This study uses genetic variation related to educational attainment as a quasi-experimental instrument (Mendelian randomization) together with Norwegian registry data to estimate the causal return to an additional year of schooling on labor-market earnings.
0:00Welcome to Base by Base, the paper cast that brings genomics to you wherever you are. Thanks for listening, and don't forget to follow and rate us in your podcast app. Thanks for joining us. So, from the time you are like basically old enough to carry a backpack, you are told one universal truth.
0:17Stay in school and it will pay off. Right. The ultimate conventional wisdom. Exactly. But actually proving that proving that sitting in a classroom directly causes you to earn more money later in life rather than just, you know, correlating with it. That has haunted social scientists for decades.
0:32Oh, absolutely. It is one of the most stubborn problems in economics. Why is it so hard to just look at the data? Well, I mean, observing that people with college degrees tend to make more money, that's simple arithmetic, but proving that the degree itself is the engine generating that wealth, rather than just, uh, a shiny certificate handed to people who are already destined to be successful, that requires disentangling human biology, family dynamics in sheer luck.
1:01Which is really tough to do. But today, we celebrate the work of the research team behind a massive new paper published in the proceedings of the National Academy of Sciences. Yeah, PNAS. Right. published on April 8, 2026.
1:13The poker is titled, Estimating Returns to Education using the genetic lottery. And the researchers essentially use the DNA of over 100,000 Norwegians to finally solve this. It's basically a giant natural time machine.
1:27Yeah, to put a concrete causal price tag on exactly what one extra year of schooling is worth. But before we get into how human DNA solves an economic problem, we really need to understand why this question has been so notoriously impossible to answer.
1:41Yeah, we have to start with how economists usually measure the value of a degree. They use these traditional observational studies, which are often called OLS models. Okay, OLS models. Right. And historically, if you run the numbers on the general population using these standard models, they show a really clear premium. The baseline data suggests that an extra year of schooling boosts a person's earnings by about 5.9%.
2:04Okay, so a 5.9% bump per year of school. That sounds like a, well, a solid definitive number. You see that and you just think the case is closed. You would think so. But that number is totally plagued by the classic correlation versus causation trap.
2:20Right, of course. In statistical terms, that 5.9% is contaminated by what we call unobserved confounding. Unobserved confounding, meaning like the invisible variables that secretly drive the results, but that your spreadsheet just isn't tracking.
2:35Precisely. I mean, think about all the complex intersecting factors that might cause someone to stay in school longer. Like having rich parents. Yeah, they might come from a wealthy family that can afford tutors, or, you know, they might have a high baseline of raw innate ambition, or even just excellent physical health.
2:53Yeah, or highly supportive home environment where their parents read to them constantly. Exactly. And the problem there is pretty obvious. Those exact same factors. The ambition, the health, the wealthy family connections, those are the very things that directly cause someone to land a higher paying job later in life.
3:09Right. So when you look at a spreadsheet, you can't tell if the extra education actually caused the higher salary, or if ambitious, well-supported people just happen to do both things. Yeah, they stay in school and they get high paying jobs.
3:24It is fundamentally a sorting problem. The data conflates the value of the education with the traits of the person receiving it. I always think about this like, um, trying to measure the health benefits of joining a gym.
3:37Oh, that's great analogy. Yeah, because if you simply look at health outcomes. People with gym memberships are obviously gonna have lower resting heart rates and better cholesterol than people without them.
3:47Right, but you can't automatically credit the treadmill for that. Exactly. You have to ask, are they healthy because of the gym? Or do inherently health conscious, active people just tend to be the ones who buy gym memberships.
3:59Yeah, the membership is often just a signal of an underlying lifestyle. Right. And for years, a university degree was viewed the exact same way. Perhaps it's just a signal of underlying capability, not the actual thing building the capability.
4:12Yeah, the gym membership effect is a massive hurdle in social science. And researchers have spent decades trying to design studies that strip away that underlying lifestyle signal. Well, if the underlying lifestyle, like the family background and resources is what ruins the data, it seems like the logical fix is to study people who share the exact same background.
4:33Which is exactly what they tried next. Like studying brothers. Or, better yet, identical twins who grew up in the exact same house with the same parents. Shouldn't that completely level the playing field?
4:45Yeah, family-based designs were the next major evolution in this research. By comparing siblings or identical twins with the exact same DNA, you effectively neutralize that family level confounding? Because you control for the wealthy parents, the neighborhood, the books and the house, everything.
5:01Right. And when researchers ran these sibling and twin models on this exact Norwegian data set, the financial return of an extra year of school just plummeted. Wow, really? Yeah, it dropped from that observational average of 5.9% down to somewhere between 3.2 and 5.3%.
5:16Oh, wow. And when they isolated it strictly to identical twins, the financial return dropped all the way to 3.2%. That is a staggering drop. The return on education gets nearly cut in half simply by controlling for the household you grew up in.
5:32Yeah, it's huge. Which heavily implies the gym membership theory is largely correct, right? Like, a huge portion of your financial success is just your background masquerading as the value of your diploma.
5:44That's certainly what it looked like. But since this new deep dive goes much further than twin studies. There must be a fatal flaw in relying solely on siblings. So what's the flaw? The flaw lies in what psychologists call the non-shared environment.
5:58Even within the exact same household. Siblies just do not experience life identically. I mean, anyone with a sibling knows that's true. Exactly. One twin might randomly encounter a highly influential high school teacher who sparks a love for math, while the other twin gets a substitute teacher and just completely loses interest.
6:15Or one sibling gets sick as a kid and falls behind. Right. A minor childhood illness that sets them back slightly in their development, and those individual differences would drive both the decision to stay in school and their later career success.
6:28Oh, I see. So the twin who met the great math teacher goes on to get a master's degree and becomes a lucrative engineer, but the difference wasn't the degree itself. Exactly. It was the spark of ambition from the teacher.
6:40So even with identical twins. You still haven't isolated the pure, untainted effect of the classroom. We're still stuck in the trap of unobserved confounding just at the individual level this time. Yeah, precisely.
6:53So to permanently escape this trap, researchers realize they needed a truly randomized variable. Okay, what does that mean in this context? They needed an event in a person's life that dictates their education level, but is completely entirely disconnected from their family's wealth, their personal ambition, or their childhood experiences.
7:12Finding a randomized variable like that in the messy real world sounds literally impossible. It does, but this is where the researchers introduce the concept of the genetic lottery. The genetic lottery.
7:23Okay, this is the core methodology of the paper. Yeah, they turn to a technique called Mendelian Randomization. It bridges molecular biology and economics by taking advantage of how human reproduction works at a microscopic level.
7:34Let's get into the mechanics of that. How does Mendelian randomization actually provide this perfectly randomized variable? Well, during conception when DNA is passed from parents to a child, the exact genetic variants the child inherits are shuffled, and they are allocated completely at random.
7:51The biological coin flip. Exactly. Now, over the last decade, geneticists have mapped out tiny variations in our genetic code, these are known as single nucleotide polymorphisms or SNPs. Okay, to clarify, SNP is basically a single typo or variation in the 1000000000s of letters of DNA code.
8:10Yes. And the researchers focused on what, 335 of these specific variations? Yeah, they identified 335 specific S&Ps that possess a statistical association with how long someone stays in educational systems.
8:22Wait, when you hear phrases like genes linked to education, it is dangerously easy for the mind to jump straight to dystopian biological determinism. Oh, absolutely. You immediately picture some sci-fi smart gene that dictates whether someone is destined to get a PhD or drop out of high school.
8:39We really need to be rigorously clear about what these 335 variations actually do to a human being. That is such a vital distinction. Let me be clear. There is absolutely no single smart gene. Biological determinism has no place in this model whatsoever.
8:55Okay, good. So what are they? These 335 variants are incredibly subtle predispositions. They do not force anyone to get a university degree. They merely act as a microscopic random nudge. A nudge, like how?
9:08Well, perhaps they make sitting in a structured classroom slightly less agonizing for a kid. Or they subtly influence a person's dopamine response when they complete a complex task. Okay, so it isn't destiny at all.
9:20It's just a tiny, invisible tailwind. The tiny tail one is the perfect way to describe it. But here's the thing. For an econom nutrition. That random biological tailwind is the holy grail. Because it's totally random.
9:31Exactly. Because these genetic variants are assigned at conception via a literal roll of the dice, they create a massive, naturally occurring, randomized controlled trial. So some people randomly inherit a slightly higher dose of these education linked genes, and some get a lower dose.
9:49Right. And it's completely independent of whether their parents were billionaires or blue collar workers. That random assignment provides a clean mathematical instrument to separate the pure effect of the education from the messy environment.
10:02You got it. If a person stays in school an extra year, solely because of this random genetic coin flip, then their family wealth or their neighborhood didn't contaminate the data. And this introduces a crucial concept in this type of research, which is called the compliers.
10:17Compliers. Like they comply with the gene? sort of, yeah. The Mendelian Randomization model is not measuring the financial return on education for every single person in Norway. It specifically isolates and measures the economic returns strictly for the compliers.
10:32Meaning the specific individuals in the data set, who were nudged to stay in school longer, specifically because of this random genetic predisposition. Exactly. They complied with the biological nudge.
10:42Okay, so we now have this perfect mathematical instrument, a natural, randomized trial using the DNA of over 100,000 people. It's an incredible data set. When the researchers finally run this population through the genetic filter, to strip away every ounce of confounding variables, what actually happens to the financial value of a year of school?
11:02Okay, here are the key findings. The Mandelian randomization reveals an 8.0% increase in lifetime earnings for every single extra year of schooling. 8.0%. Yeah. The paper translates that to about 53,622 Norwegian Kroner, which is roughly 5500 US dollars per year, compounding every year of your working life, just for one extra year of school.
11:28And it gets even more pronounced. The researchers went a step further and built a localized genetic model just for Norway. Okay, why would they do that? They built a polygenic index, entirely from Norwegian siblings, to ensure the genetic markers perfectly mapped to the local population's unique history.
11:44Oh, that makes sense. And when they ran that highly localized model, the financial jump was 9.7% in earnings per year of school. Wait, I am looking at these percentages and there is a glaring contradiction here.
11:56Yeah, I know what you're going to say. When researchers used identical twin studies to remove family background, The financial return dropped from the 5.9% average down to 3.2%. Right. But when they use the genetic lottery to remove family background, it rockets up to 8.0% and even 9.7, why would stripping away the rich family advantage make the financial return go up compared to the traditional average?
12:22It is arguably the most counterintuitive finding in the entire paper. And the researchers synthesized 2 brilliant explanations for why the traditional data was artificially suppressing the true value of education.
12:33Okay, what's the 1st explanation? The 1st one brings us back to the compliers. Right. The individuals who stayed in school specifically because of the genetic nudge. Exactly. It is highly probable that these compliers are individuals who derive a massively disproportionate payoff from extra schooling compared to the general public.
12:50How so? Well, these might be individuals from lower socioeconomic backgrounds, where one extra year of high school or a college is totally transformative. It radically alters their career trajectory. Ah, I see.
13:04Because for a wealthy student with endless safety nets, one extra year of school might barely move the needle on their lifetime wealth. Right. The genetic model zeros in on the people where the education actually does the heavy lifting.
13:16That makes perfect logical sense. The people who need the educational nudge the most, see the highest returns. But what is the 2nd explanation, because that alone can't fully explain why the national average is sitting way down at 5.9%.
13:29No, it can't. The 2nd explanation requires understanding the macroeconomic context of Norway, which introduces this fascinating concept called negative confounding. Negative confounding, meaning the invisible variables are actively lowering the statistical average.
13:44Exactly. In traditional data, we assume unobserved traits, like ambition, artificially inflate the value of a degree. But in Norway, those same traits might actually be dragging the average down. How does that happen in the real world?
13:59Well, Norway has a highly egalitarian compressed wage structure. The gap between the lowest and highest earners is much smaller than in a country like the United States. Okay. And crucially, Norway also possesses incredibly lucrative natural resource industries.
14:15We're talking maritime operations, deep sea fisheries, and a massive oil sector. Oh, wow. Okay. I think I see where this is going Right. In that specific economic environment. Consider what a highly ambitious, fiercely entrepreneurial teenager might do.
14:28They might look at the landscape and decide they just don't need a university degree. So they drop out early? Yeah, they drop out, go work on a deep sea oil rig or start a maritime logistics business, and they amass a total fortune.
14:39The highly ambitious go-getters are actually bypassing the educational system entirely. And because they still get incredibly wealthy, They make the less educated statistical pool look artificially rich in the basic data.
14:52The oil rig entrepreneurs are skewing the traditional averages down, completely hiding the true value of the classroom. Exactly. But Mandelian randomization mathematically strips those ambitious, wealthy dropouts out of the equation, because it relies strictly on the random genetic assignment at birth.
15:11That is brilliant. So once you isolate the pure effect of the education, untainted by the skewed data of natural resource entrepreneurs, the true financial return to the schooling itself is revealed to be a robust 8.0%.
15:25The economic context of the country literally camouflaged the value of the degree until they looked at the DNA. It really did. An 8% lifetime bump is staggering. But we have to ground this in reality because that wealth absolutely does not materialize the moment you receive your diploma.
15:40I mean, anyone who has been a 22 year old college graduate knows you're usually not rolling in cash. Oh, definitely not. And the researchers map this out perfectly through what they call a lifecycle analysis.
15:49Okay, what does that look like? They track the age earnings profile across decades. And the data confirms exactly what you were describing. In your early 20s, the financial returns on education are heavily negative.
16:00Seriously, if you are listening to this, and you have ever been a 23-year-old grad student, sitting in a library eating instant noodles, while scrolling through social media, watching your high school friends who started working at 18 buy new cars, and put down payments on houses.
16:17Oh it hurts. Yeah, this is that exact painful reality showing up in the econometric data. It is the brutal mathematics of opportunity cost. While the student is sitting in a lecture hall, their less educated peers are fully integrated into the labor market.
16:31They're accumulating years of real world experience. Yeah, they are securing promotions, and most critically, they are earning compound wages while the student is earning absolutely zero, or actively accumulating debt.
16:43It is the agony of the 20s. So when does the math flip? At what age does this 8% causal engine actually kick in and allow the educated group to overtake the people who went straight into the workforce?
16:54The longitudinal data reveals a highly specific crossover point. It's age 27. Age 27. So it takes nearly a decade after high school for the investment to break even. Yeah, right around your 27th birthday, the earnings trajectory of the educated cohort steeply accelerates and permanently surpasses the non-colle educated group.
17:14And stays that way. From that age onward, the wealth gap just continues to widen into retirement. And to definitively prove that enduring the negative returns of your 20s is actually worth it, the researchers calculated the internal rate of return, or IRR.
17:29We usually hear IRR in the context of like venture capital or buying stocks. It essentially measures the annualized yield of an investment. Exactly. By treating an education like an investment asset, balancing the early financial losses against the massive gains in a person's 40s and 50s, the genetic model shows the internal rate of return for education is 10.one%.
17:50Wow. Yeah. And when you benchmark that 10.one% against the standard market interest rate, which the paper notes was around 2.3% during this period. The conclusion is undeniable. Education is a wildly successful, highly lucrative financial asset over a complete lifecycle.
18:07Absolutely. And the paper also noted a really significant demographic split in how those returns are distributed, specifically regarding gender. Oh, interesting. What did they find? The models consistently demonstrated that women experience considerably higher financial returns to schooling than men do.
18:24Really? Yeah, the percentage increase in lifetime earnings they secure for every extra year of education is significantly larger. That likely points to education acting as a much stronger, more necessary signaling mechanism for women in the labor market, right?
18:39Like to overcome baseline structural disparities. is exactly the implication. Well, we now have a definitive causal answer. An 8% lifetime return, a 10.one% internal rate of return, and a crossover point at age 27.
18:52But rigorous science requires attempting to break your own model. The entire foundation of this deep dive relies on the genetic instrument working perfectly. What if the Mendelian randomization itself is fundamentally flawed.
19:04That is the big question. And the researchers spend a massive portion of the paper violently stress testing their own DNA model. How so? The single biggest existential threat to this study is a biological concept called pleotropy.
19:19In econometric terms, cleotropy violates the exclusion assumption. The exclusion assumption, meaning the model assumes those 335 genetic variants do one thing and one thing only, nudge a person to stay in school.
19:33It assumes the genes affect earnings exclusively by keeping you in a classroom. Yes, and that is the fragile linchpin of the entire study. So let's test that. What if those 335 genetic variants are multitaskers?
19:45What if they don't just nudge you to study, but they directly alter your biology in ways that cause you to earn more money through totally unrelated pathways? Like what? Imagine these genes make you slightly taller, or they fortify your immune system so you literally never take sick days.
20:00Or they make you naturally charismatic. If you possess those genetic traits, you would earn drastically more money as a salesperson, and it would have absolutely 0 to do with the fact that you hold a college degree.
20:10That scenario is the ultimate nightmare for genetic economists. Because if the genes are directly causing the wealth, via charisma or physical health, completely bypassing the educational system. Then the 8% return we just discussed is a total illusion.
20:25The methodology collapses. So how do the researchers possibly prove that these genes aren't just secretly churning out highly charismatic salespeople? They deployed a gauntlet of highly technical sensitivity analyses.
20:39These are statistical models with names like M.R. Eger, M.R. Median, and M.R. Gorge. Okay, we don't need to get bogged down in the acronyms. But we need to understand the mechanism. Sure. These models are basically designed to mathematically hunt down and penalize multitasking genes.
20:54How does a math equation catch gene multitasking? Well, the models analyze the relationship between the genetic variants across the population. If the algorithm detects that a specific genetic variant is massively boosting a person's earnings, but is only weekly associated with their time spent in school, the model flags it.
21:12Ah, it flags it as a pleotropic multitasker. Exactly. It realizes that gene must be boosting wealth through an outside channel, like your hypothetical charisma. And then what? The model then mathematically suppracts that specific non-educational financial boost from the final calculation.
21:28Wow. Yeah. It actively strips away the wealth generated by height or health. leaving only the pure financial boost that flows strictly through the classroom. So they forcefully handicap their own data to see if the education benefit survives on its own.
21:42Did the model hold up? It held up with astonishing resilience. The researchers calculated exactly what percentage of the genetic effect would have to be secretly bypassing education to completely ruin their findings.
21:55And what was the threshold? They discovered that even if a staggering 89% of the genetic effect operated entirely outside of the classroom. Wait, 89%. Yeah. Even if 89% of the financial benefit was just from the genes making you taller, healthier, or more charismatic, the causal financial return of the education itself still remains positive and statistically significant.
22:16That is wild. So even if the genetic variants are 89% busy doing other biological tasks, the remaining 11% that flows through the school system still irrefutably proves the degree makes you richer. The mathematical tool is just incredibly robust.
22:32The researchers through every possible statistical anomaly at their own findings to see if the causal link between the classroom and the bank account would shatter. And it simply did not. No, it didn't.
22:43The act of receiving the education is genuinely directly causing higher earnings. Let's distill the ultimate takeaway for you listening. We started with the oldest advice in the book, Stay in school, but for decades, the data was polluted by family wealth, raw ambition and environmental look.
22:59By using the random genetic lottery at conception as a literal time machine, researchers managed to isolate the pure, uncontaminated effect of education. They proved that a diploma is not just a hollow signal of existing privilege or innate talent.
23:13No, it is a direct causal engine for generating wealth. Yielding an 8% return per year that wildly outpaces standard market investments. And honestly, the methodology is just as profound as the finding.
23:25It demonstrates how integrating molecular genetics into social sciences can finally resolve generational debates that have been stuck in the mud for 50 years. It totally redefines how we measure human potential.
23:37But before we wrap up this deep dive, There is a slightly provocative thought worth mulling over. Oh, what is it? We have spent this time marveling at how researchers use the random genetic lottery to conclusively prove that education creates wealth.
23:53But let's flip the lens for a second. Flip it how? What does this mean for meritocracy? Think about it. If the literal biological coin flip at your conception randomly nudges your likelihood to stay in school, and that schooling, in turn, heavily dictates your lifetime wealth.
24:09It forces a deep philosophical tension. It really does. We love to conceptualize our economy as a pure landscape, where hard work and grit are the sole drivers of reward. Yeah, the self-made myth. Exactly.
24:20But if our economic success is fundamentally anchored to microscopic biological coin flips that subtly influence how easily we adapt to rigid educational systems. How much of our success is truly our own doing.
24:34And how much of it is just the look of the draw. Right. Just something to think about the next trim. look at a diploma hanging on a wall. Absolutely. This episode was based on an open access article under the CCBY 4.0 license.
24:48You can find a direct link to the paper and the license in our episode description. If you enjoyed this, follow or subscribe in your podcast app and leave a 5 star rating. If you'd like to support our work, use the donation link in the description.
25:00Now stay with us for an original track created especially for this episode and inspired by the article you've just heard about. Thanks for listening and join us next time as we explore more science base by base.
