Be careful what you ask for. You might get it.

In economics and public policy, there’s something known as the law of unintended consequences. “The law of unintended consequences, often cited but rarely defined, is that actions of people, and especially of governments, always have effects that are unanticipated or ‘unintended,’” writes business writer Rob Norton.

In fact, these actions can even make matters worse. The effect was first studied in 1936 by Robert Merton, but there are dozens of examples of laws that ended up making matters worse. There’s even a name for it: the Cobra Effect, named after one of the best-known examples, where Indian citizens who were offered a bounty on cobras ended up actually raising cobras to turn in for the bounty – and when the government eliminated the bounty, the people released the cobras, resulting in even more cobra deaths.

So, what does this mean to you? Especially if you’re not an economist or a bureaucrat and don’t deal in cobras?

The point is that the law of unintended consequences applies to any sort of policy where there is some sort of reward that is predicated on human action. The classic business consultants W. Edwards Deming and Peter Drucker have both been said (inaccurately, apparently) to be the source of the statement, “If you can’t measure it, you can’t manage it.” Consequently, managers — perhaps overly influenced by Frederick Taylor — love to look for “objective” things to measure to gauge employee performance.

The problem is, once you start measuring something – particularly if you start grading people’s performance or salary based on it — people will change their behavior accordingly. One of the classic examples is setting up a help desk and grading staffers’ performance based on how long the calls last. Consequently, the staffers make calls as short and quick as possible, even if they don’t solve the user’s problem. This means that users sometimes have to call back a couple of times to get their problem solved, resulting in user dissatisfaction. (Here are some other great examples.)

It doesn’t need to be a measurement, either. For example, making Internet policies too restrictive can mean that users will look for ways around restrictions rather than following them. The result is that you end up with even less control over what users are doing with protected corporate information.

It can be tough to figure out a reasonable proxy to measure that doesn’t end up resulting in an unintended consequence. (In the case of the help desk, for example, some companies have determined that what they really need to measure is the number of users whose problem is solved on the first call, no matter how long that first call lasts.)

The point is, people are people, and they don’t always react the way you expect them to. Or, they do — you just have to think about what motivates them, and pick your measurement accordingly.

Related Posts