Worldbuilding 205: Perverse Incentives
Greetings and sexton! ...Get your mind out of the gutter; it has to do with things far more holy. So, I hope all are doing well. Despite the name of this blogpost, it will have nothing perverse in of itself; we do SFW stuff on this blog! It is about perverse incentives, and what is that? …You know by now that the definition was a breath away.
Definition
I would define perverse incentives as follows:
Perverse incentives are incentives that are meant to achieve a specific goal, but due to their structure, end up working against the desired goal.
Some would lessen this to just be “unexpected results”, but I feel in a lot of the cases the “unexpected result” is, in fact, actively against the desired goal.
Intentional vs accidental
A question that can arise when discussing perverse incentives–I am getting quite tired of repeating that phrase, I feel icky–is if they can be intentional, and I would argue that is pure nonsense.
You set up incentives for a goal, a goal you have in mind and want to be achieved. Perverse incentives happen when the incentives cause people to do things that go contrary to the goal you desire. So, by this, perverse incentives cannot ever be intentional. After all, if the goal is achieving what you get… the incentives work, and incentives that work cannot work against you at the same time. Now, the results you get might be against the stated goals, but we’re talking about when the results go against the genuine, sometimes hidden goals.
Proximal measurements
The question naturally becomes, then, where do perverse incentives come from? Well, there is a common thread in a lot of perverse incentives, namely, what I call proximal measurements. The world is big, it is difficult, and when you want to achieve a goal, you need to measure how successful what you do is.
But not everything is easily measurable in a direct manner. You cannot measure how much knowledge a person has in a direct way. You don’t have a fancy mind-reading device that can tell you if they know the multiplication table. So, in a lot of cases, you have to take something else and use it as a proxy for what you want. If you pick good things to measure that are directly connected to what you actually want, there will be no problems!
But we wouldn’t be here if those were easy to find. They are infamously difficult to find in a lot of cases. So one has to use proxies instead, and to continue the example of studying, what proxy can be used for it? Tests are the classical method of measuring the knowledge a student has acquired. Are they good?
As a teacher, I can say… No, they are quite bad. My 🇺🇸 friends can probably tell quite loudly that tests have a lot of problems; should we run through the main ones real fast?
Cheating
Stress decreases viability
Memorisation of answers is not knowledge
These are not all, but the most prominent ones, and as for the rest? I don’t remember them, so don’t ask. This isn’t a test. But on a test, you can cheat, and we are not trying to measure your skill at deceiving people and smuggling in help. In fact, cheating can be seen as a consequence of perverse incentives. The incentive is to have good grades, so you want the student to study hard to get good grades and achieve in life, but that also incentivises cheating in order to get ahead in life.
I bet many of my beloved readers have experienced test stress: the stress of taking a test and being under time pressure to perform. Again, what is it we are trying to measure? We are trying to measure how well you have retained and understood knowledge, not your ability to withstand stress and pressure. But a huge component of taking a test is to endure this pressure and stress. This can be so bad that many people would rather skip a test and not fail (nor succeed) than have to endure the immense crushing pressure during the test.
And the last bit might appear to serve our purpose, but does it? Let me ask you a question: if a student memorises everything 10 minutes before the test, and then 10 minutes after, no longer remembers anything of it, have they really acquired any knowledge? I am sure we all can agree that it is not learning knowledge or thinking about it, it is blind memorisation without thinking, and then forgetting. Fun fact, in 🇸🇪 we call just stuffing in knowledge to not think about and such “sausage stuffing.” This is, in fact, also one of the results of perverse incentives where school funding is tied to student performance.
So even the most obvious ways of measuring can measure the wrong things, and that can, in turn, result in perverse incentives becoming real.
Historical Examples
A few examples help get the point across, right?
One anecdote is that of the British Raj, where the British wanted to get rid of the cobras–let’s just pretend it was entirely for good purposes, because we know it was almost certainly not for the benefit of the locals. So, how would one measure the decrease in cobras? Well, you can’t go around counting them because if you count them, you have to capture them, and counting all cobras is a giant pain.
So, they had the magnificent idea of counting the number of cobra heads they got in. A cobra head means a dead cobra, so the number of cobras has to go down, then, right? Well, that is true, as long as the increase in population size by new ones being hatched is less than the number of heads being collected every month. Would it surprise anyone that this worked for a short time, and then the numbers shifted?
Hindsight is 20/20, but you would think that people could see an issue this obvious. People were paid for each cobra head they handed in in order to incentivise them to kill cobras on sight. Of course, then people started breeding cobras. Why bother chasing cobras when you can just breed a large quantity of them? Chasing is dangerous, but breeding is safer and gives lots of money! And when the British caught onto this and cancelled everything… the cobras were released, and the problem was amplified instead, and the British were poorer…. Poor colonisers.
Another example happened in French Indonesia, today’s Vietnam. They wanted to get rid of the pest of rats, so tails served the same function as cobra heads. Except people caught rats, chopped off their tails, handed them in, got money, then released the rats into the sewer to breed new rats they could catch. The French there noticed that the rat population was not decreasing… but the tailless rat population was rapidly increasing. I wonder why…
A more hilarious example was in the UK in 2017. They have laws that prevent old buildings from being changed too much, and there was an old building up for being put onto the list of protected buildings, which meant the owner couldn’t do much to change it without a lot of work and paperwork. Real pain. So, the day right before the inspectors and more would come to decide if it was put onto the list, the owner destroyed the ceiling in order to prevent it from getting onto the list. That building was 400 years old. Whoops.
Another was from the Soviet Union, where speed and quantity of production were encouraged and incentivised. You had to do 40 train wagons of lumber? You stacked the trees inside such that it looked overfull when the inspector came, and the moment the train moved, it collapsed and was nowhere near enough.
So yeah, this has been done a lot in history… Some are more obvious than others about ending badly.
Worldbuilding purposes
Now, why would you ever want these in your world? Because it happens in real life, and it would naturally happen in your world too! Duh! That is why you are here 😀
In fact, a lot of the things you complain about in life can be derived from perverse incentives. A lot of problems in the 🇺🇸 “Healthcare” system exist entirely because it has perverse incentives. And I don’t mean the fact that it all costs your arms, legs, and gonads; that is by design.
So, in your world, you will likely have a lot of bad things happening socially because of perverse incentives. The question is, however, where? Well, anywhere you have bureaucracy or proximity measurements. Let’s say you are building a kingdom that has just discovered “new land” that definitely doesn't have virtually defenceless inhabitants in it due to technological discrepancies… Anyway, they want to colonise and get people over there and, for reasons, forcing them is not an option; they are not COMPLETELY evil… Just slightly.
So they decide to set up a system where plots of land are given to anyone willing to sail there, and to make it super official, they all get a paper that says it from the queen. Fancy stuff. The captain gets to be in charge of it and hand it out at the end of the voyage. After all, you don’t want it to go to people who are dead by the end of the journey!
Can anyone see the problem that can arise? Well, what happens if you sail out… and then throw all the people overboard? You get to keep the deeds to the land! Which means you can sail off somewhere, camp out for a reasonable amount of time, come back to the docks, sell all the deeds to people, pocket all the money, and you can repeat. What are morals, after all, for money?
My people are good!
Some people really want their people to be good… and honestly, I think you are very boring then. People can, in broad strokes, be good, but people will always be bastard coated bastards with bastard filling. Of course, some will not do cruel things no matter what, but let me ask you this: How much are your morals worth if your loved ones have to die for you to keep them? Probably not too much, and my point here is that everyone can succumb to these if presented.
How many parents have helped their children cheat? How many people follow rules by the letter, not the spirit? And so on. So, avoiding perverse incentives is difficult, as people WILL abuse it, and perverse incentives are usually when it is so blatant and bad that people are instinctively drawn to abuse it.
To avoid perverse incentives, you need to have the incentives of people and the incentives that are provided align. If people don’t mind the rats, the French Indonesian example will not work; the rats are just free money then. But I would say even if you aligned the incentives, then if you incentivise the wrong thing, this problem will keep coming, and that is often primarily exactly because the proximal measurement is giving the wrong impression, and people realise that they can exploit it.
If you really, really, REALLY want to avoid perverse incentives, the best tip I can give in a system is that there is not ONE proximal measurement; there are several that are used together. That is because it is much harder for people to figure out how to fool all of them at the same time.
Long-term effects
While most perverse incentives are rather easy to stop once you realise you have one on your hands–you simply stop doing the incentive–the ramifications can linger a long, long time. In the case of the British Raj anecdote, we have the cobra population skyrocketing the moment it stops, as an example, which means the initial problem that was supposed to be solved is now significantly worse.
In my made-up colonisation example, we have, of course, murder happening at the worst of cases, in the best of cases, people sold to slavery if said captains were a little nicer. But that is sometimes the least of the problems. The main damage that can happen is trust. Well, TO trust. The government and those in charge are meant to appear competent; after all, if they are not, why do they rule?
If the problem is exacerbated because of the ruler's incompetence to predict something, ESPECIALLY if the outcome is considered blatantly obvious by the people, it is a stain on their ability to rule. They can become jokes that everyone talks about and next time a new solution comes, it is not heeded as everyone is suspicious of how this will end up badly.
You can always put yourself into those people in the British Raj. Imagine spending hours or days dealing with and working with cobras. Even if captive, it is NOT risk-free! And then you walk to collect your money, and the British tell you to sod off. Sure, they are British, so you’d not like it to begin with, but you were promised money that you now don’t get! You are very mad, SMASH!
So, as much as they easily come about, perverse incentives can undermine your authority heavily if they put on display your incompetence.
Summa Summarum
Perverse incentives are almost inevitable in any society because proximal measurements are often the best we can ever get on many things we deem important. And where proximal measurements appear and we apply incentives, they can and will become perverse one way or another.
On top of this, perverse incentives often explain many issues that exist in a society, but hardly all, mind you. So if you have a problem in your society, imagine what perverse incentive might be causing it to exist.
That will be all for today. Have a great weekend!
Want to dive into a discussion about Stellima or the art of writing on Discord? We’d love to have you! And if you have any topics you struggle with or that you would like to suggest for a future blogpost, we’re open to suggestions!
Interested in supporting our work? Join our Patreon and become a part of Stellima as a citizen of Mjatreonn! Or would you like to give us some caffeine to fuel our writing? Consider buying us a coffee at Ko-fi! Every contribution inspires our creativity and keeps us going. Thank you for your support!
Copyright ©️ 2025 Vivian Sayan. Original ideas belong to the respective authors. Generic concepts such as perverse incentives and the historical examples are copyrighted under Creative Commons with attribution, and any derivatives must also be Creative Commons. However, specific language or exact phrasing are individually copyrighted by the respective authors. Contact them for information on usage and questions if uncertain what falls under Creative Commons. We’re almost always happy to give permission. Please contact the authors through this website’s contact page.
We at Stellima value human creativity but are exploring ways AI can be ethically used. Please read our policy on AI and know that every word in the blog is written and edited by humans or aliens.
