![](https://blog.lidd.ca/hubfs/Imported_Blog_Media/20140916-Cobra-Effect-1.jpg)
The guys from Freakonomics did a podcast on the Cobra Effect which explains the myriad ways in which folks engineer incentives to create one outcome and end up with an entirely different outcome.
The “cobra” in Cobra effect refers to a story where a British governor thought up a clever scheme to rid Delhi of its plague of cobras: he offered payment for every cobra skin brought in to municipal offices. Months later, his offices had paid out a small fortune for cobra skins but the population of snakes did not diminish. In fact, Delhi citizens complained that they were dealing with more snakes than ever. The governor sent investigators out to see what was going on. On the outskirts of Delhi, they discovered a booming cobra snake farming industry.
The incentive did nothing to curb urban snakes, it turned snake breeding into a lucrative field. When the governor learned that his incentive failed, he cancelled the skin reward. The snake farmers on the outskirts of town soon had worthless snake herds. So they let them free and Delhi’s cobra problem tripled overnight.
The Cobra effect refers to deliberate incentives that go horribly awry, whether that’s rats in Hanoi or pigs at Fort Benning. Less dramatic but far more routine, human organizations suffer from incentives that no one deliberately created. Everything is an incentive and bad behaviors are almost always traceable to an inadvertent incentive. This applies to supply chains as much as anything else.
For example, consider a food distributor that receives about 200 trailers of perishable product every day. Their receiving overtime was out-of-control, their shorts on product that should have been in house were terrible and the highest levels of the company wanted to understand what the problem was.
We looked at three months of data. For each inbound load, we had the timestamp of its arrival, its receipt, the end of its receipt and its departure from the site. We also had its appointment time. The first crunch of data set the problem out, of all inbound loads:
- 40% showed up early for their appointment
- 20% showed up on-time
- 40% showed up late
Clearly, no one was respecting the schedule. But why? Our second data crunch provided the answers:
- When a truck showed up early, it had a 66% chance of being received early
- When a truck showed up on-time, it had a 50% chance of being received late
- When a truck showed up late, its average wait time to be received was the same as for those trucks that showed up on time
The inbound crew’s behavior was creating an incentive for carriers to disrespect the appointment schedule. If a truck could be early, then there was an incentive for being early. But there was no incentive to be on-time nor was there an incentive to avoid being late. The warehouse was creating its own problems and no one appreciated this until the numbers sat on a page before their eyes.
A wise man once said to me, “when people are behaving wonky in an organization, look at how they are compensated, that’s the key to their behavior”. Everything is an incentive. If your workforce seems to be working against a system, policy or best practice, there is almost certainly a hidden incentive making that so.