measure becomes a target, it ceases to be a good measure.

The bonus can even make the problem worse. In 1902 the French rulers of Hanoi paid a bounty for each rat tail handed in. People soon cut off tails and let the rats go to breed, and some farmed rats for the bounty (Michael Vann 2003). The same kind of story is told of cobras in Delhi under British rule, giving the name cobra effect, though that version may be a legend.

So think one step past the rule: how will people respond to it, and what will that do? Pay for the result you want rather than the activity that usually goes with it, and watch what the shortcut would harm. No measure is perfect, so look again once the rule is running.

Why it matters

People are good at meeting the measure they are paid on. When the measure and the goal part ways, the measure goes up and the goal does not, and the numbers hide it.

Away from the game

Before you set a bonus, a target or a fine, imagine someone who wants the reward and does not care about the goal. What would they do? If that is not what you want, pay for something closer to the goal itself, or measure what the shortcut damages too.