Incentives Matter – people change their behavior in response to carrots and sticks, regardless of what the designer of the incentives hopes for.
-Captain Obvious
Incentives, those little plusses or minuses that get people to do things, or not do them. It makes perfect sense: if you want someone to do something, you reward them for doing it, or threaten to punish them for not doing it. If you want your car washed, your kids might just go ahead and do it spontaneously, but if you offer them 20$ they will suddenly much more interested (more so than if you offered just the proverbial carrot). Then again, you could say ‘Wash the car or you are grounded.” You might get two very different results from the two methods, however; there is no reason to assume that the use of positive or negative incentives is going to be symmetrical in effect.
Likewise, Nature offers her own incentives. Plants produce sweet, nutritious fruit to encourage animals to distribute their seeds, hopefully with a nice little dose of fertilizer. Insects produce nasty tasting toxins to encourage birds to find something else to eat. Run around outside and the exercise will make you feel a little better, and your skin will start making vitamin D. Spend your time outside hitting hornet’s nests with a stick and chances are you will decide not to do that again, unless you are really good at the running part.
We all intuitively know the basic idea of incentives, but it is worth looking at a few different aspects because we are apt to forget some very important points.
What Motivates Humans?
Probably some other little wet bits in there too, but at some very basic level there is at least one chemical in our brains that motivates us by making us feel good, some sort of happy juice, and we like it when our brain squirts some out. The exact nature of how all that works need not concern us much here; the really interesting bits happen a cataract or two downstream.
The first thing to note is that consumption of three primary things trigger the happy juice: sugars, fats and sex. What? Yea, that seems a bit reductionist, but note that we definitely pursue those more than logic might suggest is prudent. Modern humans would probably be healthier if those were easier to moderate. Premodern humans needed to jump all over those opportunities, however.
Sweet foods contain lots of calories, and if a plant bothered to make its fruit sweet it is a really good sign that it isn’t poisonous, and probably has a bunch of other nutrients, too.
Fats similarly are energy dense and have lots of nutrients; many important vitamins are fat soluble, whether from animals or seeds. Fatty meat also suggests that the animal it came from was healthy, as fat is a luxury many animals can ill afford, and usually quickly consumed when ill.
Sex is pretty obvious: animals that don’t have an interest in reproducing aren’t going to make it far in the evolutionary race. If you don’t reproduce by budding or something, sex is the first step, so a species that is still a going concern is going to have a lot of concern around it.
But… that doesn’t get us very far with humans, does it? Sure, rats, fine, they are pretty well pleased with just those three essentials. Some peanut butter, some girl rats, and Rizzo is pretty well set. Humans do lots more things1, things that don’t even involve eating or reproducing. Hell, sometimes we even choose to do things that specifically remove sex as an option. If everyone was satisfied with a tub of ice cream and a sex partner, we’d almost certainly live in that world already and no one would be writing or reading this. Surely we have some more elevated set of reasons for doing things.
Well… at root, no. It is dopamine all the way down. But we do have a special trick that allows us to hijack our basic rat brain systems and chase after much loftier goals. That trick is…
We don’t care about the real world.
Humans, and to a much lesser extent most of the higher order animals, don’t live in the real, physical world, but rather inside our own brains in a world of abstractions and dreams. We only tangentially touch on reality, for the most part living entirely within our own heads. One of our superpowers is basically being insane.
That is a little bit of a harsh way of putting it perhaps. A nicer way of thinking about it is that, while our brains really want to reward us for making bacon and slamming hams, evolution has also taught it to reward performing actions that are likely to lead to getting sugar, fat or sex. Sometimes it rewards us even more for potential good things than the good things themselves… insert your own experience of being really excited for something and then finding out “Eh… it was only pretty good, I guess,” despite being exactly what you expected.
This pre-goal reward system spirals quickly into madness, layers upon layers of it. What might start as “Eating peaches is great!” leads to “Finding peaches is great!” and “Climbing trees to get peaches is fun!”, which leads to “Owning peach trees and keeping other people away is satisfying,” along with “Climbing trees is satisfying.” Thousands of years later you get kids who have never picked fruit from a tree in their life climbing playground equipment made of steel because it is fun, for some reason.2 Now, madness implies bad, so maybe it isn’t the best word, because it isn’t always bad here. Getting a nice big hit of happy juice when you start a new project as opposed to only when you finish it and reap the reward probably means a lot more projects get started than would otherwise, and therefore a lot more projects get completed than would otherwise. Then again, it also means that people spend a lot more time, energy and money on projects that never will finish, and probably never should have been started.
This also means that we humans are quite happy to chase things that do not directly result in the big three, but that sort of indirectly imply them or might lead to them. Making progress towards any goal, especially meaningless ones in video games, gives us a little shot of happy. Making money, or random things that act like resources for attracting mates will do it. Getting status or power within our group, something historically tied to getting sex and nicer food, still triggers the happy juice, even when the status is among grumbling internet trolls on some near dead internet fandom forum.
Maybe madness isn’t the worst word…
Anyway, long story short, we can come up with a short list of positive incentives that direct people’s behavior:
Sugar
Fat
Sex
Money
Status
Power
Intrinsic interest (other?)
Maybe we can boil that all down to:
Money/things
Sex
Status
Power
Intrinsic goals
Note that other than sex, all of it is at a higher meta level, based on internal mental states. All also tend to feed into each other (aside from intrinsic goals, maybe). If you have lots of money you tend to get power, status and sex, and vice versa. One could argue that money, status and power are only interesting because they allow you to get sex, but I would point out that people still fight over money, status and power long past the age where sex is physically functional, and plenty of people have forgone (probably) sex but chased status and power (think politics within monastic orders).
We could also argue whether freedom or liberty falls under intrinsic goals. What about children? Safety? Shelter? How do we decide whether a nice house is mostly about status or sex or something else? This is possibly all beside the point and categorization is a fool’s errand here. If you want to add a few things to the list, that’s probably fine, but I expect taking any away is going to leave an important lacuna in your thinking.
What was I talking about again? Oh, right!
So those are the things people care about getting, our carrots. What sticks do they want to avoid? Well, losing those things. Also, pain… we really don’t like that much.3 Add “people want to avoid pain” to “losing the good things”, and you are good for sticks.
Complicating things a bit, losing something does not necessarily trigger the same magnitude as gaining the thing. (Google “Loss Aversion” for one example.) Likewise, as in our car washing example above (remember that?) offering to pay your kid might result in a better job than threatening punishment for not doing it.
In fact, the varying quality of incentives is just the beginning of the complications. Timing might matter, too. Many people want to exercise and get more physically fit, but while the effort and pain is required right now, and tomorrow, and the next day, the payoffs are pretty distant. Predictably people don’t seem to like that trade much. Future or potential payoffs are a lot less desirable than more immediate rewards. Likewise for punishments: one more drink always seems like a good idea, until you wake up the next morning.
So… Why did you spend all that time telling us that people like things they like, and dislike things they don’t?
Because otherwise you can’t understand why people behave like they do.
Why do people spend lots of money on designer clothes and brand name stuff instead of wearing grey sweat suits? It would be a lot cheaper, and the whole point of clothing is to keep warm and clear of the elements, right? Oh, right, because wearing stylish clothing boosts your status. Almost the least important part of clothing’s function is what we think of as its primary function, and nearly all the money goes to making us look good and raising our status.
Why do people go to college to get low salary jobs after college while they struggle for years to pay off their loans? Why not become a truck driver or welder or other blue collar worker and make more money, faster, often with zero loans? Because those jobs pay well to offset how low status they are, and apparently not enough to fully do so considering how short on drivers and welders the US economy has been for the past twenty plus years.
Why are politicians scumbags? See items 1-3 on the second list; using their power is how they acquire those.
Why are hobbits the best ring bearers? Because they don’t care about power, and are largely indifferent to money and status once they get enough calories and (presumably) sex. Humans, not so much.
Why is “corporate culture” apparently the most important thing in determining employee behaviors? Because status, and thus power and money, within an organization is largely determined by the culture, and people will act in such a way as protects their status, and leave organizations where how they want to behave clashes with the dominant culture, regardless of the formal rules. Which rules get enforced depends on culture, after all.
Incentives matter, because while we often want people to behave in a nice fashion, what people actually do depends on their incentives, and those incentives are always present but often hard to see.
Hourly workers tend to be less efficient than workers paid piece rate, that is, workers who are paid based on what they produce. If you are paid based on time, and working harder is less pleasant, then you are more likely to relax and do less work during a given amount of time because you won’t get paid more.4 Perhaps you get around this by offering a bonus for output in addition to hourly pay, but then you have the same problems as piece work. Maybe you hire managers to keep an eye on the workers to make sure they don’t slack, but then you have to find a good manager and make sure they aren’t slacking either, moving the problem upstream a little. Maybe you hold out possibilities of promotion to the more productive workers, but then you get the problem of office politics; is the worker more productive or just more popular with the bosses?
All roads lead to what is called the “Principal/Agent Problem” which is that once you get someone to do something for you, you have a hell of a time making sure they actually do it. As grandma said, if you want something done right, you have to do it yourself. Of course that is kind of awkward when it comes to, say, abdominal surgery, so that doesn’t solve the problem. We will be addressing this issue in more detail later.5
When the incentives work out such that they encourage people to do what we (or their bosses, spouses, whatever) want them to do, we say their incentives are aligned. The incentives point in the same direction, so you work more hours you get paid more, working hard during that time means your boss likes you more so you are more likely to get promoted, resulting in more and harder work from you and more money being paid. You are happy with more money and promotions (status) and your boss is happy because you did more work for him.
If incentives are misaligned, they cause conflicts of interests between the principles and their agents. I pay someone to paint my house by the hour and they end up taking forever so as to make more money, while I have to deal with the annoyance of having painting paraphernalia all over the place. Or I pay someone a flat rate to paint my house and they rush through as quickly as possible and do a really crappy job. In both cases I need some other incentives to overcome the misalignment and get things more towards “Painters want to do a good job, and quickly” which is what I want them to do.
Incentive alignment is why we don’t generally expect people to do things out of the goodness of their hearts. People really are incentivized to be good because if nothing else they want to think of themselves as good people,6 but people are also really good at rationalizing why they are a good person while doing whatever they want, so we don’t want to rely primarily on that. Understanding incentives, and being able to really analyze which incentives are in play in a given situation, allows us to tweak our world so that we want to engage in good behavior, and avoid encouraging bad behavior.
Only you ever have all your incentives aligned towards your own best interests, and even then, future you is not the same person as current you, who isn’t the same as past you. Past you was an asshole who didn’t go for a run yesterday but instead ate a pint of ice cream and binge watched “Half in the Bag” episodes. Future you is going to be totally hard core, nose to the grindstone and get all that work you have been putting off done in no time. Current you, dying anew every instant, starts to wonder just how much future you is really going to appreciate all the hard work you are thinking of maybe doing. How much is future you really going to benefit from you running now?
I wonder if there is still ice cream in the freezer…
And to be fair, so do rats.
I suspect a similar evolutionary effect explains why people, especially children, find fire and water so fascinating. If you were an early human who didn’t like fire, chances were you were not long for this world. Likewise, bodies of water seem to have been an extremely important part of early human life, both as something to drink and a source of food. Regardless of whether the “aquatic ape” theory turns out to be true, hydrophobia was not going to be a winning strategy for early hominids.
Although that can get subverted as well…
Of course, if you are paid by the piece then you are incentivized to be a little sloppier perhaps, too. Plus your boss needs to track what you actually produce more closely. No easy solutions here.
At my current rate of writing, “later” probably is two years away… maybe see what Wikipedia says in the mean time?
One of these days I am going to do a post or ten on Adam Smith’s “Theory of Moral Sentiments,” which is probably the best book on the topic ever written. That’s a long, long project for another day, however.
Incentives are a can of worms. Look at the misalignment between the incentives for Twitter to maximise engagement while also improving the quality of discourse. And then the 'audience capture' thing. Where someone builds a following by saying interesting things about some stuff. And suddenly their hard-won audience calls the shots on what they say.
It was interesting to read this piece while aligning almost every point with how marketing works 🤔