Caster Semenya is congratulated by swimmers Cameron van der Burgh and Chad le Clos after winning Olympic gold. Archive photo: Ihsaan Haffejee
25 February 2019
Caster Semenya’s case has generated global interest. It has raised questions about sex, gender, race and fairness in sport.
This is not the first high-profile South African athlete that the Council for Arbitration in Sport (CAS) has had to consider. It also had to decide whether Oscar Pistorius could participate in “able-bodied events”. The argument, then, against Pistorius’s participation was that his running blades gave him an unfair advantage. But the CAS decided that it did not. Now it has to decide whether Semenya’s biology gives her an unfair advantage in women’s athletics.
The CAS will deliberate on the proposed IAAF policy titled Eligibility regulations for the female classification (athletes with differences of sex development). It will address athletes who do not fit into the binary sexual categories used in sports.
The IAAF has a tough job ensuring a level playing field for all athletes. No doubt enormous effort to balance various interests and rights has gone into producing its policy.
Currently sport recognizes four bases for differentiation: sex (male/female), weight (in sports such as boxing, weightlifting and judo), age (in age-specific competitions), and degree of impairment (in disabled sport). Possibly the only sport that doesn’t differentiate on sex is equestrian.
With established categories an athlete who crosses into another category is penalised by recategorisation. This is understandable for weight and in most cases for age, which is often a challenge in countries with poor birth registration systems. Degree of impairment in disabled sport is a fine balance between an objective and subjective assessment and often the assumed advantage is speculative. A higher category in disabled sport results in significant disadvantage to the athlete.
In the case of sex (not to be confused with gender), the binary division is not helpful for some athletes who cross the established boundaries. This is one of the critical issues that the CAS will consider in its ruling on the IAAF policy: How do we differentiate sex and what is the best marker?
Various methods have been attempted previously – including a nude parade by female athletes to assess external genitalia, analysis of chromosomal patterns, hormonal profiles, assessment of internal genitalia, and a combination of these – but the issue remains fraught because these analyses only throw up greater questions.
Challenges, acknowledged by the IAAF, include mixed chromosomal types, androgen insensitivity syndrome and alpha-reductase deficiency (genetically males but with female characteristics). These cases make binary classification notoriously difficult and sport has no special category for these athletes. Importantly, many of these athletes are unaware of this difference, and are frequently picked up on when they test for elevated testosterone levels.
The IAAF policy attempts to answer this complex question by reducing it to a single end-point variable: female testosterone (fT) levels. The CAS will have to address this challenging question.
Accepting the fact that some females have elevated testosterone (fT) levels for a whole range of reasons, the second question that the CAS will consider is: Do athletes with elevated fT have an advantage?
This is the question that the CAS posed to the IAAF in the Dutee Chand case in 2015 when suspending the IAAF’s previous regulation — Regulations Governing Eligibility of Females with Hyperandrogenism to Compete in Women’s Competition — granting it two years to provide evidence. The court accepted that some female athletes have higher testosterone levels but questioned whether this translates into physiological and performance advantage and therefore could be used as a basis for remedial action by the IAAF.
The IAAF policy is based on population norms (with an allowance for athletes) where male and female levels of testosterone do not overlap. The statistically normal range of testosterone in females is 0.1 – 2 nmol/l and in males it is between 7.7 and 30 nmol/l. Testosterone is considered the major differentiator in sports performance between males and females: male performances exceed female performances between 9-15%.
It is fair to assume that athletes such as Chand and Semenya have levels greater than 10 nmol/l since both were compelled by the IAAF to reduce their levels under the previous policy where the testosterone ceiling for athletes classified as female was pegged at 10 nmol/l (the new policy proposes a more stringent cut-off of 5 nmol/l). It is also fair to assume that the IAAF believes that raised testosterone levels grant female athletes with elevated fT a physiological and performance advantage. A paper published in the British Journal of Sports Medicine by the IAAF that compared athletes with high T to low T, demonstrated an advantage only in the 400m, 400m hurdles, 800m, hammer throw, and pole vault.
This raises the next critical question that the CAS will consider: if there is indeed an advantage, what is its quantum (size)? Of all the events considered, the referenced paper showed an advantage of 2.73% (400 m), 2.78% (400 m hurdles), 1.78% (800 m), 4.53% (hammer throw), and 2.94% (pole vault). It has been estimated that if Semenya was to reduce her T levels, it would affect her performance negatively in the region of 5%. This is assumed to have occurred when, in compliance with the previous IAAF policy, Semenya did attempt to reduce her fT levels through hormonal manipulation.
This would trigger the next question: what is a tolerable level of advantage? What level of fT would not grant a female athlete an unfair advantage? There must be a cut-off defined scientifically by the IAAF. (As mentioned, previously it was 10 nmol/l and now it is 5 nmol/l). But these cut-off value relate to fT levels and not a performance threshold, since the demonstrable advantage of elevated fT does not exceed 5% according to the IAAF’s own study.
The final question is what is considered an unfair advantage? Many accidents of nature are automatically classified as fair advantages: height, muscle composition, arm span, feet size, etc.
But such category comparisons are spurious since sports does not classify according to these criteria but does differentiate by sex. On the one hand athletes with different sexual development have no control over the advantage they gain, unlike doping. On the other hand evidence from the East German dataset of state sponsored doping in the 1970s demonstrated a significant benefit to female athletes doping with testosterone. The CAS may ask: if there is indeed an advantage, where would the line signifying an unfair advantage lie: Would a 10 or 20% advantage be considered unfair?
The CAS will have to grapple with these five questions, and the evidence supporting the various positions in its deliberations. The results will be made public before the end of March 2019.
The weaknesses in the IAAF case have already been exposed: the paper which forms the fundamental basis of their case has fatal flaws and has been panned by academics and statisticians. The authors themselves have acknowledged the flaws, and they had to produce an addendum to the paper.
The first flaw is that there were errors compiling their dataset (with an error rate potentially as high as 30%). The second flaw is that the study compared high fT with low fT athletes, not with normal fT athletes.
There’s also a flaw in the IAAF policy: while the article demonstrated advantage in specific events, the IAAF added additional events to their policy without scientific evidence, including all races between 400m and the one-mile. This was despite there being no evidence of advantage demonstrated in races above 800m. Also, the field events, where there is demonstrable advantage in the IAAF study, are excluded.
But it could also be argued that since testosterone increases muscle strength, power, endurance and speed, enhances cardiorespiratory function, and causes elevated red blood cell volume, it would grant advantage in all events and not only those selected by the IAAF.
Finally, despite its denial the IAAF policy attempts to assign sex, evidenced by the title of its policy: “Eligibility regulations for the female classification”.
The perennial racism question has been bandied around. Notwithstanding a remark by Olympic official Norman Cox, over fifty years ago, that there should be a separate category for black female athletes who are “unfairly advantaged hermaphrodites”, the IAAF has equally applied its standard to white and black athletes. A white athlete will testify as an expert witness for Caster on the negative effects of reducing her testosterone levels. European athletes have fallen foul of the regulations, notwithstanding the fact that the two latest high profile cases happen to be black.
The South African Department of Sport and Recreation has wagered a whopping R25-million on this case. The key challenge lies in CAS balancing genetics, anatomy, physiology, endocrinology and performance, against fairness and gender rights in their decision. Even 29 experts will not enable the CAS to produce a result that is satisfactory to all, and this case may just find itself in a higher civil court if it finds in favour of the IAAF. The emotional cost to Caster Semenya, Dutee Chand and similar athletes has been extremely high.