Dunning-Kruger, Idiocracy, and the Kid Who Rebels

Confident people must be competent. Dumb parents must raise dumb kids. Strict parents must raise obedient ones. Three unrelated ideas, one shared assumption, and it falls apart every single time you check it against an actual family.

I was talking with a friend about Dunning-Kruger the other day, half-joking that I'm probably a walking case study, and the conversation wandered three topics away from where it started. By the end we'd covered a psychology paper, a Mike Judge movie, and why the strictest parents we both knew somehow raised the wildest kids. None of it seemed related at the time. It is.

The Unskilled and the Unaware

Quick recap for anyone who's only heard the meme version. In 1999, Justin Kruger and David Dunning ran four studies at Cornell testing people on humor, grammar, and logic, then asked each person to guess how they'd ranked against everyone else. The paper, published in the Journal of Personality and Social Psychology under the title "Unskilled and Unaware of It," found a specific, lopsided pattern: people who scored in the bottom quartile overestimated themselves the most, guessing they'd landed around the 60th percentile when they'd actually landed closer to the 12th. People who scored in the top quartile did the opposite. They slightly underestimated themselves, assuming the test had been just as easy for everyone else.

The least qualified people are often the most confident. The most qualified are still checking their work.

The mechanism Kruger and Dunning proposed is the part that made it famous: the same incompetence that produces wrong answers also removes the ability to recognize wrong answers. You need to understand grammar to know you're bad at grammar. If you don't have that baseline, you have no ruler to measure yourself against, so you default to average or better.

The honest caveat: the popular version of the graph, the one where unskilled confidence swoops downward across the whole competence range, has taken real hits from statisticians. Edward Nuhfer and colleagues published work in Numeracy showing that a lot of the classic Dunning-Kruger shape reappears even in randomly generated data, because comparing a bounded self-estimate against a wide-ranging test score produces regression to the mean whether or not any real self-assessment bias exists. The effect isn't fake. It's smaller and messier than the meme, and some of what looks like "dumb people are overconfident" is just statistics doing what statistics does with noisy, bounded numbers.

Idiocracy's Math Problem

That callout matters for the second thing we talked about: Idiocracy. Great movie, genuinely funny, and I've had a specific gripe with its premise since it came out in 2006. The setup is that smart people keep having fewer kids while less-educated people keep having more, so average intelligence ratchets downward generation after generation until the country is irrigating crops with a sports drink. The fertility gap by education level is real and well documented. My problem is what the movie needs that gap to mean: that a kid born to an "unintelligent" parent comes out unintelligent too, like a direct copy.

That's not how heritable traits actually distribute, and we've known why since 1886. Francis Galton, studying the heights of parents and their adult children, found that unusually tall parents had children who were, on average, shorter than they were, and unusually short parents had children who were taller. Everyone drifted back toward the population mean. He called the paper "Regression Towards Mediocrity in Hereditary Stature," and the phrase stuck around long after anyone remembered it started as a height study. He'd accidentally invented regression to the mean, and it applies to cognitive traits the same way it applies to height: extreme parents tend to produce kids closer to average than themselves, not clones.

Intelligence is genuinely heritable, twin studies put adult heritability estimates somewhere in the 50 to 80 percent range depending on the study, but heritable and deterministic aren't the same word. Genius kids get born to unremarkable parents constantly. Kids from resourced, well-educated families turn out entitled and clueless constantly too. Idiocracy needed intelligence to be cleanly transmitted and permanent for the joke to land, and biology just doesn't cooperate with that setup.

Reactance, or Why the Preacher's Kid Runs Wild

The third thread came from something we'd both noticed anecdotally: kids raised under rigid control often become the opposite of their parents, and kids raised in chaos often become obsessive about order. I'd chalked this up to coincidence until my friend mentioned psychological reactance, a theory the psychologist Jack Brehm laid out in 1966. Brehm's claim was simple: when people perceive that a specific freedom is being restricted or taken away, they become motivated to reassert it, sometimes by doing exactly the restricted thing more than they otherwise would have.

Enforce a rule harder, and you don't get more compliance forever. You get compliance while the enforcement is watching and a growing urge to break it once it isn't. That urge tends to peak exactly when someone starts building an identity separate from their parents, which is most of adolescence and early adulthood. And a lot of identity-building at that age is just "be the opposite of the thing standing in front of me." It's the cheapest available way to feel like your own person without inventing yourself from nothing.

The One Assumption All Three Get Wrong

None of these three things are about the same subject. One is about self-assessment, one is about heritability and demographics, one is about parenting and rebellion. But every popular misreading of all three makes the identical mistake: assuming a trait transfers cleanly from cause to effect with no noise in between. Confident people must be competent. Parents' traits must become the kids' traits. Enforced rules must produce rule followers.

Reality runs noisier than that on all three counts. There's a wide spread of outcomes and plenty of exceptions, and the exceptions aren't even rare. They're just less interesting to build a movie, a meme, or a parenting theory around than the clean version would be.

Peter Benoit

Written by

Peter Benoit

UI developer and federal design-systems engineer in Florida. I build accessible interfaces for federal health infrastructure and the open web.

Keep reading

All writing