In a meritocracy, the smarter you are, the dumber you feel
And why Doug Burgum thinks he can run for president
Life in a meritocracy
There’s a “former gifted kid” meme about people who feel like they were misdiagnosed as unusually smart children. They tend to blame the education system for overusing terms like gifted.
Well, I’m here to tell you not to blame the education system1. The real criminal is our old friend, normal distributions2.
It’s like this: Over the general population, ability (when measured correctly) is distributed roughly in a normal distribution, whose PDF looks like exp(-x^2)3. But you don’t see the general distribution: You’re at a specific workplace, university, or some other kind of environment that selects for talent. You’re inside on of those trapezoidal sections on the right, in an environment filtered for your ability. If you’re an academic at a third-tier university, you’re in one of the trapezoids a bit closer to the middle. If you’re a professor at MIT, your trapezoid is all the way at the end.
Now zoom in on one of these trapezoids. What does it look like locally? Well, locally the PDF function is just a straight line, with the relative slope4 being:
that is, it gets steeper the further out you are. Which means the further out you are, the more skewed the distribution is, which means most people cluster along the left edge of the trapezoid. From your point of view, the smarter you are, the more likely you are to be in a situation where you’re tied with a lot of other people for being mediocre, while looking at a few outlier stars that make you feel bad in comparison5. If you’re at a third-tier university, on average, there’ll be some people who are worse than you and some people who are better. If you’re at MIT, you’re tied for worst with the other 99% of professors who don’t have a Nobel prize or a Fields medal.
A few caveats
First, note that this depends on the slots being set up by a society that’s at least reasonably meritocratic. If you insert yourself into one of a fixed range of trapezoidal slots, you’re probably near the left end of it. If you were making your own slot, you’d probably center it around yourself, which would mean you’d feel smarter the smarter you are - if you’re in the middle of a skewed trapezoid, you’re at a higher percentile of your local group if it’s more skewed.
Second, why are we always on the right tail? It’d be nice if we were on the left (and always tied for being the best around). But we measure ourselves by what we’re good at. I’m a pretty slow typist6, but I mostly just don’t especially care about it. I’d feel bad about myself if I was a court stenographer or something that required fast typing, but, well, it’s not a coincidence that I’m not a professional stenographer.
And importantly, this gets worse the more specialized society bcomes. In a society where 90% of people are farmers, even a perfect meritocratic sorting would result in 40% of people being on the left end of the bell curve. In a society where everyone is hyperspecialized at something they’re uniquely talented at, everyone might be in the top 1% of the thing they measure themselves by, which means they’ll all feel relatively bad about themselves.
A second model
There’s another way to model this: Instead of assuming hard cutoffs, assume that people get sorted into buckets based on a proxy of their ability. Sticking to standard jobs - a job interview uses an imperfect signal (the ratings of the interviewers) to test if a candidate is above a bar. But the higher the bar is, the more likely people who pass are false positives: With a very high bar, there’s a large cluster of people whose true ability is a bit to the left of it (who might pass if they get lucky in their interview), while false negatives are less likely (since there’s a lot fewer people just to the right of the bar).
The nice thing about this is that the resulting distribution looks less like a trapezoid and more like a normal distribution (so you’re more likely to have people around who you’re clearly better than, who just got very lucky in their interview). The minus side is that this means most people who pass the interview don’t actually meet the bar. This is one reason why imposter syndrome is so common at top companies and universities: Most people there really did get in mostly through luck.
Non-steady states
So far we’ve been assuming a steady-state situation, where people just find their natural place and stick to it. But this isn’t how things work, in practice it takes decades to reach your limits. If you’re an average postdoc at MIT, you were probably one of the best in your year at grad school, a runaway star in college, and smarter than all your teachers in high school. You have years or decades of experience at being the smartest guy around, and now you’re suddenly below average, and it feels weird.
So why is Doug Burgum7 running for president, despite being a complete unknown in a crowded field with a clear frontrunner? Well, look at it from his perspective. He’s a small town guy who somehow started his own billion dollar software company, then turned around to run for governor as a complete unknown and somehow won in a landslide. From Doug Burgum’s perspective, winning impossible competitions that everyone tells him he’s an idiot for even trying is just a normal Tuesday. He’ll probably be genuinely surprised and disappointed when he flames out and loses the primary, because this is his first time hitting his limit. And most of us, no matter how talented, hit this limit earlier and more repeatedly.
Solutions?
This is a similar problem to Scott’s parable of the talents, except the solution is reversed: Scott resolved to not be down on himself about the things he’s bad at in return for some humility about the things he’s good at. But Scott genuinely is in the rare position of being best in class at the things he’s good at8. The average famous successful blogger isn’t Scott, it’s a talented quotable writer who feels bad about not being Scott.
I don’t really have a great solution here, aside from “try to center your trapezoid on where you actually are, and not the slot imposed by social structure”. Compare yourself to your friends more than to your coworkers. It’s better to focus on doing what you can do with a reasonable amount of real effort, but not worry about being the best, and to solve the problems you actually run into and are interested in instead of the ones you feel like you should be able to do. Do be ambitious and try harder things when you can, but because they’re interesting to try, not because you want to be better than the people around you. Some red queen races are unavoidable, but avoid the ones you can.
For this specific thing. You can still blame them for, well, everything else.
This is usually the case.
Yes, I forgot a bunch of constants. They don’t matter here.
We want to look at relative and not absolute slope: You compare yourself to a fixed number of people in your environment independently of its size, so you need to normalize by the total size to get your relative position within it.
And since the stars are likely to be better known than average, this is skewed even further than the model would suggest.
It was remarkably hard for me to find an example of something I’m measurably worse than average at. This isn’t because I’m supertalented at everything, it’s because we almost never think about the things we’re unusually good at - noticing things you’re just kinda lousy at feels unnatural!
In case you don’t know who that is: Doug Burgum is the governor of North Dakota. At the time of writing (September 2023), he is running in the Republican presidential primary despite polling around 1% and no one outside of North Dakota having ever heard of him. If you’re reading this in the distant future where he’s somehow actually become president, boy do I feel silly.
Maybe not? Maybe Scott compares himself to the great writers of history and beats himself up over not being as cool and quotable as Oscar Wilde? I dunno, in I assume Scott actually is at the top of whatever trapezoidal reference frame he puts himself in.