Why I Changed My Mind About the Caster Semenya Case

I used to think that rules regulating testosterone in sports made sense. Now I’m not so sure.

Caster Semenya smiles and gestures.
Caster Semenya at the Athletics Gauteng North Championships in Pretoria, South Africa, on March 13. Phill Magakoe/AFP via Getty Images

This story also appears in David Epstein’s newsletter, Range Report.

Last week, Caster Semenya lost her appeal of a rule by World Athletics (the governing body for international track and field) that established a testosterone limit for certain women in certain events. Semenya has a rare condition that elevates her natural testosterone level above that limit. She will not be allowed to defend her Olympic title in the women’s 800 meters unless she takes testosterone suppression medication, which she has said repeatedly she is not willing to do. The me who first wrote about Semenya in 2009—when she literally ran away from the field at the World Championships—would have felt last week’s ruling was the right one, albeit unfair to Semenya. Now, I feel differently—still torn, but more amenable to the argument that testosterone regulation should be off the table for athletes like her.

In my book, The Sports Gene, and in Sports Illustrated, I delved into the voluminous scientific rabbit holes of this debate. I won’t do that here. Instead, I’m going to share a few high-level concepts, and try to explain why I changed my mind.

I want to start by describing two friends of mine, a pair of scientists on opposite sides of the Semenya legal battle.

Joanna Harper is a transgender woman and medical physicist. She’s also an accomplished age-group distance runner. I first got in touch with Harper for a Sports Illustrated article on transgender athletes, co-authored with Pablo S. Torre. When she began hormone therapy to suppress her body’s testosterone in 2004, she started collecting data. She was getting slower and weaker by the end of the first month. In 2003, before hormone therapy, Harper ran a half-marathon in 1:23:11. In 2005, after a year of testosterone suppression, she ran the same half-marathon in 1:34:01. Relative to her age and gender category, those were practically identically good performances. In 2007, she won the women’s 50–54 age group at the USA Track and Field Club Cross Country Championships, but age and gender-graded performance standards continued to show that Harper was similarly competitive against women after testosterone suppression as she had been before it against men. Harper has compiled (and published) data from other distance runners who transitioned and found the same pattern.

Like no one I have ever seen, Harper has an ability to engage in controversial conversations without pandering or making others feel attacked. She managed to insult Tucker Carlson on his own show without starting an argument and left him with what appeared to be grudging respect. I wrote the foreword to Harper’s authoritative and fascinating book, Sporting Gender. I could hardly admire her more. Harper believes that, absent medical intervention, Caster Semenya should not be allowed to compete in the women’s 800 meters. She testified for World Athletics in the Court of Arbitration for Sport, or CAS, hearings that upheld the testosterone regulation.

Ross Tucker (aka @scienceofsport on Twitter) is a South African sports scientist. The more controversial the topic, the more likely Tucker is to be part of the discussion, usually highlighting an angle no one else has spotted. For years, he pointed to evidence suggesting that the carbon-fiber blades worn by double-amputee sprinter Oscar Pistorius gave him an unfair advantage in Olympic competition. When Tucker began arguing this point, Pistorius was still an inspirational and beloved A-list celebrity in South Africa. (This was before Pistorius shot and killed his girlfriend in 2013.) You can imagine how kind the nation’s media and sports fans were to Tucker, but he stood his ground. I’m a massive admirer of his, even when (perhaps especially when) I disagree with him. In the Court of Arbitration for Sport hearings, Tucker was an expert witness for Athletics South Africa, which argued, along with Semenya’s team, to throw out the World Athletics testosterone rule.

Harper and Tucker both agree that:

• Testosterone is the primary source of the male advantage in elite sports (the CAS decision upholding the testosterone rule was 2–1, but all three judges agreed that “testosterone is the primary driver of the sex difference in sports performance”).

• The lower limit of the typical male testosterone range is about four times higher than the high end of the typical female range.

• Testosterone levels are the best (that is, the least imperfect) marker for separating the men’s and women’s competitive classifications.

Both Harper and Tucker believe that the female classification in sport is important to maintain and that there should be some criteria for entry into it that protects the integrity of women’s competition. Part of the challenge is that women like Semenya with differences of sex development, or DSDs, sometimes have testosterone in the typical male range, or that falls between the typical male and female ranges. (One common pro-Semenya argument is that Michael Phelps’ long arms gave him an advantage over other swimmers, but nobody talked about kicking him out of the Olympics on account of his oversize limbs. I find that unconvincing. We don’t separate swimmers by arm length. There are, though, separate competitive classifications for men and women in track and field. Given that those classifications exist, they need to have some meaning, and there needs to be some dividing line between them.) When Harper and Tucker testified for opposite sides in the Semenya proceedings, it was regarding the research that World Athletics relied upon to create its particular testosterone rule. Tucker felt that research was insufficient. Harper supported World Athletics and later wrote that “the verdict is congruent with my belief that both inclusion and meaningful sport for all women are important.”

It’s important to understand that the World Athletics rule doesn’t set a testosterone limit for all female competitors. It sets a limit specifically for women with certain DSDs—those who have XY chromosomes, internal testes and not ovaries, testosterone levels in the typical male range, and whose bodies can respond to that testosterone. It also doesn’t set a testosterone limit for these women in all events. The rule only forces them to lower their testosterone levels if they want to compete against women in the 400, 800, and 1,500 meters. Those are the events, World Athletics says, in which testosterone levels and performance, combined with other evidence—like the historical overrepresentation of women with DSDs in the 400, 800, and 1,500—suggest the clearest testosterone-linked advantage. If Semenya opts to lower her testosterone, she can compete against women in those events. If she doesn’t, she can compete against women only in other events.

But the World Athletics research also shows an advantage for women with elevated testosterone levels in the hammer throw and pole vault. Those events have not been regulated, at least not yet. Why?

In a paper that Harper co-authored, researchers wrote that World Athletics may not have included the hammer and pole vault because athletes with DSDs have not been as significantly overrepresented on the Olympic medal stand as they have been in the other events. On the one hand, it is responsible scientific thinking to apply a rule only where the confluence of evidence is strongest. It also suggests that World Athletics realizes it is not drawing some bright line between men and women for all sports. On the other hand, it’s an awkward fit with the idea that testosterone is the primary driver of the male advantage in all events, and with the pretty consistent 10 percent performance gap between the fastest men and the fastest women at all distances. (In the paper, Harper and her co-authors speculate on some of the potential reasons that women with DSDs may not have been as historically successful in the hammer throw and pole vault.)

For most people, their genes, chromosomes, reproductive organs, and gender identity align into the categories typical for cisgender men or women. But that isn’t true in every case. Dividing athletes into male and female groups requires imposing a clear binary on a messy reality. There is no science that will change that. As Tucker has said: “This is easily the most complex issue that sports has had to deal with, perhaps ever.” It’s so complex that Harper and Tucker, who are also friends, can agree on the social and scientific principles and yet disagree about the bar for evidence that leads to regulations.

Over time, I’ve become more drawn to the argument that athletes, like Semenya, who have been raised as women and who have lived as women and who identify as women should be allowed to compete as women. But Joanna Harper wouldn’t be Joanna Harper if she didn’t push me to think very hard about why I feel that way.

Harper also pressed me to think about how the “ ‘born female, raised female’ argument,” as she refers to it in her book, would even work. We can’t know gender identity at birth, so in practice, people are assigned female when they’re born without a penis. But external genitalia aren’t a good way to divide male and female athletes. Until the guidelines were changed in 2016, the International Olympic Committee said that transgender women who have penises should have them surgically removed to compete with cisgender women. Given that nobody thinks a penis is the source of the male advantage in sports, that made no sense, along with being inhumane.

Christie Aschwanden, one of the best science writers in the world, a former elite cross-country skier, and another friend of mine has argued that “recogniz[ing] Semenya under the gender identity that she has inhabited since birth” is sound policy. Any kind of testosterone testing, she argued, will harm not only women with high testosterone, “but also every other woman athlete who looks too ‘manly’ or otherwise does not conform to someone else’s notions of what a woman should be.”

I can count on my hands the number of people whose opinions I value as much as Harper’s and Aschwanden’s. And I don’t think either one is necessarily wrong. Our society—and, as a consequence, our sports—hasn’t been set up to accommodate gender-nonconforming individuals. That is not fair or just, and it puts us in the impossible position of selecting which kind of unfairness and injustice we believe is the least harmful to the fewest individuals. As Aschwanden put it: This is a trolley problem, and World Athletics—and anyone who cares about Semenya and track and field—has to acknowledge that no matter what decision gets made, someone is going to get run over.

I started running the 800 meters late in high school; within a few years, my best time was faster than the women’s world record. It doesn’t seem fair for the most-accomplished, hardest-working women in the sport to compete against someone with my physiology for their livelihoods and podium spots. World Athletics has argued that, as it pertains to the 800, Semenya has all the physiological advantages of a male runner. (Her times in the event wouldn’t be nearly world class for a man, but they would be good enough for her to compete on a men’s college team.) I think it’s good for athletes, sports fans, and society to have fair women’s competition at the elite level. The Court of Arbitration for Sport ruling acknowledged the harm of the testosterone rule to athletes like Semenya but said that “such discrimination is a necessary, reasonable and proportionate means of achieving the legitimate objective of ensuring fair competition in female athletics in certain events and protecting the ‘protected class’ of female athletes in those events.”

Semenya has been an exemplary champion, performing admirably on and off the track in the face of tremendous pressure and scrutiny. The initial handling of her case was abysmal and inhumane. After she won the African Junior Championships in 2009, Athletics South Africa—at the request of World Athletics—conducted sex-verification testing on Semenya; she was not given an explanation and thought the tests would be for doping. When she became world champion in Berlin that same year, Semenya said that World Athletics took her to a local hospital where she “had no choice but to comply” with more testing. Leaks of medical information and tin-eared public remarks from World Athletics officials enmeshed Semenya in what she called “the most profound and humiliating experience of my life.”

World Athletics’ approach has improved enormously in substance and in tone since then, but I still have lingering mistrust. I’m inclined to have the governing body proceed by the precautionary principle and get involved in sex testing as little as possible. That is, given that World Athletics is the one driving the trolley, we should make sure that it’s moving along the tracks incredibly slowly and deliberately.

My experience doing investigative reporting on medicine has also made me feel skittish about medical interventions for asymptomatic people, even if they’re generally safe. If Semenya wants to try to defend her Olympic title, she would have to undergo testosterone suppression for no medical reason. I think it’s reasonable to suspect that that could change her life in ways that she finds undesirable. In fact, she has said as much about a five-year period when, at the behest of World Athletics, she underwent testosterone suppression; she remained world class but got steadily slower. (Alternatively, she could compete in other events. Sprinter Aminatou Seyni, who hails from Niger, would have been a World Championships medal contender in the 400 this year without a testosterone rule; she opted not to lower her testosterone and instead competed in the 200, where she made the semifinals.)

And now for one more layer of complexity. Harper and Tucker may soon be on different sides of another legal battle in sports, about transgender athletes in rugby. And this time, they’re on sides opposite to what you might guess based on their roles in the Semenya case.

Tucker collaborated on a report that concluded that allowing transgender women to play on women’s teams carries at least a “20 percent to 30 percent greater risk” of injury to women who never went through typical male puberty. Based on that report, World Rugby just proposed banning transgender women from women’s competition. Harper once told me that “for cardiovascular factors, trans women go from typical men to typical women after transition, but with strength, they go from typical men to somewhere in between typical men and typical women.” Still, she thinks the evidence for a ban is shoddy.

Harper and Tucker—and I—all think that having guidelines in Olympic sports and the NCAA that require athletes transitioning from male to female to undergo testosterone suppression (if they want to compete in the women’s category) is reasonable. In that scenario, the athlete is proactively seeking to switch her competitive classification, and I think it makes sense to have rules for managing that. But even there, I’m not perfectly consistent; I don’t think the same guidelines should apply to Olympians as to youth athletes. Harper has floated the idea of allowing transgender girls who have not undergone testosterone suppression to compete against girls in high school track but not at the state championship level. Last week, Harper explained to me, “I think that in most—but not all—matters, we should use gender identity to divide people into male and female categories.”

I used to be troubled by this feeling of inconsistency. I don’t love it now, but I’m more comfortable with it. I think Harper and Tucker must be comfortable with it too, which is why they can take positions in track and rugby that might look contradictory from the outside. They’re not actually contradictory at all. In its decision, CAS called the testosterone regulations a “living document”; experience may show the rules need to be altered for appropriate implementation, or new evidence may show that fewer or more events should be regulated. I’m sure that both Harper and Tucker would love to see new research, and that both would be open to changing their views if warranted. They’re just trying to follow the best available (necessarily imperfect) evidence in each individual case, while keeping humanity and the values of sports in mind.