Intro. [Recording date: December 22, 2025.]
Russ Roberts: It’s December 22nd, 2025, and before I introduce today’s guest, a quick reminder to visit econtalk.org and cast your votes for your favorite episodes of the year.
Today, we’re joined by mathematician and author David Bessis. He last appeared in October 2025 to discuss his intriguing book, Mathematica.
Our discussion today will revolve around a thought-provoking essay from his Substack, linked in our notes. The title? “Twins Reared Apart Do Not Exist: The Shaky Science of Genetic Determinism.” Welcome back to EconTalk, David.
David Bessis: Hello, Russ! It’s great to be back.
Russ Roberts: Your book, Mathematica, presents what many might consider a controversial perspective on mathematics. Unlike Henri Poincaré’s assertion that “mathematicians are born, not made,” you contend that nearly anyone can master math with the right approach. You suggest that the so-called great mathematicians aren’t inherently genius; rather, they possess a powerful, learnable method combined with relentless focus and effort. You reference figures like Descartes, Einstein, and Grothendieck to support your argument, rejecting the notion that genetic determinism defines our capabilities.
Yet, you found yourself troubled by a claim regarding twin studies that assert genetics dictate destiny. You encapsulated this in your essay, which states: ‘When you separate two identical twins at birth, raise them in two random families, and test their IQs in adulthood, the two results are barely more divergent than two different tests of the same person.’ Essentially, the idea is that regardless of nurturing differences, genetic predisposition may dominate, suggesting we are hardwired for intelligence from birth.
Can you elaborate on why this notion was unsettling for you? In your essay, you admit to having a stake in this debate.
David Bessis: Absolutely. This debate is often framed too neatly into two extremes: either everything is genetically determined, and you’re powerless to change your fate, or the blank slate theory suggests that anyone can become anything. I reject both extremes. I think it’s critical to acknowledge the complexity of the issue. While genetics undoubtedly play a role—some genetic conditions can severely impair cognitive function—it’s not the be-all and end-all. We see variability in traits like brain size, which, while correlated with outcomes, does not dictate them. The heart of the matter lies in the distinction between genetic influence and total determinism.
Many of the twin studies you mentioned were conducted decades ago and have gained traction, especially on social media. A prominent figure like Paul Graham shared a graphic summarizing these findings, leading many to accept the conclusion without question. If the studies indeed show no significant differences between twins raised apart, it would imply a profound truth about genetic determinism.
This contradicted my personal experiences and understanding of mathematical talent, which I articulated in my writings. When I saw Graham’s post while finalizing my book, I felt a pang of doubt. If that claim were true, then perhaps my perspectives—and those of influential figures like Einstein and Descartes—were misguided. I had to dive into the science.
What I discovered is that the science surrounding this topic is far more nuanced than the prevailing narrative suggests. The ideal experiment, as described, simply doesn’t exist in real life, which was a surprising revelation.
Russ Roberts: It’s worth noting that there’s a chance Einstein, Descartes, Grothendieck, and even you might be exhibiting a touch of false modesty. Perhaps you’re more gifted than you’re letting on.
Interestingly, you chose not to embrace the narrative of inherent genius. Many people prefer to attribute their successes to extraordinary talent rather than hard work. Your self-assessment is refreshingly unique.
David Bessis: It raises intriguing questions about motivation and perception. When my book was released, I was interviewed by Quantum Magazine, where I explained my belief that genius isn’t innate but developed through unique journeys. I don’t suggest everyone can become a genius, but those who do often undertake extraordinary paths.
To quantify that, let’s consider a conjecture about the odds of becoming a one-in-a-million genius. Individuals like Einstein or Grothendieck may indeed be extraordinary, but it’s plausible they simply belong to the top one percent. Their unique journeys and experiences shaped their abilities. For instance, Bill Thurston, who faced challenges with his vision, had a supportive upbringing that allowed him to hone his perception and ultimately become a renowned geometer.
While it’s tempting to label someone as a genius, acknowledging the journey and effort behind their success offers a more plausible narrative.
Russ Roberts: As a graduate student, I noticed many peers preferred the narrative of inherent talent over the hard work they put in. It’s an interesting psychological phenomenon.
Russ Roberts: Shifting gears, let’s discuss the percentages you mention—30%, 50%, and 80%—in relation to the correlation of innate traits and their real-world outcomes. It’s fascinating how these figures are often discussed with an air of scientific precision, yet they lack the nuance required for a complete understanding. For instance, claims that IQ is 80% genetically determined oversimplify a complex reality.
Even if we accept these numbers at face value, you argue that 30% or 50% still leaves ample room for personal agency and choice in shaping one’s destiny, while 80% veers dangerously close to a deterministic viewpoint. Care to elaborate?
David Bessis: Your mathematical background certainly helps in discerning the nuances of this discussion. The conversation is muddied by numerous misconceptions. When we say, for example, that IQ might be 50% genetically determined, it’s essential to ask what that truly means for each individual, as it only captures a portion of the population.
In my essay, I included visualizations to illustrate these correlations. The heritability percentages indicate how closely linked genetic factors are with outcomes like IQ. A 50% correlation suggests a trend, but it remains weak. If you possess an average genetic predisposition, there’s still a possibility of reaching the top 1%, although it’s not a guarantee.
Moreover, the question arises: what remains unaccounted for? It’s common for parents of two siblings to observe significant differences in outcomes and leap to genetic conclusions. Yet, numerous factors contribute to these divergences, including birth order, parental dynamics, and family context.
Socioeconomic factors certainly play a role, but they don’t encompass the entirety of influence. Each individual’s life experiences, unique encounters, and even chance events can significantly shape outcomes.
Turkheimer’s Third Law emphasizes this variability—there’s a substantial amount of individual influence that neither genetics nor family backgrounds can fully explain.
Russ Roberts: I’d love for you to elaborate on Turkheimer’s laws for our audience.
David Bessis: Eric Turkheimer, a behavioral geneticist, articulated three laws of behavioral genetics, which shed light on the influences of genetics on behavior. The First Law states that everything is heritable to some degree; this is widely accepted. The Second Law posits that the family you’re born into often matters less than your genetic makeup, challenging the blank slate perspective.
Finally, the Third Law reveals that a significant amount of variability in traits comes not from genetics or family but from individual life experiences. This principle underscores the complexity and unpredictability of human development.
When I delved into Turkheimer’s work, I found his insights immensely valuable, particularly regarding the unique experiences that shape us. As he noted, no two siblings grow up in identical circumstances. Life is inherently complex, and often, what happens in your life can’t be easily categorized or predicted.
Russ Roberts: So, in essence, the relationship between genetics and environmental factors is both intricate and multifaceted.
Russ Roberts: I find Law Number Two somewhat troubling. While it may hold for easily quantifiable traits like height, where genetics plays a significant role, intelligence is a much more complex trait influenced by numerous environmental factors.
David Bessis: That’s true. Height is a straightforward example of a trait heavily influenced by genetics, but intelligence operates differently. While genetics certainly contribute, the environment also plays a critical role, making intelligence less predictable.
Russ Roberts: Exactly. My point was that while genetic factors are indeed significant, they don’t negate the effects of environmental influences. Intelligence, happiness, and other traits are challenging to quantify, complicating any deterministic narrative.
Russ Roberts: Let’s pivot to the science behind twin studies, particularly regarding IQ. I consulted Claude, our AI assistant, for a summary of the relevant studies. It’s clear that many studies often cited are not as reliable as they might seem.
You’ve pointed out that several twin studies have significant flaws, including issues with sample sizes and even outright fraud. However, a few studies do boast relatively larger samples. Can you walk us through the most notable findings?
David Bessis: Certainly. For instance, one of the most notable studies by Thomas Bouchard examined 56 pairs of identical twins and suggested that their IQ scores were astonishingly similar, regardless of the environments in which they were raised. However, it’s important to note that the circumstances surrounding these studies were unique and not representative of the broader population.
Identical twins, by nature, share not just their genes but also their wombs for nine months. The environment they’re born into, even when adopted into different families, may still share similarities due to cultural and socioeconomic factors. Thus, drawing broad conclusions from such limited and specific data is problematic.
Russ Roberts: The allure of these studies is understandable—they appear to offer a clear-cut answer to the nature versus nurture debate. Yet, the reality is far more complicated, and the pursuit of definitive conclusions can lead to misinterpretations.
You mentioned the documentary Three Identical Strangers, which, while dramatic, can also be somewhat misleading. The stories of separated triplets who share numerous similarities are captivating, but they may not accurately reflect the complexities of human development.
David Bessis: Yes, the Jim twins are a classic case from Bouchard’s research. Their uncanny similarities, from names to life choices, seem to suggest a genetic basis for behavior. However, upon closer examination, these parallels often arise from shared cultural backgrounds rather than genetic determinism.
When we unpack their experiences, it becomes clear that many of the similarities stem from societal norms and familial structures rather than genetic programming. This realization complicates our understanding of the implications of such studies and highlights the importance of context in interpreting results.
Russ Roberts: What’s particularly striking about the design of these studies are the flaws that often go overlooked. Can you elaborate on some of the issues with the Bouchard study and its methodology?
David Bessis: On paper, the twin studies appear to be well-structured natural experiments. However, they fail to account for critical factors, such as the shared prenatal environment. The twins may not have been separated at birth as often claimed, with many being separated only a few months later, which can skew the results significantly.
Moreover, the circumstances leading to their separation could indicate underlying familial dysfunction, further complicating the interpretation of their similarities. A thorough control for these biases is essential for accurate conclusions.
Russ Roberts: Let’s clarify the rationale for using fraternal twins as a control group. If we had thousands of both identical and fraternal twins raised apart, how would you use their data to validate the findings of the identical twins?
David Bessis: That’s a great question. In an ideal scenario with large samples of both identical and fraternal twins, the approach would involve comparing the correlation coefficients of IQ between the two groups. If genetics play a significant role, we’d expect a distinctly higher correlation for identical twins compared to fraternal twins.
For instance, if identical twins exhibited an 80% correlation in IQ while fraternal twins showed a correlation of only 50%, the difference would provide insight into the heritability of intelligence. However, due to the small sample sizes often encountered in these studies, drawing definitive conclusions remains problematic.
Russ Roberts: The implications of these twin studies are profound. If the data were to show no significant difference in the IQ correlation between identical and fraternal twins, it would suggest that genetics might not be as deterministic as previously thought.
There’s even criticism from figures like Jay Joseph, who argues that some of these findings lack significance. Can you weigh in on this?
David Bessis: Yes, Joseph’s perspective is noteworthy. He suggests that when examining the data from the Bouchard study, it may not show a meaningful genetic correlation, which further underscores the limitations of the research. The small sample sizes and the unique circumstances of the participants can lead to misleading conclusions.
Ultimately, while genetic influences certainly exist, the extent to which they dictate intelligence remains an open question, complicated by the methodological flaws and biases inherent in the studies.
Russ Roberts: As we wrap up, it’s worth reflecting on why these debates are significant. Parenting, for example, plays a crucial role in shaping children’s lives. Many parents wonder whether their efforts truly impact their children’s development, especially in light of conflicting research on the influence of nature versus nurture.
What are your thoughts on the broader implications of this debate?
David Bessis: This discussion is vital, as it touches on our understanding of agency and influence in human development. While genetics undoubtedly contribute to our capabilities, it’s crucial to recognize that parenting and environmental factors also play meaningful roles. We must avoid deterministic thinking that dismisses the importance of individual experiences.
Ultimately, while we may not have total control over our destinies, acknowledging the non-zero influence we can have on our lives encourages a more nuanced understanding of human behavior.
Russ Roberts: Thank you for sharing your insights today, David. It’s clear that the relationship between genetics and environment is complex and multifaceted. I appreciate your time and perspective.
David Bessis: Thank you, Russ. It’s been a pleasure!
