The Cobra Effect: When the Fix Makes the Problem Worse

The cobra effect is what happens when a fix creates an incentive that feeds the problem it was meant to solve. The name comes from a story about colonial Delhi. Worried about venomous snakes, the British administration offered a bounty for every dead cobra. People began breeding cobras to collect it. When officials noticed and cancelled the bounty, the breeders let their now-worthless snakes go, and the city ended up with more cobras than before.

It's a perfect story, and that's the first thing to be careful about. The economist Horst Siebert popularised it, and the term, in his 2001 book Der Kobra-Effekt, but historians haven't found colonial records confirming it. Treat it as a parable. The pattern it describes, on the other hand, is very well documented, including one case from Hanoi with paperwork to prove it.

What is the cobra effect?

The cobra effect is a perverse incentive: a rule or reward that makes people behave in ways that worsen the original problem. It belongs to a wider family the sociologist Robert Merton described in 1936 as the "unanticipated consequences of purposive social action."

At its root it's a failure of second-order thinking. First-order thinking asks what the fix will do: pay for dead cobras, get fewer cobras. Second-order thinking asks how the people in the system will respond: if dead cobras are worth money, someone will start producing them.

The cobra effect in the real world

The Hanoi rat bounty, 1902

The historian Michael Vann reconstructed this one from French colonial archives. Hanoi's new sewers had become a highway for rats, and with plague spreading in the region the authorities paid a bounty per rat. To save officials handling carcasses, hunters only had to hand in a tail.

Soon there were reports of tailless rats in the city. Hunters were cutting off tails and releasing the rats to breed more. Officials also found people farming rats outside the city for the bounty. The scheme was dropped.

Wells Fargo's sales targets

In 2016, US regulators fined Wells Fargo after finding that employees, under pressure from aggressive cross-selling targets, had opened accounts customers hadn't asked for. A later review put the number of potentially unauthorised accounts at about 3.5 million. The targets were meant to measure customer relationships. Employees, facing the threat of being fired for missing them, met the targets instead.

Mexico City's driving ban

In 1989, to cut air pollution, Mexico City introduced Hoy No Circula, which barred each car from the roads one weekday based on its number plate. The economist Lucas Davis studied air-quality data from before and after and found no measurable improvement. One likely reason: many households bought a second car to drive on their banned day, often an older and more polluting one.

Test scores in Atlanta

When school funding and staff careers became tied to standardised test results, some Atlanta teachers and administrators began altering students' answer sheets. A 2011 state investigation found cheating at 44 schools. The scores went up; what they were supposed to measure didn't.

Your own goals

The same thing happens at personal scale. Set a goal of fifty books a year and you may find yourself skipping long, difficult books for short ones that keep the count moving. The number goes up while the reason you set it, reading things that change how you think, quietly goes down.

You may have heard a version about a software company that paid developers for every bug they fixed, so they wrote bugs to fix. That one comes from a 1995 Dilbert strip, not a documented company. It's funny because it's plausible.

Why the cobra effect keeps happening

Two well-known laws describe it.

Goodhart's law. The economist Charles Goodhart observed in 1975 that a statistical regularity tends to collapse once it's used for control. The anthropologist Marilyn Strathern later put it in the form most people know: "When a measure becomes a target, it ceases to be a good measure."

Campbell's law. The social scientist Donald Campbell wrote in 1979 that the more a number is used to make social decisions, the more it will be corrupted, and the more it will distort the process it was meant to watch.

Both point at the same mistake. We design fixes as if a system were a machine that does what it's told. But a system with people in it adapts. Introduce a reward, and the people inside will find the cheapest way to earn it, which isn't always the way you meant.

How to avoid the cobra effect

Keep asking "and then what?"

Before you introduce a rule or reward, ask how a smart person with nothing to lose might game it. What's the cheapest way to hit this target without achieving the goal? If you can think of one in five minutes, so will the people it applies to. It also helps to resist false dilemmas that make one fix look like the only option.

Measure more than one thing

A single number is easy to game. Pair a target with a check on what you actually care about: call-handling time with customer satisfaction, bugs fixed with bugs reported by users, accounts opened with accounts actually used.

Try it small first

Before rolling out an incentive across a whole team or city, try it somewhere contained, where it's easy to reverse. That turns it into a two-way-door decision. Watch what people actually do, and look out for the breeders.

Watch incentives play out

The Paradox Game includes the Prisoner's Dilemma, the classic case where each person's sensible response to the incentives leads to an outcome nobody wanted. Playing it is a quick way to feel how a well-meant structure can push reasonable people somewhere bad.

Respect the system

The cobra effect isn't an argument against trying to fix things. It's an argument for humility about how fixes land. Before you change a rule at work, at home or in your own habits, ask what it rewards. Make sure you aren't paying anyone, yourself included, to breed the snakes.

Frequently Asked Questions

What is the cobra effect in simple terms?

It's when a solution creates an incentive that makes the original problem worse. The name comes from a story about a cobra bounty in colonial Delhi that led people to breed cobras.

Is the Delhi cobra story true?

Probably not as told. Historians haven't found records confirming it. The very similar Hanoi rat bounty of 1902 is documented in French colonial archives, and modern cases like Wells Fargo's sales targets are well established.

What is an example of the cobra effect in business?

Wells Fargo is a well-documented one: aggressive cross-selling targets led employees to open millions of accounts customers hadn't requested. The targets were met while customer trust collapsed.

How do you prevent the cobra effect?

Ask how people could game a rule before you introduce it, measure the real goal alongside the target, and try new incentives on a small scale first, where they're easy to reverse.

Sources

1. Siebert, H. (2001). Der Kobra-Effekt: Wie man Irrwege der Wirtschaftspolitik vermeidet. Deutsche Verlags-Anstalt.
2. Vann, M. G. (2003). Of rats, rice, and race: The Great Hanoi Rat Massacre, an episode in French colonial history. French Colonial History, 4, 191–203. https://doi.org/10.1353/fch.2003.0027
3. Davis, L. W. (2008). The effect of driving restrictions on air quality in Mexico City. Journal of Political Economy, 116(1), 38–81. https://doi.org/10.1086/529398
4. Consumer Financial Protection Bureau. (2016). Consent Order: Wells Fargo Bank, N.A. (File No. 2016-CFPB-0015).
5. Merton, R. K. (1936). The unanticipated consequences of purposive social action. American Sociological Review, 1(6), 894–904. https://doi.org/10.2307/2084615
6. Goodhart, C. A. E. (1975). Problems of monetary management: The UK experience. Papers in Monetary Economics, 1. Reserve Bank of Australia.
7. Strathern, M. (1997). "Improving ratings": Audit in the British university system. European Review, 5(3), 305–321.
8. Campbell, D. T. (1979). Assessing the impact of planned social change. Evaluation and Program Planning, 2(1), 67–90. https://doi.org/10.1016/0149-7189(79)90048-X