Australasian Science: Australia's authority on science since 1938

Ten ethical flaws in the Caster Semenya decision on intersex in sport

By Julian Savulescu, Visiting Professor in Biomedical Ethics, Murdoch Children's Research Institute; Distinguished Visiting Profess

Caster Semenya is legally female, was from birth raised as female and identifies as a female. Jon Connell on flickr , CC BY-NC

This essay is part of our occasional series Zoom Out, where authors explore key ideas in science and technology in the broader context of society and humanity.


Middle-distance runner Caster Semenya will need to take hormone-lowering agents, or have surgery, if she wishes to continue her career in her chosen athletic events.

The Court of Arbitration in Sport (CAS) decided last week to uphold a rule requiring athletes with certain forms of what they call “disorders of sex development” (DSD) – more commonly called “intersex” conditions – to lower their testosterone levels in order to still be eligible to compete as women in certain elite races.

The case was brought to CAS by Semenya, as she argued discrimination linked to a 2018 decision preventing some women, including herself, from competing in some female events.

This ruling is flawed. On the basis of science and ethical reasoning, there are ten reasons CAS’s decision does not stand up.

But first let’s take a quick look at the biology involved.




Read more:
Caster Semenya: how much testosterone is too much for a female athlete?


Semenya underwent medical testing in 2009: at the time she was told it was a doping test. The results are confidential, but it has been widely reported that she does have an intersex condition. It seems reasonable to assume she has XY chromosomes, as she is covered by the CAS ruling. Her testosterone levels have not been disclosed, but since the ruling applies to her, they must almost certainly be in what they classify as the “male range”.

According to CAS, the DSD regulations require athletes who want to compete in some female events, who have XY chromosomes and in whom testosterone has a biological effect to reduce their natural testosterone levels to an agreed concentration (below 5 nmol/L).

In women referred to as “46 XY DSD” – the most common intersex condition among female athletes – the presence of a Y chromosome causes the development of testes. These do not descend from the abdomen but do produce testosterone. However the receptors for testosterone are abnormal, with the result that the individual develops as female with a vagina, but no ovaries or uterus. Circulating testosterone may have no biological effect in the case of complete androgen insensitivity syndrome (AIS), or some effect in partial AIS.

Now let’s consider what’s wrong with the ruling.

1. It confuses sex with gender

Sex refers to biology, and gender refers to social role or self-identification. In sport, the definition of male and female used to be based solely on sex. This was assessed anatomically in the 1960s, then by biological tests such as the presence of a structure called a “Barr body” in cells (found only in genetic females), or the gene for testicular development.

Sex determination was abandoned in the 1990s in favour of gender. From the 2000 Sydney Olympics forwards, there were no tests of gender other than self-identification.

Caster Semenya’s gender is uncontroversially female. She is legally female, was from birth raised as female and identifies as a female. So, on the current definition, Semenya is a female. Indeed, there has been no question of her gender.

Sex determination itself is not simple, with chromosomal, gonadal (presence of ovaries or testes), or secondary sex characteristics (physical) all possible definitions that would include or exclude different groups.




Read more:
What makes you a man or a woman? Geneticist Jenny Graves explains


The CAS decision relates to “XY females with disorders of sexual development.” XY denotes the male sex chromosomes. This reverts back to the old biological categories. Behind this ruling is the view that Semenya is really a man competing in the women’s category. This view is embodied beautifully in an article entitled “A victory for female athletes everywhere.”

But Semenya is a female by the rules used by the International Association of Athletics Federations (IAAF) – so she should be allowed to compete to the best of her potential in her category.

An alternative is to retreat to the old sex-based definition based on the presence of a Y chromosome. But that carries its own questions on definitions, and also comes at great political and individual cost. It would imply that Semenya is a male with a disorder of sexual development.

2. It discriminates against some forms of hyperandrogenism

Hyperandrogenism is a term used to describe high levels of testosterone.

But the CAS decision does not cover all forms of hyperandrogenism. It only refers to women who have XY chromosomes, such as partial androgen insensitivity syndrome (AIS).

It does not cover a condition called congenital adrenal hyperplasia, which can cause elevated levels of testosterone in women with XX chromosomes.

The implication is that XX females are real women, while those with XY chromosomes are not.

3. It’s based on inadequate science

The significant problem in partial AIS is that although testosterone is elevated in the blood, the receptors for testosterone do not respond to the hormone in the usual way. That is why these individuals have typical external female physical characteristics.

While the testosterone may have some impact on how the body works, it is impossible to quantify how much effect it is having. For example, the difference testosterone makes between males and females in all events is estimated to be up to 12% (all other items being equal). But Semenya’s best time is only 2% faster than her competitors. It is not possible to determine how much of this 2% is due to testosterone, and how much due to other factors about her as an athlete, or her psychology.

The study on which the current decision is based contains only correlations and is flawed in several ways, with a call for its retraction on scientific grounds. It is a single study, conducted by the IAAF and the full data have not been released for independent replication. The sole ground for the claim that Semenya derives “material androgenizing effect” (that is, biological impact) appears to be the “statistical over-representation of female athletes with 46 XY DSD” in the relevant events, as documented in this single, poorly conducted study.

Even if Semenya’s times were to drop after the reduction of testosterone, this could be a side-effect of the drugs used to reduce testosterone, or a function of reductions in mental or physical functions which are themselves legitimate entitlements of the athlete.

Her body has grown up in the presence of a certain level of testosterone of uncertain function. Our bodies are complex, and still poorly understood. A change of this kind may lead to unexpected results. Some of these reductions in functions may be unjust.




Read more:
Testosterone: why defining a 'normal' level is hard to do


No one has given a complete description of the role of testosterone in someone like Semenya, nor how much it ought to be reduced to achieve a supposedly fair outcome. The comparisons are only with XX chromosome women, who have a very different physiology and normal functioning testosterone receptors.

Put simply, a level of 5 nMol/L testosterone is meaningless in Semenya’s case because the receptors are not responding in the usual way. It does not achieve a “level hormonal playing field”.

This is an example of “decimal point science smokescreen.” There is the impression of much greater confidence and sensitivity than the science warrants by appealing to figures with multiple decimal points. The science around testosterone in intersex conditions is poorly understood, let alone as it applies to individuals. This is a level chosen for convenience, not a level that will negate any perceived advantage, but go no further.

4. It’s inconsistent with values of sport and human rights

The self-professed values of sport include the development of one’s own talent .

Yet Semenya is asked to cobble her natural potential as a female competitor. She must take risky biological interventions to reduce her performance.

The United Nations Human Rights Council has stated that the regulations contravene human rights “including the right to equality and non-discrimination […] and full respect for the dignity, bodily integrity and bodily autonomy of the person”.




Read more:
It's not clear where human rights fit in the legal ruling on athlete Caster Semenya


5. It’s inconsistent with treatment of other athletes

Other women with disorders resulting in higher than expected levels of testosterone, such as congenital adrenal hyperplasia, are not required to reduce their biological advantage.

Competitors with genetic mutations causing increases in red blood cell mass, and who experience enhanced oxygen-carrying capacity as a result, are not required to reduce their biological levels.

The Finnish skier Eero Mäntyranta had a genetic mutation that boosted his red blood cell count by 25-50% (he produced more blood hormone erythropoetin, or EPO). He and won several Olympic medals with this natural form of doping.

6. It’s unjust

The decision is unjust in several ways.

Firstly, it was the IAAF which moved from sex to gender definition of female in 1990s. Semenya has entered competition, trained and competed fairly under the rules. To change them now will be undermine her capacity to compete, work and live, after a lifetime of investment.

If the rules are to be changed, they should not affect athletes who agreed to the current rules, but future athletes. There should be a “grandmother clause” for current athletes, like Semenya or else they are unfairly burdened by the bungles of the IAAF. Even if these rules could be considered justified, they should apply to future athletes as soon as possible after puberty.

Secondly, justice is about giving priority to the worst off in our society – but this ruling adds disadvantage to the worst off. Those with intersex conditions are already stigmatised, discriminated against, in many cases cannot bear children even if they want to. They are the socially disadvantaged. This ruling adds further discrimination and disadvantage.

Thirdly, it sets back integration of intersex people, by stigmatising and marginalising them. We have told them: be yourself, society will accept you. But this sends the message: you are really male, we don’t accept you, you should be castrated.

7. It is an inappropriate reaction to fear of a ‘slippery slope’

At the heart of this decision is the fear of displacement of cisgender women on the podia by increasing debate over transgender athletes. The concern is that if “XY females” are allowed to compete in the female category, formerly male transgender females will follow and rob cisgender women of their medals.




Read more:
Explainer: what does it mean to be 'cisgender'?


This is a separate issue. Transgender athletes have normal testosterone receptors and would have grown up in the presence of male levels of testosterone acting on normal receptors. Intersex athletes have not grown up in this way and are typically raised as female.

The perceived problem of transgender domination of female sports can be dealt with by separate rules that do not disadvantage existing intersex athletes, though they will raise contentious issues of their own.

8. It is disproportionate and unreasonable

All methods of reducing testosterone involve some risk. For example, the administration of high-dose birth control medication involves risk of clots, including fatal lung clots.




Read more:
How to choose the right contraceptive pill for you


These interventions interfere with a normally functioning organism for highly uncertain benefits to other people. This is disproportionate and unreasonable.

9. It can’t be implemented

The World Medical Association has advised doctors not to administer testosterone-lowering interventions, describing the regulation as “contrary to international medical ethics and human rights standards”.

Their use would be “off label” and is for purposes other than the athlete’s health. The rules involve “strict liability” which means the athlete is responsible for any failure to comply, even if unintentional and outside of the athlete’s control.

10. There are fairer, safer alternatives

I have argued athletes should be able take performance-enhancing substances within the normal physiological range. This would mean cisgender female athletes could take testosterone up to 5 nMol/L. This would reduce any advantage Semenya may have.

It would also deal with the problem that up to 40% of elite athletes are currently doping anyway. Semenya received the London 2012 800m gold medal after the original winner was disqualified for doping. It is highly likely that some of her current competitors are also doping.

No doubt part of the resistance to allowing Semenya to “naturally dope” is that it will encourage other athletes to engage in doping. But they already are, and a better approach to “de-enhancing” Semenya is to regulate and monitor the enhancement of other athletes.

Spectacular fail

Rarely does a public policy fail so spectacularly on so many ethical grounds.

CAS acknowledged that its decision constituted discrimination:

“The panel found that the DSD Regulations are discriminatory but the majority of the panel found that, on the basis of the evidence submitted by the parties, such discrimination is a necessary, reasonable and proportionate means of achieving the IAAF’s aim of preserving the integrity of female athletics in the restricted events.”

The UNHRC has refuted this claim of proportionality: “there is no clear relationship of proportionality between the aim of the regulations and the proposed measures and their impact”.

This ruling is neither necessary, reasonable nor proportionate. It is simply unjust discrimination.




Read more:
Caster Semenya's impossible situation: Testosterone gets special scrutiny but doesn't necessarily make her faster


Thanks to Michelle Telfer and Ken Pang for comments

This article builds on arguments presented in the paper Time to re-evaluate gender segregation in athletics?.

The Conversation

Julian Savulescu receives funding from Uehiro Foundation on Ethics and Education and the Wellcome Trust.


Originally published in The Conversation.