Can you suffer your way to success? In a recent Financial Times column, Pilita Clark describes the harrowing procedure of cosmetic surgical leg lengthening now performed around the world hundreds of times a year. Patients (nearly all men), pay up to $150,000 to have their femurs extended by three inches plus — another three for their tibias. The process begins by breaking the bone, extracting the marrow with a device called a reamer, and inserting pins in the bone to be anchored in a Soviet-developed mechanism known an Ilizarov frame. As the Canadian-born Las Vegas surgeon who has been refining the procedure put it, bone is a uniquely self-renewing material that will fill in the space left by gradually increased gaps. (In his boundless enthusiasm for his specialty, the doctor reminds us that bone is a not just tissue, but rather “the most exciting tissue in the world.”)

While Clark doubts the pain is worth the gain and notes the long recovery and potentially agonizing complications of leg-lengthening, she repeats the statistics concerning height and business success suggesting the patients are “not entirely mad.” She cites one study concluding that 10 centimeters (4 inches) of height yield 3 percent higher pay for men, along with the research showing that the average height of CEOs in Sweden and the United States exceeds six feet.

A No-Brainer, Except It Isn’t

What Clark doesn’t mention is that the causal relationship between height and success turns out to be far more complex as well as interesting when you drill down. A more subtle mechanism explaining the correlation was ingeniously revealed by a trio of University of Pennsylvania economists, Andrew Postlewaite, Nicola Persico and Dan Silverman, almost two decades ago.

Not all tall men, it turns out, enjoy an earnings advantage — only those who were already tall for their age at 16. Equally tall men who had experienced later growth spurts took home no more than the average. Conversely, men who had been taller than their peers at 16 but grew to only average adult height nonetheless earned as much as the tall men. 

After ruling out nutrition, socioeconomic background and native intelligence, an intriguing explanation remained. Adolescent height, it seems, helps form critical formative experiences, at least for men.

Shorter male teenagers are less likely to participate in extracurricular high school activities like clubs, student government and especially athletics, which help build self-confidence, teamwork and leadership skills. Obviously, there are exceptions. But these potential deficits may persist even after later growth, while the early benefits of height are conserved even after young men’s growth tapers off to average stature.

The researchers benefited from the existence of two massive databases, the U.S. National Longitudinal Study of Youth and the British National Child Development Survey. But the most important factor in ferreting out this striking result was the economists’ determination to look beyond conventional interpretations of the height premium to understand the importance of exceptions.

The historian of science Thomas Kuhn was famous for urging his students to scan for oddities in the texts they were reading as clues to a deeper consistency. Many great and small inventions from X-ray imaging to Post-It Notes have been inspired by surprises in experiments. The same thinking might be applied to data analysis — especially when there is a long-established, almost self-evident conclusion ripe for revision. (While the 58 percent advantage of taller U.S. presidential candidates also mentioned in Clark’s column sounds impressive, another analysis cited in Wikipedia suggests the real difference is a “round-off error” that vanishes entirely when the one election with a female candidate — 5’6” Hillary Clinton — is excluded.

As with virtually every important study, the Penn researchers’ answers lead to more questions. How does height affect the success of women and groups already disadvantaged by discrimination, especially Black men and women? (Unsurprisingly, there is evidence that police discriminate in street searches against taller Black and Latino men.) Has belief of the inherent advantage in being tall become a self-fulfilling prophecy, and might wider knowledge of the Penn results strengthen self-confidence without surgery? The great challenge in the age of big data is spotting missing but crucial variables such as the outlier behavior of non-college graduates in the 2016 election — an absence that led pollsters to underestimate the following of Donald Trump.

• • •

It is probably too early to judge the long-term socioeconomic consequences of leg-lengthening. There is evidence from South Korea, arguably the global capital of cosmetic surgery, that patients do not recover the costs of the procedures through additional income — though, of course, there may be non-pecuniary psychological benefits. What we can infer from the tale of height and income: it is essential not only to have good data, but also to have models of behavior that capture the elusive complexity and unsuspected linkages of homo sapiens.