Sex Testing in Elite Sport

It's still all about testosterone.

|The Volokh Conspiracy |


If we're going to define the women's category in elite sport on the basis of sex, we have to be prepared to sort athletes in and out on those grounds. The Olympic Movement has a long and complicated history with this work, which has involved continually updating the approach to atypical cases so that the eligibility rule remains consistent with sport's core commitments and its evolving values. The current iteration, which rests exclusively on testosterone (T) levels, is illustrative. It represents a renewed commitment to a women's category based in female sex-linked traits and, within that framework, a new commitment to a pathway into the category for male-bodied athletes who identify as women.

Here are the specifics:

The IOC's Transgender guidelines require transgender women and girls who want to compete in the women's category to drop their T levels to below 10 nmol/l for at least a year before their first competition. (See my post on Tuesday for details about the male and the female T ranges.) It is expected that the required T levels will be revised to below 5 nmol/l.

The IAAF has taken the lead developing the eligibility regulation for 46 XY males with differences of sex development (DSDs). Athletes with DSDs are often described as "intersex" but for sport's purposes, the only relevant conditions are those affecting biological males, i.e., athletes with testes, male T levels, and functional androgen receptors. The IOC is waiting for the outcome of the Court of Arbitration for Sport's decision in the case Caster Semenya has brought challenging the regulation to align its rule with the IAAF's.

The IAAF's eligibility rule requires 46 XY males with relevant DSDs who identify or are legally identified as female to drop their T levels below 5 nmol/l for at least 6 months before entering women's competition. In other words, as I explained in the NYT last year, the IAAF's rule "limit[s] entry into women's events to athletes who have testosterone levels that are capable of being produced solely by ovaries."

Both rules permit male-bodied athletes who identify as women to compete in the women's category—they are no longer sex tested and then excluded because they are male—but they cannot enter as superwomen. Here are some useful additional facts:

5 nmol/l represents a generous reading of the outer boundary of the female range, as it captures outlier results from 46 XX females with PCOS (polycystic ovarian syndrome). The normal female range is generally described as no higher than 3 nmol/l.

The time frames in the two rules are based on evidence about how long it takes for the body to wind down the physiological advantages that account for important aspects of the performance gap between male and female athletes. For a summary of these advantages, see my post on Tuesday.

The rules are also designed to address concerns about prior iterations, including the overinclusion of 46 XY males with complete androgen insensitivity (CAIS) when eligibility was established via chromosome analysis; and unnecessary intrusiveness when eligibility was established via external examination and then via the medical standard of care / differential diagnostic for DSD.

In Sex in Sport, I argued that the eligibility rule for both transgender women and males with DSD should require all athletes competing in the women's category to have T levels in the female range. And—consistent with Martina Navratilova's position—I argued that the category should "not be open to intersex and trans athletes who had testes and testosterone in the male range through puberty, since the point of the women's category in elite sport is to provide a space free of competition from athletes with male bodies." Male puberty builds the male body in the respects that matter for sport, including the development of the secondary sex characteristics responsible for the performance gap. Winding down the physiological advantages of male T levels post puberty significantly reduces the male advantage, but it cannot erase it entirely—particularly as to its structural aspects.

Nonetheless, I agree with the IAAF and the IOC that because they are committed both to protecting the category for female-bodied athletes and to including post-pubertal male-bodied athletes who identify as women, using T levels to do this work is the best, i.e., the most accurate and least intrusive, approach. In the language of anti-discrimination law, the policy goals are important, and the means chosen to accomplish them are narrowly tailored and proportional. I testified to this effect in the Semenya case at CAS.

The rules are, of course, subject to criticism. This includes the critique Navratilova and I have made that I've just described and that's further developed in my last two posts. And it includes Semenya's and Rachel McKinnon's critique (summarized in yesterday's post) that gender of rearing and/or identity should be determinative, not biology. Finally, it includes finer points about over- and under-inclusion, and about intrusiveness and proportionality, that I don't have the space to develop further here, and so this list will have to do:

The argument that the rule is over-inclusive is that the governing bodies should but cannot (or do not try to) prove that the performance of a particular male-bodied athlete is due to their male T levels, as opposed to some other endogenous or exogenous factor(s). Generally, this argument is made with respect to male-bodied athletes who are about the same as or only slightly better than the females in the field, e.g., all of the boys and men surrounding Allyson Felix in the 400 meters figure I provided in Tuesday's post. As the detailed public analyses of Semenya's performances demonstrate, this work is not impossible, but I'll leave you to ponder the administrative burdens of such a case-by-case charge. And the categorical problem that is permitting non-elite males to compete without condition for the highest prizes in women's sport.

Arguments that the rules are under-inclusive include concerns that:

  • The 10 nmol/l threshold for transgender women doesn't limit entry into the women's category only to women with T levels in the female range.
  • The 6 months period for at least some categories of DSD athletes is too short.
  • The IAAF's pending DSD regulation only applies to certain "restricted events"—the long sprints and middle distances—when it should apply to all events affected by the performance gap.

Arguments that the rules are intrusive and disproportionate include:

  • Even if confidentiality is maintained, the rules can have the incidental effect of revealing private facts about the athlete's sex or sex traits, or triggering suspicion about those traits. This can be especially damaging when the athlete comes from a traditional society.
  • The rules require athletes to alter their endocrine profiles for purposes of sport when this is not necessarily consistent with their financial, psychological, or physical best interests.
  • Consent to treatment in these circumstances is not truly voluntary, which is to say it's given under a form of duress, becauseo f the athletes' desire to compete in the women's category.