Thoughts on how we do incentives poorly and don’t understand second-order effects.

Subscribe: Apple Podcasts / Overcast / Spotify / Stitcher


Hello, I’m Nick, welcome back.

One of the things I’ve been thinking about recently is incentives. I’m got a 3-year old…and anyone who has kids knows the joys of teaching them consequences and how to motivate them to do the right thing.

We been looking at public and private pre-K schools around us to see which ones are quote unquote good. Of course, you read the ratings and reviews, talk to other parents, and you try to learn why or why they don’t like their kids schools. Eventually the conversation gets to how older kids have to go through testing and whether students really learn the material or learn to take the test.

In theory, tests can be a good thing, right? We are trying to make sure all students learn the basics, a bit of math, reading comprehension, whatever. But we run into problems with 2nd order effects where in some places, we end up incentivizing teachers and school districts to teach the test, so the students get better scores, and maybe schools get a bit more funding, and maybe some folks get little pay bumps.

These 2nd order effects happen…a lot. There are a bunch of names for it too. One is Good Heart’s Law, which says: When a measure becomes a target, it ceases to be a good measure.

Another one is the cobra effect. When the British ruled India, there were lots of cobras all over the place. Pretty dangerous, right? So the British put out a bounty for dead cobras. What did people do? They started breeding cobras to kill them and get the rewards. When the British figured this out, they ditched the reward system, which meant the cobra farmers freed all the snakes which made the problem even worse.

Similarly, we have the rat effect. In Vietnam, under the French…they had a rat problem. The French created a bounty program for each rat tail people turned in. So folks would chop off the tails of rats, but let them go and breed, so more rats would be born and more rats tails could be chopped off, and on and on.

We have this problem in the business world too. Being innovative is all the rage. But the incentives aren’t there. Most CEOs are only going to stick around for a few years. Their pay structure is tied to short term goals like their stock price. It’s much easier to cut costs than actually try being innovative which will be a longer process, probably require re-thinking business models, maybe re-structuring the organization. The idea of breaking down silos in the process of innovation sounds simple, but people don’t like change, people tend to be territorial. It always feels like the change is taking power or safety. Plus you have the problem of large companies having to report their financials every 3 months. When your neck is on the line every 3 months, why worry about top-line growth when its easier to try to simply cut costs?

You see these 2nd order effects happening in digital advertising all the time. Brands pay for awareness, get their name out there. We end up with ad tech that puts up banner ads in low-cost, inappropriate places, running bots to drive views and clicks. We end up with the situation where Google says an ad has been viewed if 50% of it is on the screen…for 1 second. Come on, that’s not view. We talk about how personalization is the holy grail, and that concept has become so ingrained in all of us, that it’s perfectly ok that it takes 8 clicks to turn off Google location tracking…ON a friggin’ Android phone.

Back to companies. We all love talking about how culture beats strategy. If we have the right culture in place, we will have motivated and engaged employees, folks who will put in the extra effort, be collaborative, think outside the box. This is all funny because we have essentially widgetized people. Everyone is a replaceable cog in the machine. We don’t hire like we want innovative people. HR usually has a list of skills to hire for…with abstract measures like… “oh, are you good team player?” But then we hire and fire based on how many support calls employees handled in an hour, how much did they upsell to a customer. These incentives are designed for executives and managers to maximize their benefits by being in charge instead of taking care of people in their charge.

So be careful about incentives. We all need to think about What are the possible consequences of the incentives we are thinking about. What are the possible consequences of those consequences? How do we minimize the probability and cost of unintended consequences? Because those unintended consequences will happen. If it might be abused, it will be abused. If you are trying to evaluate incentives from the outside looking in, there are even more questions to ask. Do the people setting the incentives only benefit from them or do they feel some of the pain if something goes wrong? How will we know if the incentives are failing? Sometimes there is a time delay between putting incentives in place and the consequences. Or are the incentives being put in place for the sake of doing something and looking busy without actually understanding the real problem?

Alright, that’s all I’ve for today. Wish me luck with my 3-year old.